Pages:
1
2
3
4 |
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
Exploding wires
This post is prompted in response to the post here
http://www.sciencemadness.org/talk/viewthread.php?tid=11756
much of this is informatively discussed at length in the reference
given by damn2 here _
http://www.sciencemadness.org/talk/viewthread.php?tid=11756&...
Succinctly outlined here _
http://www.teledynerisi.com/0products/8td/page06.html
http://www.teledynerisi.com/0products/8td/index.html
http://www.teledynerisi.com/0products/index.html
It was my understanding the notion of exploding electronic board resistors
was put to rest in this other thread _
http://www.sciencemadness.org/talk/viewthread.php?tid=7266&a...
here we go again _
First some definitions and misconceptions.
There are two applications of electric firing and these are mutually exclusive,
ignition, or alternatively , detonation ( of secondary explosives )
'Xenoid' describes above an elaborate apparatus to ignite rocket motors. I can't help thinking
so much effort for so little gain. I can't think of any reason to do so in quite that way except
that is just how he likes to do it. One can very expediently make a single use improvised
" Hot Bridge Wire " squib ignitor made from a flashlight or torch light bulb. Heating it with a
blowtorch to poke it open with a nail , and fill it with powdered match heads. The flashlight
itself is the only power supply needed and an arbitrary length of hookup wire.
Bulbs are manufactured with small values of resistance, which of course increases many fold
when turned on , limiting the working current. This complicates calculation of adding the
measured ohms of the firing cable. To determine the voltage needed to obtain the working
current its easiest to just test the bulb connected to the firing cable to light it before making
it into an ignitor. Using one more 1.5 volt battery than what is normally required should be
sufficient. Moderately higher voltage drives excessive current through the bulb lessening
it's life in normal use , but as it burns brighter that may be advantageous in this application.
http://www.bulbtown.com/1_0_2_99_Amp_Light_Bulbs_s/665.htm
http://www.bulbtown.com/1_4_9_Volt_Light_Bulbs_s/764.htm
[url] [/url]
Primary ( initiating ) explosives ( substituted for match heads ) can be detonated in this same
way. http://en.wikipedia.org/wiki/Blasting_cap
Multiple charges are routinely connected by detonation cord set off from a single blasting cap.
Apart from wearing high impact resistant eyeglasses and face visor , ballistic vest , and tongs
or forceps for handling , thoughtful precaution is necessary when making an electric blasting
cap as the wire leads are effectively a dipole antenna and as such susceptible to radio frequency
broadcast in the vicinity which could resonantly induce current sufficient to heat the bridgewire
( filament in this case ). From the time it is first loaded with explosive and initially been tested
for continuity ( read the following procedure ) to the time when it will be used , the exposed
lead ends must remain twisted together. To ensure there is no hangfire , immediately prior to
use ( by having the cap inside a pit dug into the ground covered with a sand bag ) it is tested
for continuity , a tricky feat that requires a milliampere low voltage source of just under 1 volt.
The cap lead ends are then connected at this time to the cable or wire that will be used for
firing it, which is also first tested for continuity being already shorted and disconnected at the
power end. The final act is to bring the cap from under the sandbag to the main charge of
secondary explosive and inserted into a previously formed opening there , then promptly leaving
the scene to a safe area. The last thing before firing is to un-short the cable end and test again
for continuity before hooking up to power for firing . This is to distinguish an electrical failure
mode from a dud hang fire which may perhaps be still smoldering , ready to explode , in that
case a wait of 15 minutes or more is warranted before approaching the item for inspection.
For a high degree of safety one could instead electrically light a fuse first to then set off the
blasting cap. ( Individuals who snicker and think all this is excessive , are by definition kwels )
[url] [/url]
__________________________
The EBW ( exploding bridge wire ) is an altogether different device that initiates detonation of
secondary explosives in direct contact without the use of primary explosives. This is accomplished
by vaporizing a small metal wire by the rapid Joule heating produced by a nearly instantaneous
( very low impedance ) applied electric current. Immediately evident by the very L O U D bang
it makes , note there will be little sound if the wire only melts and drips away much as a fuse.
( See here fusing current ) - http://www.ndrr.com/rmr_faq/Introduction/Fusing-Currents.htm
and also this - http://www.litz-wire.com/New%20PDFs/Fusing_Currents_R2.011609.pdf
* Note that both the above references omit to state that the wire diameter must be C U B E D
. before taking the square root , so , √ ( D X D X D ) , only then multiplied by the coefficient " A "
EBW devices must be within arms reach of the pulsed power source and supplied by a short
low impedance cable typically within 40 - 50 cm. This can be :
Solid copper core RG-8U Coaxial cable ,
[url] [/url]
Braided or Rutherford Litzendraht ( Litz ) wire
[url] [/url]
http://en.wikipedia.org/wiki/Litz_wire , http://www.litz-wire.com/applications.html ,
http://www.hmwire.com , http://www.hmwire.com/sm.html ,
Goertz flat ribbon cable ,
[url] [/url]
similar available plain and much cheaper from _ http://www.alphacoredirect.com/contents/en-us/d53.html
or 2 plys of 12AWG from http://www.decorp.com
[url] [/url][url] [/url]
http://www.flatwirestore.com/mm5/merchant.mvc?Screen=CTGY&Store...
http://www.tycoelectronics.com/catalog/cinf/en/c/11930/1486
http://www.tycoelectronics.com/catalog/minf/en/152
http://www.ampnetconnect.com/product_groups.asp?grp_id=1660&pat...
http://www.ampnetconnect.com/product_cut_sheet.asp?grp_id=2285&...
even 2 edgewise parallel flat cross section solid grounding strap would do.
http://www.keison.co.uk/furse/conductors_flat_tape.htm
Fralock transformer coil copper tape
http://www.surplussales.com/RF/RFSilv-CoppS.html
_____
Perversely , the uninitiated continue to believe that the bridge wire itself is a resistive element ,
- - W R O N G - -
A bridge wire is just that , an UNINSULATED wire having as low a DC current resistance as
possible. Due to the extremely short time rise of the current , initially conduction occurs only
along the surface , known as " skin effect ", as the wire heats up the resistance increases,
then as the wire melts the resistance drops again and conduction resumes now through the
entire cross section and continues to increase as it becomes vapor. High resistance material
inhibits this progression so that the surface is cooked off forming a vapor boundary layer which
is more conductive , dissipating the discharge as a heat flash , not shock.* Note this is a
principle reason why very fine wire is used in pulses of very short time frames , the current
simply cannot penetrate into the depths of the wire where it needs to be to rapidly heat the wire.
The applicable equation of state is , I = V / R , current I = amps , V =
volts , R = ohms
Arithmetically R must be as small as possible to maximize the current. This is why kilovolts
are typically applied , to assure a huge current surge and rise time.
[url] [/url]
Another misconception is the available energy provided to the wire.
A capacitor's stored energy according to E = CV^2 /2 , E = Joules , C =
farads , V = volts
83 Joules = (.00082 F X 450 X 450 ) / 2
may be adequate to the task were it all to arrive instantaneously - which it cannot , being
a function of time and restrictive circuit characteristics. The determinant will be Power ,
the rate at which the energy can be delivered by the cable to the EBW. Because this is a
rate per second , at initial discharge apparent power seems as if megawatts are on tap.
Read Overview here _http://en.wikipedia.org/wiki/Pulsed_power
Dividing 83 Joules by the discharge time .000160 sec = 0.518 Megawatt average power
discharge time is calculated as follows : see - RC time constant , below
[url] [/url]
- RC time constant
Determining the time during which the initial discharge occurs is given by
TC = RC , R = ohms , C = farads ,
TC means Time Constant , a factor which relates the rate at which the capacitor discharges.
Specifically , 63 percent of charged voltage during the first TC leaving 37 percent remaining ,
86 percent of the original after 2 X TC , leaving 13 percent of the total remaining , after the
third TC 95 percent discharge has occurred. Complete discharge is assumed after 5 X TC.
See middle of page here - http://www.kpsec.freeuk.com/capacit.htm
also bottom of this page - http://www.bcae1.com/capacitr.htm
RC TC in depth - http://www.electronics-tutorials.ws/rc/rc_1.html
RC Time Constant Calculator - http://www.cvs1.uklinux.net/cgi-bin/calculators/time_const.cgi
Capacitors exhibit resistance which limits their rate of discharge, known as ESR ( Equivalent
Series Resistance ). Smaller ESR means a shorter RC Time Constant and more power
( energy delivered per unit time ). For the capacitor type and size range applicable typically
around 0.1 ohm or less. Reducing ESR shortens the TC and the current rise time without needing
to have higher voltage to compensate for higher resistance.Very short pulse durations
are only acheived by capacitors having exceptionally low ESR. As these do not have large
capacitance, they must have a very high voltage to store significant energy.
What is a Capacitor ? - http://www.sofia.usra.edu/Edu/materials/activeAstronomy/sec5_capaci...
http://www.bychoice.com/capacitor_DF.pdf
http://www.illinoiscapacitor.com/uploads/papers_application/F8CD49C...
http://www.cartage.org.lb/en/themes/sciences/physics/electromagneti...
http://www.cartage.org.lb/en/themes/sciences/physics/electromagneti...
Equivalent Series Resistance (ESR) of Capacitors - http://www.low-esr.com/QT_LowESR.pdf
Equivalent Series Resistance of Tantulum Capacitors , Both the following PDF's are the same
- http://www.avx.com/docs/techinfo/eqtant.pdf
- http://www.avxtantalum.com/pdf/EQTANT.PDF
- In the opening post of this thread -> http://www.sciencemadness.org/talk/viewthread.php?tid=11756
given 820 uF or .00082 F, assuming a 2 cm 28 AWG EBW
Resistance of EBW = .004 ohm , plus Resistance of cable = .0267 ohm ( sum of core and shields
resistance of 3 meters 10 AWG RG-8U coax , see footnote ## ) , estimated ESR
0.1 ohm ( note that ESR is 3 times greater than the rest of the circuit resistance ) the total
circuit resistance = .131 ohm then TC = RC = (.131 )(.00082 ) =.000107 ~ 107 microseconds
- discharged at 450 volts after one TC the remaining voltage is 167 ( 37 % ) , so by E = CV^2/ 2
( .00082 X 167 X 167 ) / 2 = 11.5 Joules , dividing by the rating of 83 Joules ( 11.5 / 83 ) = 13.8 %
100 % - 13.8 % = 86.2 % of the energy is delivered during the first 107 microseconds. The magic
factor is 1.5 times the TC when the remaining charge is 100 volts ( 22.2 % ) 95 % of the energy
of the capacitor has gone into the line in just 160 millionths of a second. What this means is if
conduction is still occurring only along the surface of the EBW , there will be no loud B A N G !
http://www.probertencyclopaedia.com/cgi-bin/res.pl?keyword=Effectiv...
- Given that frequency is the reciprocal of time , f = 1/ t , then the 160 millionths of a second
translates to a frequency of 6.25 kHz. Well below the 170 kHz where skin effect reduces
cross sectional saturation. Look up 28 AWG , here -> http://www.powerstream.com/Wire_Size.htm
Note that this value would be a mere 2.6 kHz for the 10 AWG coaxial core were it just bare wire
instead of coaxial cable.
_____
Joules needed for vaporization of a 2 cm 28 AWG wire = ( weight of EBW / molar weight Cu ) X
( * *Enthalpy of vaporization of Copper 300 KJ / mol + Heat of fusion 13.1 KJ / mol ) =
( .0145 gm / 63.546 gm/mol ) X ( 313100 J/mol ) = 71.4 Joules
By comparison an equivalent weight of TNT yields 65.5 Joules (.0145 X 4519 Joules/gm )
* * From http://www.webelements.com/copper/thermochemistry.html
weight of bridge wire is easily obtained from this applet here _
http://circuitcalculator.com/wordpress/2007/09/20/wire-parameter-calculator
Resistance of EBW = .004 ohm , Resistance of cable = .0267 ohm ( see footnote ## )
The energy is consumed along the entire circuit's resistance. The portion which will act on
the EBW is: .004 / (.004 +.0267 ) = 0.03 , alternatively a 25 cm length of cable is .0022 ohm
recalculating , .004 / (.004 +.0022 ) = 0.645 , then .645 X 83 Joules = 53.5 Joules
An 83 Joule rated capacitor will not be adequate with inherent circuit losses , and still
be marginal with somewhat thinner wire. A thinner wire has less mass and will require less
energy to vaporize while presenting more resistance and consume a greater portion of the
available energy in the circuit. Worth a try with an aluminum foil ribbon and very short cable
run ~ 25 cm. Remember it must make a very loud bang , if it does not sound like a rifle shot
you have not achieved detonation status.
12AX7 observed in this other post
http://www.sciencemadness.org/talk/viewthread.php?tid=6032&a...
the only energy that matters in a system is that which actually achieves the intent.
_____
Capacitive reactance and Inductive reactance the two components of circuit impedance
behave as an additional form of resistance which interferes with rapid discharge of the
capacitor and transmission of that power as a short pulse. Coaxial power cable limits this
to a maximum of it's characteristic impedance of 50 ohms regardless of length. This value
is only a transmission line design parameter , whereas the actual line impedance of your
particular firing cable depends entirely on the wavelength of the discharge waveform
relative to the cables length , and what is attached on the ends. Given that wavelength
is the speed of light ' c ' divided by the frequency , λ = c / f, or alternatively
multiplying
by the time of discharge , wavelength = λ = c X t = ( 300,000,000 X .00016 ) = 48000 meters.
The actual wavelength within the cable is always shorter than this because it slows down
( similar to light when it enters a denser medium ) known as " velocity propagation factor ".
For RG-8U it is 0.84 of light speed. A pulse that is lengthy due to a long discharge time
becomes shorter by the propagation delay inherent to the cable. A cable of just 3 meters
behaves here as a short circuit , exhibiting practically no line impedance to or attenuation
of the discharge pulse. See footnote below - IMPEDANCE EXPLAINED -
Xc = .159 / f C , Xc = capacitive reactance of the capacitor expressed as ohms
f = frequency , C = farads. So , Xc = .159 / ( 6250 X .00082 ) = .031 ohm ( vanishingly small )
* Note the absurdly overpriced Goertz " speaker cable " exhibits a characteristic impedance
of just 2 to 4 ohms by further reducing line induction.
Unrelated to the present topic , a rational perspective on such expensive attributes.
http://www.leedsradio.com/technical/snakeoil-cables.html
Ideally the EBW will have melted by full discharge as the current surge drives through
the lowering resistance and vaporizes the wire. We have observed that the ESR of the
capacitor is the most pronounced inhibitor of circuit performance , throttling the power
available. By the end of the next post it will be evident how the effect of ESR can be
minimized to improve the response and performance by at least two fold.
_____
The power feed cable as selected below must be of large gauge to minimize resistance
( ## 3 meter length of RG-8U coax ) or a braid of Litz cable. Welding cable can
suffice for only very short lengths but the induction becomes significant in the circuits
performance. This is because it acts as a transformer primary , magnetically coupling
with anything electrically conductive even moist ground.
Given everything going for it one can see why this has no application other than for
use in ordnance where the EBW can be snug up against the pulse generator.
A setup for outdoor use will need to have the pulse generator close by which puts it
at hazard if used to detonate explosives. Charging and firing control can be done from
a distance with a 3 wire extension cable. ( See the circuit diagrams - next post - below )
* 28 AWG is a common size used in " wire wrap " electronic board prototyping.
Click thumbnails then click again the pictures here -> http://en.wikipedia.org/wiki/Wire_wrap
My suggestion of 28 AWG is 3 to 4 times thicker than what is usually used for an EBW,
and with much bigger caps. Given that the 1 mil to 3 mil wire otherwise used is AWG 40
and smaller which is much finer than hair one should not stray too far from 28 AWG wire
as selected above. Much smaller wire size as noted requires less effort to explode but
becomes increasingly more difficult to physically work with and prepare. Larger wire
naturally requires greater energy to explode but more importantly the lower frequency
at which skin effect becomes significant will affect cross sectional current saturation.
Fine wire is cheaply obtained from ordinary stranded electric line cord. Count the number
of strands in the wire bundle of a known gauge ( 14 AWG being common ) to derive the
single strand size verified here _ http://www.interstatewire.com/WireTable.htm
http://www.seas.gwu.edu/~ecelabs/appnotes/PDF/techdat/swc.pdf
==> One experimenter's results with videos
http://members.tm.net/lapointe/Wire_Explosions.html
Viewing this setup one immediately can see despite success with fine copper wire that
such a large loop of single conductor must sap considerable energy by inductive losses.
This can't be stressed enough , the faster the current rise time , the greater the need
to minimize induction in the circuit !
Viewing the following videos I ask myself what's wrong with this picture. The widely
separated cables is what , which is in effect a single turn solenoid the induction of
which is responsible for the mediocre results observed.
Capacitor like woelen's - explosion -
http://www.youtube.com/watch?v=CY8bkf7QcVE
same guy with new General Atomics capacitor
http://www.youtube.com/watch?v=QMWRvAI4o2E
Banks of capacitors in parallel have one fatal flaw :
ALL of the energy will discharge into the capacitor that fails destroying it and
perhaps damaging those that are alongside.
[url] [/url]
It's essential to design a capacitor bank so that the individual caps can have at
least one terminal disconnected from the common conductor with the other caps
so that it can be individually tested for "leakage current" and signs of impending failure.
http://www.fullnet.com/~tomg/esrscope.htm
Invest in a good RCL bridge meter that can measure this. Better yet get it all in one
package , the most prized of all , Sencore LC 75, 76, 77, 101, 102, 103 ( may be
seen on EBay for ~ 150 to 250 dollars and up ) These can even re-form deteriorated
aluminum electrolytics. The oxide insulating layer will tend to deteriorate in the absence
of a sufficient rejuvenating voltage, and eventually the capacitor will lose its ability to
withstand voltage if voltage is not applied. A capacitor to which this has happened can
often be "reformed" by connecting it to a voltage source through a resistor and allowing
the resulting current to slowly restore the oxide layer.
( See my attachment Equivalent Series Resistance & Maximum Leakage.rtf
next post below in Pulse Power - Firing circuits )
http://bama.edebris.com/download/sencore/lc102/LC102.pdf
Cheaper expedient methods (with oscilloscope ) are available
http://octopus.freeyellow.com/esr.html
http://octopus.freeyellow.com/99.html
All the way at the bottom see ' Scope ESR ' here _
http://www.anatekcorp.com/ttg/tiptrick.htm
_____
- IMPEDANCE EXPLAINED -
http://www.eskimo.com/~ddf/Theory/Char_Z.html
http://www.speedingedge.com/PDF-Files/BTS002_Characteristic_Impedan...
The activity in a 3 meter cable at the discharge frequency of 6.25 kHz being very much
less than a quarter wavelength is seen at the extreme right side of this diagram here_
http://www.tpub.com/content/neets/14182/css/14182_150.htm
http://www.allaboutcircuits.com/vol_2/chpt_14/5.html
http://www.epanorama.net/documents/wiring/cable_impedance.html , See the following
- Cables characteristics at high frequencies
- Why attenuation figures tend to increase with increasing frequency ?
Impedance calculations ( scroll down )
http://ocarc.ca/coax.htm
http://www.mantaro.com/resources/impedance_calculator.htm
http://hamradio.arc.nasa.gov/coaxcableloss.html
http://74.125.93.104/search?q=cache:xfy31beGKowJ:hamradio.arc.nasa.gov/coaxcableloss.html+%22Belden+9913%22&cd=58&hl=en&ct=c
lnk&gl=us
FOOTNOTE
## 3 meter length of Belden 9913 10 AWG solid copper core RG-8U coax , .0089 ohms / meter
( sum of both core and shield resistance )
RF9913 - $ 0.76 cents ft ( recommended vendor )
- http://www.radiobooks.com/products/rf910.htm
- http://www.radiobooks.com/rwcoax.htm
25 ft cable + UHF connectors, or custom size ready made - $39.95
- http://www.radiobooks.com/products/ca9913.htm
An RG-213 stranded conductor 7 X 21 AWG + UHF connectors, ready made cable
may be suitable to sustain a multi kilojoule pulse for modest power throughput
( rated for 5 kilowatt ( 250 V X 20 amp ) due to the effective cross section being
half that of solid core RG-8U, is cheaply available here _
http://www.buckscom.com/catalog/index.php?main_page=popup_image&...
http://www.buckscom.com/catalog/index.php?main_page=product_info&am...
This one is preferable for applications at kilovolts having a very short time pulse.
Belden 9913 - 0.79 cents/foot ( alternative more expensive vendors )
http://www.theantennafarm.com/catalog/index.php?main_page=product_i...
http://www.signalpros.com/index_files/Page3071.htm
http://www.progressive-concepts.com/info/item.html?id=68
RG-8U Cable specifications and data -
http://www.contactcables.com/products/Belden/Belden9913datasheet.pd...
http://www.alliedelec.com/Images/Products/Datasheets/BM/BELDEN_WIRE...
http://sigma.octopart.com/30985/datasheet/Belden-9913-010500.pdf
http://www.belden.com/pdfs/03Belden_Master_Catalog/06Coaxial_Cables...
( page 6.70 )
Fusing current
http://www.ndrr.com/rmr_faq/Introduction/Fusing-Currents.htm
http://www.litz-wire.com/New%20PDFs/Fusing_Currents_R2.011609.pdf
http://www-d0.fnal.gov/hardware/cal/lvps_info/engineering/wirefusin...
http://www.interfacebus.com/Reference_Cable_AWG_Sizes.html
http://www.interfacebus.com/Aluminum_Wire_AWG_Size.html
Wire chart
Solid conductor
http://amasci.com/tesla/wire1.html
http://www.powerstream.com/Wire_Size.htm
Stranded conductor
http://www.interstatewire.com/WireTable.htm
http://www.seas.gwu.edu/~ecelabs/appnotes/PDF/techdat/swc.pdf
Wire calculator
http://circuitcalculator.com/wordpress/2007/09/20/wire-parameter-calculator
Unit conversion
http://www.allconversions.com
__________________________________________
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
Pulse Power - Firing circuits
All the previous analysis presupposes the capacitor is a stand alone power source connected in a
closed loop with the EBW. Energy storage capacitors as these below, while suitable for exploding
wires , are large and unwieldy. A high voltage power supply is also needed to charge them.
240 uF 5000 Volts - 3000 Joules
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=350128516137
[url] [/url]
http://www.gaep.com/capacitors.html
To provide a large power pulse , a current surge generator is a better arrangement.
By intentionally causing the a.c. power line to short circuit through the EBW , this enables
a very large power pulse to be directly applied with minimal , relatively inexpensive hardware.
In this scheme the capacitor voltage discharges through a spark gap to complete a circuit
with the a.c. power line , resulting in conduction of a huge current surge ( termed " in rush
current ") through the firing cable. This is known as '" Power OR " mode of operation , a
" shunt " diode channels this " follow current " from the a.c. line conducted by the arc.
See F O O T N O T E S bottom of the next post following
Also read here beginning the 3rd paragraph at left _
http://img11.imagevenue.com/img.php?loc=loc284&image=05761_Magnetizing_.jpg
A current surge of hundreds of amps during one half cycle of the a.c. power line
frequency can be tolerated only as a self terminated transient event to prevent damage
to the firing circuit and in particular the a.c. power line. The EBW additionally serves as
the circuit protection fuse , taking only just the power used to explode.
A 300 amp surge through a 2 cm 28 AWG wire by P = R I^2 = .004 ( 300 X 300 ) = 360
watts ,
more than 5 times what was determined necessary to be vaporized. Note that much more
power is consumed by the rest of the power line circuit which heats up considerably though
manageably in a single shot event. Line transients as these are commonly handled by
commercial surge suppressors discussed here - http://www.zerosurge.com/PDF/PQ94.pdf
_____
Drawing ' A ' ( attachment ) restricts the capacitor current to only the firing circuit by having
the capacitor ' C ' parallel with the a.c. line. Synchronizing a.c. line commutation to occur
in phase with the capacitive discharge is facilitated by a triggered spark gap or gas tube ' gg '
as part of the capacitor EBW circuit - Drawing ' B '
See footnote - Triggered Switches , in the post following this one and
- http://www.angelfire.com/80s/sixmhz/trigatron.html
This allows the peak voltage of the a.c. line to trigger the gas gap into conduction.
A diode ' d ', restricts triggering to the phase when the a.c. line voltage has
the same polarity as the caps. This arrangement does not receive the power boost
provided by having the capacitors in series with the a.c. line and will not be considered further.
3 terminal gas tube - 2 to be used in parallel & need to have leads thickened - $ 3.02 each
- http://www.mouser.com/Search/Refine.aspx?Keyword=871-B88069X8440B20...
- http://www.epcos.com/inf/100/ds/t81a90xx8440b202.pdf
Drawing ' C ' shows the capacitor ' C ' in series with the EBW and the a.c. power line. The
problem of a.c. commutation synchronization is solved by a spark gap ' G ' which conducts
in phase at peak a.c. power line voltage. Having the a.c. power line in series with the
capacitor has an additional benefit that the a.c. line voltage and capacitor voltage add
together increasing the voltage to the EBW , improving the current rise time. A large
diode ' D ' provides a path for the resulting current surge generated in the firing circuit.
_____
* * * Note that all the following circuits are divided into a charging and switching part shown
on the left side of each drawing , and the power storage part shown on the right side of the
same drawing. The entire circuit is interconnected across some distance with a 3 conductor
cable serving as an extension cord. Each conductor is labeled
' Top ' , ' Mid ' , ' Bot ' ( Top , Middle , Bottom )
This arrangement also charges the extension cable to the same static voltage as the capacitors.
which adds the cable's capacitance to the firing circuit , reducing the damping at discharge.
' Top ' and ' Mid ' conductors flow electron current to the left.
' Bot ' conductor flows current to the right.
( This is true for all these circuits , although charging phases may have different current paths. )
Suitable cable is of the type used for connecting RV ( Recreational Vehicles )
and trailers to the electric utility main grid , RV 30 Amp Generator Cord , 10/3
( The leading number is the gauge AWG followed by the number of conductors )
Characteristic impedance for twisted pair of this type would be somewhere 100 to 200 ohms
See http://www.mantaro.com/resources/impedance_calculator.htm
Conductor inductance calculations ( nice to know not really all that useful here )
http://www.kolb-net.de/pulsedpower.html - - scroll down
A very good price is less than a dollar a foot. 500 foot loop circuit resistance = .509 ohm
http://wesbellwireandcable.com/PortableCord.html
http://wesbellwireandcable.com/SOOW/SOOW10-3.html - 250 ft - $ 247.50
http://www.americord.com/bulk-cable/prod_537.html - 250 ft - $ 217.50
Cord Nomenclature
http://www.americord.com/glossary
http://www.cables.com.tw/yuefeng/glossary.htm
Crimped and soldered terminal lugs with screws are to be used to hook up the
3 conductor connecting cable.
Plugs and receptacles to connect the charging and switching circuit to the a.c line.
http://en.wikipedia.org/wiki/IEC_connector , IEC 320 C19 / C20 , cords are not available
ready made in 30 amp rating, but can certainly be wired with 10 AWG wire and provide a
secure connection , the pins are robust enough and provide better moisture seal.
[url] [url]
The following rewireable plug must have the 2 piece mating seam sealed with epoxy
http://www.mouser.com/Search/Refine.aspx?Keyword=4789.1200
http://www.schurterinc.com/pdf/english/typ_4789.pdf
http://www.mouser.com/Search/Refine.aspx?Keyword=4797.0015
http://www.schurterinc.com/pdf/english/typ_4797.pdf
http://www.apcmedia.com/salestools/SADE-5TNRML_R0_EN.pdf
http://www.feller-at.com/English/Feller-GP-E.htm
NEMA Standard connectors ( If you must , use these , but are comparatively very large )
http://www.hubbellcatalog.com/wiring/catalogpages/section-b.pdf
see B-17 to B-19
_____
Drawing ' D ' shows a practical circuit having the minimum components.
Single Pole Single Throw / NO ( normally open ) switch ' S ' connects power to charge
the capacitor(s) ' C '. Electron current flows through resistance ' R ', and parallel
resistor ' r ' with series ' LED ' light emitting diode , through diode ' dx ' and
' Mid ' conductor , collecting in the negative ( - ) terminal of the the capacitor(s) ' C ',
displacing current out of the positive ( + ) terminal flowing back through ' Top ' conductor
and shunt diode ' D ' to the a.c. source. As it is charged , light emitting diode ' LED '
momentarily flickers on then darkens indicating that charging has occurred. Resistance ' R '
is to protect the small diode ' dx ' from excessive current when charging starts ).
* * * For added safety a 10 amp fast acting circuit breaker ' X ' provides redundant
protection for domestic wiring , serving also as a switch it now does the actual firing
when turned on. The charging circuit is F I R S T switched off by normally open switch ' S '
before the charged circuit can be fired.
Circuit breaker ' X ' fires the circuit by applying the a.c. line voltage in series with the cap(s) ,
effectively doubling the circuit voltage which arcs into conduction the Gas Gap arrestor ' gg '
near the peak of the a.c. sine wave. ' Top ' conductor channels the resulting current surge
left and down shunt diode ' D ' to the a.c. source.
In practice the 325 volt peak is minimal and very marginal to produce an arc across a spark gap ,
better is a " Gas Gap Arrestor " ( use 2 or even 3 in parallel and solder thicker leads on each )
' gg ' - Gas Gap Arrestor - 444-GT-230L - $ 3.09
http://www.mouser.com/Search/Refine.aspx?Keyword=444-GT-230L
http://www.mouser.com/catalog/specsheets/XC-600002.pdf
' gg ' - Gas Gap Arrestor - (RMO) CG2-230L , middle of this page - $ 2.75
http://www.surplussales.com/Semiconductors/SemiCTransSup.html
[url] [/url]
http://sigma.octopart.com/510848/datasheet/Littelfuse-CG2230L.pdf
http://octopart.com/search?q=CG2-230L , cheapest at Mouser
http://www.mouser.com/Search/Refine.aspx?N=254139&Keyword=CG2-2... - $ 1.89
' X ' - Single Pole circuit breaker - $ 10.99
http://www.mouser.com/Search/Refine.aspx?Keyword=845-1BU10R
http://www.altechcorp.com/PDFS/R-Series.pdf
http://www.mouser.com/catalog/637/1795.pdf
- @ $2 each set of 3 breakers held by pins this has two 15 amp breakers which can work
and is cheap enough to try out but needs to be dismantled
http://www.sciplus.com/singleItem.cfm/terms/348
[url] [/url]
' S ' - Switch SPST / no - (SWP) 1175 , lower third this page - $ 3.75
http://www.surplussales.com/Switches/SWPushB-1.html
[url] [/url][url] [/url]
' R ' - 30 ohm 5% 15W Power Resistors - $ 0.63 cents X 8 resistors
( 2 sets of 4 are to be wired in series for a total equivalent resistance of 15 ohms )
http://www.mouser.com/Search/Refine.aspx?Keyword=280-CR15-30-RC
http://www.mouser.com/catalog/specsheets/XC-600041.pdf
' r ' - 8.2 kohm 5% 1/4W - $ 0.22 cents
http://www.mouser.com/Search/Refine.aspx?Keyword=30BJ250-8.2K
' LED ' - Green LED - $ 0.26 cents
http://www.mouser.com/Search/Refine.aspx?Keyword=351-5502-RC
' LED ' - Green LED - (SDI) LTL-4233 - $ 0.15 cents
http://www.surplussales.com/Bulbs-Incan-Panel/LEDDisplay.html
_____
Available capacitors that have a higher working voltage than the ~ 162 Volt peak
that comes from the wall socket , require this voltage to be stepped up.
Drawing ' E ' has the capacitor charged to a higher voltage with a transformer included in the circuit.
Switch ' S ' flows electron current through diode ' dy ' into primary coil ' p ' of transformer '
T ' and
' Bot ' conductor back to the a.c. source , inducing a current in secondary coil ' s ' of transformer '
T '.
Current flows through diode ' dx ' into ,' Mid ' conductor collecting in the negative ( - ) terminal of
capacitors(s) ' C ' charging the elevated voltage , displacing current out of the positive ( + ) terminal
flowing back through ' Top ' conductor down resistor ' R ', and parallel resistor' r ' with series
light
emitting diode ' LED ', into 'secondary coil ' s ' of transformer ' T '.
( * * where the wire crosses ' Mid ' conductor there is N O connection )
Light emitting diode ' LED ' momentarily flickers on then darkens indicating that charging has occurred.
Firing as already explained is done by circuit breaker ' X ' with the charging circuit F I R S T
switched off by normally open switch ' S ' before the charged circuit can be fired..
_____
Alternatively as explained in Drawing ' F ' using another capacitor as a half wave
voltage doubler , ramps up the charging voltage to 2 times the a.c. line voltage.
In phase - 1 - the capacitor is charged , then in phase - 2 - the reverse voltage
of the a.c.line and the just charged capacitor add together in series aiding effectively doubling.
Illustrated here _
http://www.gallawa.com/microtech/doubler.html
Motor ' start ' capacitors are bipolar electrolytics , also called Non-Polarised or NP capacitors
and are the most suitable for use in voltage doubling as they are designed for a.c. operation ,
but only if run briefly. Those with a Voltage rating higher than the ~ 125 VAC minimally necessary
can be run for just a little longer before overheating ( over 85º C , impairment will occur )
(Cy) ' 270-324 uF, 250 VAC ( the range specifies plus or minus 10% variance from 300 uF )
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=200326166634
http://www.amazon.com/dp/B000LERFO4 , ( the point here is shop around )
These here are perfectly adequate _
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=230281260691 - $ 4 each
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=370198861276
Motor ' run ' caps can operate continuously but have much less capacitance, or are
much more expensive for equivalent capacitance, and larger.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=310028799378 - $ 12 each
Capacitor testing - http://toad.net/~jsmeenen/capacitor.html
U.S. a.c. power lines typically achieve a peak of ~ 162 volts which provides
the equivalent of 115 volts direct current, called R.M.S. ( root mean square ).
You may wonder why a.c. capacitors have the seemingly odd rating of 125 volts.
60 cycle a.c. is 120 half cycles of alternating polarity, voltage rises from zero
to peak voltage after 1/240 of a second (.004167) ( yellow portion shown below )
and over the next 1/240 of a second drops back to zero again ( green portion )
repeating this rise and fall but in reverse polarity for the following half cycle.
( orange and lime )
[url] [/url]
This sets the time available for the percentage of the supplied line voltage a
capacitor will charge up to during the rise of one half cycle. It is not instantaneous.
Re-arranging the folrmula : Xc = .159 / f C , C = farads
f = frequency , Xc = capacitive reactance of the capacitor expressed as ohms ,
thus : XcC = .159 / f , = .159 / 60 , = .00265 , which is effectively the minimum
RC time constant ( TC ) for any capacitor at this frequency ( without considering
additionally the ESR which will make the time constant longer ) It does not matter
what size capacitor or what voltage is applied to it, this value is set in stone, it
is characteristic of 60 cycle a.c. and cannot be changed. Dividing the rise time
.004167 by .00265 , = 1.57 TC ( Time constants ) works out to 79 percent of
peak a.c. line voltage , or 128 volts. This in practice will never be seen due
to the resistance present in the capacitor, this despite that the capacitor will
still be charging slightly even as the a.c. line voltage is dropping , until both are
equal. As the applied a.c. voltage decreases the capacitor will discharge through
the a.c. source. The capacitor will not be completely discharged when the a.c.
line voltage reverses polarity, but as the polarity is now the same as the capacitor
polarity, the voltages aid, quickly discharging the capacitor the rest of the way.
When the applied a.c voltage reverses direction, the capacitor is charged again,
and the entire process is repeated with the next half cycle of the a.c. waveform.
The waveform of an applied sinusoidal a.c.voltage will remain unchanged and only
its amplitude ( peak value ) will be affected. Hence when measuring the voltage
pole to pole on a capacitor in series with an a.c. line you detect a voltage drop,
just as though it were a resistor, this is the capacitive reactance.
http://www.electronics-tutorials.ws/capacitor/cap_8.html
In a half wave voltage doubler , diodes are used to block one phase of the
a.c. waveform so that the capacitor remains charged when the polarity is
reversed. When the polarity again changes , charging continues where it left
off incrementally raising the voltage charged in the capacitor ever closer to
the a.c. supply voltage.
_____
Drawing ' G ' shows this scheme.
- In the first a.c. phase
current flows up from ' Bot ' conductor through resistor ' R ', and parallel resistor ' r '
with series ' LED ' light emitting diode , to motor start capacitor ' Cy ' , diode ' dy ', and
switch ' S ' back to the a.c. source. ( Resistance ' R ' is necessary to protect the
small diodes ' dx , dy , dz ' from excessive current when charging starts ).
- In the next a.c. phase
switch ' S ' flows current through diode ' dx ' , into ' Mid ' conductor,
collecting in the negative ( - ) terminal of capacitor(s) ' C ', displacing current out of
the positive ( + ) terminal flowing through ' Top ' conductor , and down diode ' dz '
( * * NOTE where the wire crosses ' Mid ' conductor there is N O connection )
into motor start capacitor ' Cy ' where the a.c. line voltage acts in series aiding , to
raise the charging voltage on capacitor(s) ' C ' to double a.c. line voltage. As it is
charged , light emitting diode ' LED ' momentarily flickers on then darkens indicating
that charging has occurred. Firing is as before done by circuit breaker ' X ' and switch ' S '
turned off F I R S T before the charged circuit can be fired.
When fired , the a.c. line voltage applied in series aiding effectively raises the circuit
voltage to 3 times the a.c. line voltage , enough for spark gap ' G ' to arc into
conduction , current surges into the positive ( + ) terminals of cap(s) ' C ' displacing
current from the negative ( - ) terminals of capacitor(s) ' C ' left through
' Mid ' conductor and circuit breaker ' X '. ' Top ' conductor channels the
current surge left and down shunt diode ' D ' to the a.c. source.
' dx , dy , dz ' - Small ~ 6 amp diodes - (SDI) MR756 - $ 0.30 cents
( 2 of these are to be wired in parallel for each of ' dx , dy , dz ' a total of 6 diodes )
found down a bit from the top , use the " find on this page " function of your browser
http://www.surplussales.com/Semiconductors/Diodes-Rectifiers-5.html
' G ' - Air Gap - 350 to 600 volts - (RMO) WP438 - near the bottom of this page - $ 3
http://www.surplussales.com/Semiconductors/SemiCTransSup.html
[url] [/url]
Using Rectifiers In Voltage Multiplier Circuits
http://www.eettaiwan.com/ARTICLES/2001JUN/2001JUN14_AMD_AN2009.PDF
_____
Drawing ' H ' Omitted for clarity but an essential part of any
such devices is a means for safely discharging the stored energy of a bank of capacitors.
This must be attached and operable before the capacitors are ever charged . Under no
circumstances E V E R handle any part of these devices when charged except to fire !
Until after you have completed the ritual described in the next 2 paragraphs the circuit must
be considered to be charged even if it has been fired - there may still be residual charge.
Attached to ' Top ' conductor parallel with the shunt diode ' D ' is a 2 meter length of wire
with it's other end soldered to a strip of metal on one edge of a small wood block ' B '.
On the opposite edge of wood block ' B ' is another strip of metal soldered with a wire that's
hooked to one terminal of an otherwise unconnected a.c. plug ' E ' ( kept safely inside of
a glass jar ) this will be inserted instead of the power cord into the surge generator.
Wood block ' B ' is to be weighed and submerged in a small bucket of water acidified with
a supermarket bought pint bottle of lime juice ( citric acid ). This water " resistor " will harmlessly
dissipate the charge.
To discharge by using this circuit instead , the entire device must F I R S T be powered off
by disconnecting from the a.c. line altogether and instead connected to plug ' E '
Switch On circuit breaker ' X ' to conduct the charge through the corresponding
terminal of the a.c. plug into the receptacle wired to the water bucket.
This above is simply a very cheaply implemented expedient contrivance to obviate
more elaborate alternative permanent wiring to the circuit which will be outlined
in the closing post. A circuit can include instead a compact drain resistor network.
Wire wound power resistors can withstand Intermittent loading 10 times their rated
power for short pulse durations provided duty cycle limitations are observed. The
best compact size and price per watt is for wire wound cement types in 15 or 25
watt rating.
Remember that 95 % of the energy will dump during the first 1.5 RC time constant.
The reciprocal of capacitance 1/C is the resistance value for an RC time constant
of one second. This means that 95 % discharge of energy will occur in 1.5 seconds.
To provide an equivalent resistance that gives a discharge time long enough to stay
within the power handling capabilities of the resistor network first calculate capacitor
bank energy. An example using 6000 uF charged at 450 volts , E = CV^2/ 2 = 607 Joules
= ( .006 X 450 X 450 ) / 2 . Divide 607 Joules by the maximum excess loading factor
10 , to obtain the combined watt rating of the resistors to be used. 60.7 is most
closely 4 X 15 watt resistors. Now divide " 10 " by the capacitance thus 10/.006 ,
to give a value of 1667 ohms ( 1600 nearest manufactured value ).
Alternatively , r.m.s. discharge voltage (.707 X 450 ) is 318. Using P = V^2/ R
re-arranging to R = V^2/ P , ( 318 X 318 )/60.7 = 1666 ohm , same thing
* Dividing V^2 by full value , 607 Joules , resistance is 167 ohm same as with 1/C .
1600 ohms will discharge 95 % in 15 seconds , total discharge after 50 seconds,
why it is advisable to overload the resistors and use much less ressitance value,
by choosing instead to divide 1.5 or 2 by the capacitance .
A series of 4 groups , each with 4 resistors in parallel , 16 total ,
costing from $ 9.12 to $ 15.84 - See Mouser catalog link below _
can be secured on both sides of a thin aluminum plate ( If all resistors are the
same value then that value will be the equivalent combined resistance ) and
will give adequate heat sinking 10 times the actual 240 or 400 watt rating.
Using P = V^2/ R ( 318 X 318 )/62 = 1631 watts , what a hair blower/dryer
consumes per second. Choosing a 62 Ohm resistor , one group of 4 in parallel can
serve double purpose as the current limiting resistance for the charging circuit.
http://www.mouser.com/Search/Refine.aspx?N=4141191+4294966099+42945...
http://www.mouser.com/Search/Refine.aspx?N=4141191+4294966099+42945...
http://www.mouser.com/catalog/specsheets/XC-600041.pdf
_____
Aluminum electrolytic capacitors provide the highest energy pulse for their size.
Capacitance and voltage vary inversely , you will find those sizes having optimal
energy storage occurs at a few thousand uF at a few hundred volts. Select one
specifically designed for Strobe and Photoflash applications, these are made with
very low ESR ( Equivalent Series Resistance ). Larger capacitance has lower ESR ,
in the range of interest typically around 0.1 ohm or less. This type can best
provide the firing pulse. Price and sizes vary greatly, shop around.
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=260350709930 - 2 X $ 35
This one here is 150 Joules in 100 cc ~ ( 40 X 80 mm )
' C ' - Electrolytic Power Capacitor - 1500 uf 450 VDC - $ 5 , salvaged
* * Marx Generator made up of 10 of these = 1500 Joules
http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=140322028747
* Note if this auction has ended, scroll down to the original offer and
. . select " View seller's other items ". This offer is regularly repeated.
Observe dummy terminal pinout-code VRD pg 13(pdf) , guidelines pg 19 , 20 , 21, (pdf)
http://www.chemi-con.co.jp/e/catalog/pdf/al-e/al-all-e1001i-090126....
available new from United Chemi-Con Caps - http://www.ucccaps.com
http://www.ucccaps.com/detail.aspx?ID=103647
http://www.ucccaps.com/detail.aspx?ID=105567
Described as " snap mount " usually implies " spade " type terminals for mounting
with spade wire lugs for easy replacement when servicing appliances. These
are not that kind and are best mounted on a board with terminals soldered.
- As we shall see , output of these caps can be doubled to 300 Joules per 100 cc.
_____
Hi Energy Cap
http://www.gaep.com/tech-bulletins/large-high-energy-density-capaci...
Electrolytic Technology & Design
http://www.elna-america.com/tech_al_principles.php
Aluminum Electrolytic Application Guide
http://www.cde.com/catalogs/AEappGUIDE.pdf
Electrolytic Capacitors - Testing, Care & Storage
http://www.jammaboards.com/guides/Capacitor_Testing.pdf
also online here - http://www.repairfaq.org/sam/captest.htm
ESR
http://www.gaep.com/tech-bulletins/capacitor-engineering-bulletins....
http://www.faradnet.com/deeley/book_toc.htm
Capacitor Service Life
http://www.liebert.com/common/ViewDocument.aspx?id=1210
These electrolytics are polar and cannot stand reverse bias , but can be discharged while
connected in series aiding , providing there is a diode ' D ' parallel , serving to prevent
reverse charging as described before. This is the most expensive item of the circuit selling
for 30 to 55 dollars U.S. A 600 volt 300 amp ( or better ) chassis mount stud type Diode
with flag terminal. one of this -> ' D ' Stud Diode - ( 300 amp - 600 volt ) - $ 30.44
http://www.mouser.com/Search/Refine.aspx?Keyword=844-300UR60A
http://www.vishay.com/docs/93508/93508300.pdf
http://sigma.octopart.com/12038/datasheet/Vishay-300UR60A.pdf, similar to these
[url] [/url][url] [/url]
A diode of lesser current rating ( and therefore cheaper ) can withstand
brief surges , but it won't last for very long through repeated firings of this type.
Alternatively several cheaper lesser rated diodes may be used in parallel , for example ,
4 of these -> ' D ' Stud Diode (SDI) 70HF40 - ( 70 amp, 400 volt ) - $ 0.50 cents salvaged
for anything higher than it's 400 volt rating , 2 in series 4 times , 8 total.
look down near the bottom , use the " find on this page " function of your browser
http://www.surplussales.com/Semiconductors/Diodes-Rectifiers-3.html
http://sigma.octopart.com/511589/datasheet/Vishay-70HF40.pdf
or 3 of these -> ' D ' Stud Diode - ( 95 amp - 800 volt ) - $ 6.82
http://www.mouser.com/Search/Refine.aspx?Keyword=844-95PF80
http://www.vishay.com/docs/93532/93532.pdf
General information on diodes.
http://www.kilowattclassroom.com/Archive/DiodeRec.pdf
_____
The seller of the 1500 uF cap listed above , correctly observes that 10 of those
arranged into a Marx generator , ( a device to charge caps in parallel and then discharge
them together in series ) will produce a 4500 volt output ( 450 volts each, times 10 ) but
at only 150 uF equivalent capacitance ( 1500 uF / 10 ) for a total 1500 Joules.
http://www.allaboutcircuits.com/vol_1/chpt_13/4.html
http://home.earthlink.net/%7Ejimlux/hv/marx.htm
http://www.kronjaeger.com/hv/hv/src/marx/index.html
http://www.electricstuff.co.uk/marxgen.htm
Greinacher Cascade Multiplier _
http://www.instructables.com/id/High_Voltage_Power_Supply_For_Marx_...
see parts (CFP) YE-TW-3 and (CFP) YE-TW-6 , here _
http://www.surplussales.com/Capacitors/motorstart.html
http://www.techlib.com/files/voltmult.pdf
http://www.amazing1.com/download/MARXIU1045.pdf
http://www.celnav.de/hv/hv3.htm
http://www.celnav.de/hv/hv4.htm
Here is a real big one ->http://skyfi.org.ru/photos/2008/marksgen/005.jpg
http://skyfi.org.ru/photos/?path=marksgen
Drawing ' I ' shows this with just 4 caps , gaps and diodes in place of resistors,
( anyone of the already outlined charging circuits can be used here )
switch ' S ' flows current through resistance ' R ', and parallel resistor ' r '
with series light emitting diode ' LED ', up diode ' dx ', into ' Mid ' conductor,
through diodes ' d1 ' collecting in the negative ( - ) terminals of capacitors ' C ',
displacing current out of the positive ( + ) terminals flowing through diodes ' d2 '
and out diodes ' dy , dz ' into ' Bot ' conductor back to the a.c. source.
( one 600 volt diode as ' dx , dy ' in line , per every 3 capacitors to resist reverse breakdown
during firing if caps are charged at just a.c line voltage , add more diodes as needed if caps are
charged at higher voltage and substitute spark gaps for the gas gaps ).
As it is charged , light emitting diode ' LED ' momentarily flickers on then darkens
indicating that charging has occurred ( this is also necessary to protect the small diodes
from excessive current when charging starts ). The charging circuit is to be switched off
F I R S T before the charged circuit can be fired. A normally open switch
accomplishes that.
Circuit breaker ' X ' fires the circuit by applying the a.c. line voltage in series with
the caps ' C '. Current is drawn back through ' Mid ' conductor from the
negative ( - ) terminal of the left cap ' C ' which arcs into conduction the left Gas Gap ' gg ' ,
in an avalanche cascade , current surges from the negative ( - ) terminals of all caps
' C ' across Gas Gaps ' gg ' into the positive ( + ) terminals of cap(s) ' C '
to arc into conduction Spark Gap ' G ' where from current surges into the positive ( + )
terminal of the right cap ' C '. ' Top ' conductor channels the resulting current surge
to the a.c. source down shunt diodes ' D1 , D2 ' in series to resist reverse brealdown
during firing.
______
Given the example of the 1500 uF 450 Volt cap , it can be charged at 3 times the
peak domestic U.S. a.c. value of 162 volts (* 115 volts X √2 ) to 486 volts ( note
that overcharging slightly ~ 8.5 % is perfectly alright providing it is discharged
promptly and since it is static and not subjected to cyclical deep discharge at
a.c. line frequency which will cause overheating and breakdown failure ).
Voltage of the ten caps ( 10 X 486 ) + 162 peak a.c. line voltage add to 5022 volts
By E = CV^2/ 2 , E = Joules , C = farads , V =
volts
1891 Joules = (.00015 F X 5022 X 5022 ) / 2 , , or as Dirty Harry ( Clint Eastwood ) would say
- " will blow your head clean off " - , this is not an exaggeration , 1891 Joules is equivalent to
1395 foot pounds of muzzle energy , a proof load for the .44 caliber remington magnum ,
and nearly the equal of the large heavy power cap shown above.
Loud blowing up of fruit http://www.youtube.com/watch?v=aWf-V2-kr9Q
Here is what happens with 9 kJ - equivalent to a .50 caliber Browning Machine Gun round.
from http://www.powerlabs.org
Just think , that could be you all over the street.
![](http://www.powerlabs.org/images/watermelon_small.jpg)
9kJ is what a hair blower/dryer consumes in only six seconds The short time interval
of the pulse makes all the difference. Read Overview here _
http://en.wikipedia.org/wiki/Pulsed_power
- This is continued in the following post -
___________________________________________
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
- continued from the post above -
If 2 capacitors ' C ' are placed in a circuit so that both are in series , as in the
Marx layout , equivalent capacitance then becomes 750 uF , but the voltages in series
add 486 + 486 , plus the a.c. power line voltage of 162 peak comes to 1134 volts at firing.
By E = CV^2/ 2 , (.00075 F X 1134 X 1134 ) / 2 , = 482 Joules
Drawing ' J ' - T H I S . I S . T H E . O N E . T H A T . M A T T E R S
Alternatively if both capacitors are placed instead in parallel as in circuit Drawing ' J ' , the
equivalent capacitance is the sum of both , 3000 uF. The firing voltage remains the
same , charged to 486 volts ( 3 times peak a.c. line voltage ) plus the a.c. power line
voltage boost at firing comes to 648 volts.
By E = CV^2/ 2 , (.003 F X 648 X 648 ) / 2 , = 630 Joules
THE LESS ELABORATE CIRCUIT OUTPUTS . 1 4 8 . JOULES MORE POWER !
Employing the higher voltage rated caps ~ 450 volts requires a voltage tripler circuit
and an additional charging a.c. motor start capacitor. This is outlined in Drawing' J '
Single Pole Single Throw switch - SPST / no ( normally open ) ' S ' connects the
charging circuit of a.c. motor start capacitors ' Cx , Cy ' & diodes ' dx , dy '.
( * * NOTE where the wires cross ' in between ' Cx , Cy ' there is N O connection ).
Half cycle ' 1 ' charges the caps and half cycle ' 2 ' applies both voltages in series with
the a.c power line voltage to diodes ' d1 , d2 ' and the firing capacitors ' C '
charging them in parallel to 3 times the a.c. voltage. Electron current flows through diode
' d1 ' into ' Mid ' conductor collecting in the negative ( - ) terminals of caps ' C '
displacing current out of the positive ( + ) terminals flowing through ' Top ' conductor
and down into diode ' d2 ' ( * * NOTE where the wire crosses ' Mid ' conductor
there is N O connection ).
Resistance ' R ', is necessary to protect the small diodes from excessive current
when charging starts. Parallel with ' R ' is an a.c. voltage rated light bulb ' B ' to
provide visual indication when it momentarily flickers on then darkens indicating that charging
has occurred. Firing as before is done by circuit breaker ' X ' with switch ' S '
turned off F I R S T before the charged circuit can be fired.
Circuit breaker ' X ' fires the circuit by applying the a.c. line voltage in series with the cap(s) ,
which arcs into conduction the Spark Gap ' G ' near the peak of the a.c. sine wave.
' Top ' conductor channels the resulting current surge left and down shunt diode ' D '
to the a.c. source.
I want to emphasize that it is possible to achieve equally high energy and
power surge levels using fewer components than for the Marx generator mentioned above.
Compared to the 1891 Joule Marx generator of 10 - 1500 uF caps ,
just 6 of these caps in parallel as in Drawing ' J ' would be 9000 uF , then By E = CV^2/ 2 ,
1890 Joules = (.009 F X 648 X 648 ) / 2
EQUIVALENT OUTPUT OF JOULES USING FEWER COMPONENTS !
The circuit performance is improved by eliminating the extra resistance that is present
with the extra components and wiring. By having a bank of capacitors in parallel , just as
with resistors in parallel , equivalent resistance is greatly diminished , ( the exact opposite
of series resistance! ). http://www.electronics-tutorials.ws/resistor/res_4.html
http://www.electronics-tutorials.ws/capacitor/cap_6.html
http://www.electronics-tutorials.ws/capacitor/cap_7.html
The same as with batteries , for capacitors in series the voltages add , for capacitors in parallel
the currents add. The RC time constant remains unchanged for either configuration because the
ESR of individual caps does not change. TC = R X C = ( 0.1 X .0015 F ) = .00015 sec
- | - For six 1500 uF caps in series the errected capacitance is ( 1500 / 6 ) = 250 uF, the ESR
value of each adds so that 6 X 0.1 ohm = 0.6 ohm , TC = R X C = ( 0.6 X .00025 F ) = .00015 sec.
- | - For six 1500 uF caps in parallel capacitance adds totaling ( 1500 X 6 ) = 9000 uF, equivalent
ESR of the bank reduces to ( 0.1 ohm / 6 ) = .0167 ohm , TC = ( .0167 X .009 F ) = .00015 sec. also
By graphing power relative to time as it appears on an oscilloscope, energy is represented by
the area under the power trace. Capacitor ESR establishes how brief the RC time constant is
and how narrow the pulse. By varying the voltage , the only thing that can be altered is the
leading slope of the pulse ( and the height ) not its width ( the area remains constant because
the energy is the same ) Very short pulse durations are only acheived by capacitors having
exceptionally low ESR. As these do not have large capacitance, they must have a very high
voltage to store significant energy.The increased resistance of the series circuit depletes more
power and is less efficient. More importantly with lower voltage one can dispense with heavy
high voltage insulation requirements to obviate systemic breakdown ( arcing ) which occurs at
kilovolt levels. At kilovolts everything is huge http://www.celnav.de/hv/hv1.htm
Of course lower voltage produces comparatively longer current rise time ( all things being
equal ) but it won't matter given the following current surge.
The rated energy per cap at 450 volts is 150 Joules times ten caps is 1500 Joules.
By slightly overcharging to 486 volts plus the boost from a.c. line voltage applied in series
during it's discharge , output is increased to 3149 Joules = (.015 F X 648 X 648 ) / 2
slightly more than the 165 pound 240 uF high voltage capacitor mentioned at the beginning
of this post. Yet this device is small enough to fit into a shoe box and weigh less than a
3 liter bottle of soda.
How to size a capacitor bank , specific for ultracapacitors but generally applicable.
http://www.powerdesignindia.co.in/STATIC/PDF/200807/PDIOL_2008JUL31...
See attachment below - Equivalent Series Resistance & Maximum Leakage.rtf -
One can determine all the values for one capacitor and then simply
- - multiply by the number of caps to obtain the output of all. Thus:
By E = CV^2/ 2 , (.0015 F X 648 X 648 ) / 2 = 315 Joules
- - 315 X 6 = 1890 Joules
ESR of one 0.0015 Farad cap is ~ 0.1 ohm R in value
TC ( Time Constant ) by R X C = TC , 0.1 X .0015 = .00015 sec
Average power of a single cap is ( 315 joules / .00015 ) = 2,100,000 watts
- - times 6 ( caps ) is 12.6 megawatts combined average power.
Mean voltage is close to the RMS ( Root Mean Square ) value
( Peak voltage is ~ 648 ) X 0.707 = 458 volts
Dividing 2,100,000 watts by 458 volts = 4585 amps average current
- - 4585 X 6 caps = 27510 amps.
A related thread on this topic http://www.sciencemadness.org/talk/viewthread.php?tid=6032
Pulsed power has many varied applications. http://www.nessengr.com/links.html
A pulse can serve to energize a laser, or a maser ( its equivalent in the microwave region ).
Jolt a coil generating intense magnetic flux or provide the feed current to a magnetic flux
compression generator. This remains an area of endeavor into high energy physics within
the means of shoe string budgets.
http://en.wikipedia.org/wiki/Explosively_pumped_flux_compression_generator
Coin shrinking - http://205.243.100.155/photos/shrinker5.pdf
applying the data just derived above on a 20 turn coil of 1/2 inch ( 12.5 mm )
using flat tape conductor as this _
http://www.keison.co.uk/furse/conductors_flat_tape.htm
The more squat the coil the higher the induction
Flux Density (in Gauss ) = N I / (2.02 × L)
Where:
N = # of turns of wire on coil
I = current flowing in coil ( Amperes )
L = length of coil ( Inches )
Flux Density (in Gauss ) = ( 20 )( 27510 A ) / (2.02 × 0.5”) = 54.5 Teslas
54.5 times the 10,000 gauss a Neodymium magnet has in a closed armature , which is
5 times what the 2000 gauss open air or gap flux will be , 272 times more. In other
words, the field of a rectangle of Neodymium magnets 16 by 17, 272 total, compressed
into the space of one Neodymium magnet. No wonder the coils explode. Due to the rapid
rise and fall of the field , the induced eddy current in the coin will be far greater. Which
is why you do not want the induction from a wide loop of connecting cable to couple to
the surroundings where it does nothing but subject the tape collection you keep in the
room above the garage to an EMP , : -O
The energy must remain in the wires all the way to where it has to act.
More than you probably want to know _
http://www.mse.eng.ohio-state.edu/%7EDaehn/metalforminghb/tabofcont...
- V E R Y , V E R Y , I M P O R T A N T -
It is not my responsibility to protect you from your irresponsibility.
If you do not understand what is detailed here which is as basic as it gets , or what a
ground fault is and how not to create one , you are at serious hazard of electrocuting
yourself when you become an unwitting part of the circuit. ( " Electrocute " is a word
derived from the invention of the electric chair, meaning electro - execution , if you
auto - electro - execute , you are automatically eligible for a Darwin award. )
A 20 kilovolt discharge when it is the result of static buildup from walking on a carpet
is merely startling due to the miniscule power. A several hundred Joule pulse at this
same level will kill you instantly if received in one hand and out the other hand.
If only one arm or leg is affected it will knock you out cold and due to the resulting
nerve damage you may never be able to use that arm or walk well again.
Standing on a dry rubber mat when outside is good practice. Not directly touching the
firing switch is life saving practice , use a stick.
http://www.repairfaq.org/sam/safety.htm
In order not to cause havoc and collateral ruin , the a.c. branch line used must be
directly wired to the main service panel power transfer switch, from the utility.
Nothing else will do. Preferably with its own circuit breaker since you W I L L trip
a circuit breaker. Do not bypass the utility meter, it's surge arrestor safely adds another
cutoff point. There cannot be any lighting or running domestic appliances even in standby
mode , such as televisions , air conditioners , refrigerators , including ones you have not
thought of such as the water heater , kitchen stove ignitor ( without gas pilot ), timers ,
clocks , or alarms. Everything off on that branch line , preferably disconnected. If you
have any doubts at all consult an electrician first. Finally, house wiring of aluminum is not
suitable for this method, nor if it has been de-rated from original installation, is more than
40 years old or is cotton asphalt insulated. All these pose potential risk for starting fire.
____________________________
Triggered Switches
Below is collected wisdom from the internet on over-voltage gaps
The spark gap length / voltage relationship varies with electric field distribution, but
spark lengths usually range roughly from .28 to .9 mm. per kilovolt of peak voltage at
voltages in the 4 to 35 KV range. Significantly longer spark lengths in millimeters per
kilovolt can occur at peak voltages near or above 50 KV. This is for air at normal sea
level atmospheric pressure. Longer sparks can occur at high altitudes.
A spark gap can be triggered in many ways, but there are two main types, those
which trigger via localized ionization ( via a secondary spark, point corona, UV laser,
ionizing radiation, flame etc.) and those which trigger via altering the breakdown
potential of the gap along the Paschen curve ( such as spark gaps which are initially
pressurized and then bled to provoke conduction ).
The x-axis of the Paschen curve is measured in mBar-mm ( the product of pressure
and separation ), with the y-axis being in kV. If the separation is constant and the
pressure is reduced the breakdown potential decreases until the gap fires, likewise if
the pressure is constant and the separation is reduced the breakdown potential
decreases until the gap fires. Being that such a gap triggers itself by lowering the
separation ( but not touching ) it can be said to fall into the latter of the two types.
If the contacts touch to conduct then it would be a mechanical switch.
Gas discharge tubes are filled with special gases with low dielectric potential designed
to arc-over ( that is, start conducting electricity ) at predictable low voltages. In
other words, the gas-tube is a closed environment that allows lightning-like pulses of
electricity flash through. By selecting the right gas and separation of electrodes in the
tube, engineers can set a precise flashover voltage.
Gas tubes can conduct a great deal of power--thousands of kilowatts--and react
quickly, typically in about a nanosecond. They are faster than MOVs ( Metal Oxide
Semiconductors ) and less likely to be damaged by large surges.
On the negative side, a gas tube does not start conducting ( and suppressing a surge )
until the voltage applied it reaches two to four times the tube's rating. The tube itself
does not dissipate the energy of the surge; it just shorts it out, allowing your wiring to
absorb the energy. Moreover, the discharge voltage of a gas tube can be affected by
ambient lighting ( hence most manufacturers shield them from light ).
Worst of all, when a gas tube starts conducting, it doesn't like to stop. Typically, a gas
tube requires a reversal of current flow to quench its internal arc, which means that the
power going to your PC could be shorted for up to 8.33 milliseconds, the length of a
single half cycle of utility power. Sometimes gas tubes continue to conduct for several
AC current cycles
__________________________
F O O T N O T E S
TM 5-692-2 Maintenance of Mechanical & Electrical Equipment
http://www.army.mil/usapa/eng/DR_pubs/dr_a/pdf/tm5_692_2.pdf
From top page 28-2
A lightning stroke to a power system develops very high surge voltages across
equipment and line insulation systems. If these voltages exceed the insulation strength, a flashover
occurs. Once lightning enters a power system, the surge current is unlikely to cause any damage.
Although the current may be extremely high, it is very short lived and can easily be handled by a small
conductor. The largest recorded conductor to be fused or vaporized by a direct stroke was an American
Wire Gage (AWG) No. 10. The size of conductors, installed expressly for conducting lightning currents,
is usually determined by mechanical strength considerations, rather than by current-carrying capacity. On
some rare occasions, overhead ground wires have been severed by lightning at the point of contact. This
is probably due to the stroke channel heating the conductor at the point of impingement, rather than from
simply conducting the lightning current.
_______________________
Transient Overvoltages In Electrical Distribution System & Suppression Techniques
http://www.nalanda.nitc.ac.in/nitcresources/ee/lectures/VoltageTran...
From middle paragraph page 12
In applications where there is a normal operating voltage, as in the AC mains, there is a
possibility that the gas tube will not reset itself once it has fired and suppressed the transient.
This condition is known as " follow on " current and is defined " as the current that passes
through a device from the connected power source following the passage of discharge current ".
Follow on current will maintain conduction of the ionized gas after the transient has disappeared
and the concern is that the follow on current may not clear itself as the a.c current drops to zero
and will result in a permanently destroyed gas tube. In an AC mains application it is not sufficient
to rely solely on the crossings of the sinusoidal voltage to extinguish the follow current.
_______________________
Grounding, Bonding, & Shielding For Electronic Equipment Vol II & I
http://www.tech-systems-labs.com/books/grounding.pdf
From middle page 1-68 , Vol II
(1) Follow current. The typical discharge (arc) voltage across a spark gap is 20 to 30 volts while it
is in full conduction. Because of the low arc voltage, the voltage and current available from the ac power
supply would maintain the spark gap in an on state after a transient was dissipated until the first zero crossing
of the power supply or until a supply line fuse opened, a line burned open, the spark gap burned open, or the
service transformer burned open.
__________________________________________
Attachment: Equivalent Series Resistance & Maximum Leakage.rtf (264kB) This file has been downloaded 2006 times
|
|
GoatRider
Harmless
Posts: 21
Registered: 3-11-2008
Location: Minnesota
Member Is Offline
Mood: No Mood
|
|
I blew up a switching power supply regulator when I misread the pinout and hooked +V to ground. It had a ceramic package, there were bits of ceramic
thrown across the room.
|
|
zed
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2281
Registered: 6-9-2008
Location: Great State of Jefferson, City of Portland
Member Is Offline
Mood: Semi-repentant Sith Lord
|
|
Interesting. My friend Dr. Crazyfingers built an exploding wire device many years ago. Folks like Magicians and the special effects people from the
Lucas related "Industrial Light and Magic" where interested in the effect.
Unfortunately, he was trying to achieve the effect with a very small gauge, round, copper wire(nearly invisible).....and his results were
inconsistent.
During development, his potential clients found a string form of chemical explosive that was reliable and easy to handle, and he reluctantly abandoned
exploding wire experiments.
He did however, impart a little high voltage wisdom to me, while the project was ongoing......."Don't point at that giant capacitor....Damn it"
"This stuff points back!"
[Edited on 23-6-2009 by zed]
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
Available again - Get'em while they last !
' C ' - Electrolytic Power Capacitor - 1500 uf 450 VDC , 150 Joules in 100 cc ~ ( 40 X 80 mm )
- $ 5 , salvaged -> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=14033...
![](http://i20.ebayimg.com/03/i/001/07/d4/e16f_35.JPG)
.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Electrolytic? Pffifle. Too much inductance.
http://webpages.charter.net/dawill/tmoranwms/Elec_CapBank.ht...
I estimate this bank has 20.0uF capacitance, 0.1uH ESL and around 0.003 ohms ESR. That's a time constant of about 2.3 microseconds. The caps are
rated 630VDC max, 1.5 x overload for short periods, so could in principle run up to nearly 1kVDC. (That's all of 10 joules stored energy.) A short
directly across the terminals would take 2.3us to turn that 1kV into 0V at approximately 14kA (minus losses). If voltage falls linearly while current
rises linearly during this event (a gross approximation to the damped sinusoid of reality), then the peak power will be on the order of 3.5 megawatts.
(After the first 1/4 cycle, most of the energy (about 8-9J) will be in magnetic form, then after another quarter cycle it'll be back in the
capacitors; etc.)
Tim
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
@ 12AX7
See the detailed explanation in the opening post a bit down from the top
where it says Joules needed for vaporization of a 2 cm 28 AWG wire
on how to determine the joules needed to blow the volume of metal
comprising a bridgewire. You can calculate the weight with this _
http://circuitcalculator.com/wordpress/2007/09/20/wire-param...
Knowing this you can now proceed to the next phase , of determining
the capacitance and voltage to provide the needed energy
You stated
~ 4 Joules = {( 20 uf ) X ( 630 volts ) X ( 630 volts )} / 2
10 Joules at 1000 volts is not realistic , if the rating is
substantially exceeded the caps will fail , count on it.
Of course you can get around this by placing 2 caps in series
to withstand 2 X 630 volts. The energy will remain ~ 4 joules
for the reduced equivalent capacitance of 5 uf
Discharge time will not change , although the actual output
certainly will be less. Read my third post above for why that is.
The ESR ( Equivalent Series Resistance ) of the capacitor(s) determines
what the discharge time will be. By your numbers the ' RC ' time constant
has to be (.003 ohm ) X (.00002 farad ) = 60 nanoseconds , where 2.3
microseconds comes from I don't see.
If you don't know or have a ' reasonable ' idea of what the ESR
is ( from the manufacturer perhaps ) any ' surmise ' of a discharge
time is just that, surmise.
Voltage is irrelevant to anything but the determination of what
the energy of the capacitance is. Higher voltage will only squash
more of the energy into the leading part of the discharge profile.
For the established ESR the pulse width remains unchanged
regardless of any other changes. A higher energy pulse may be
inferred from the truncated leading portion from what follows.
Extremely short ( a few millionths of a second ) discharge times
unless there is some other need besides just exploding a wire, is
wholly uncalled for and unnecessary.
A very big problem which rears up with short discharge times is
the skin effect which overshadows any stray inductance by at
least another 2 orders of magnitude. ( 100 times or greater )
You cite 2.3 usec ( I'll humor you for the moment )
frequency is the reciprocal of time , f = 1/ t , then the 2.3 millionths of a second
translates to a frequency of 435 kHz. You will need at the very
least a 32 AWG wire or much preferably smaller to lessen attenuation
of the pulse.
See this chart - http://www.powerstream.com/Wire_Size.htm
posted just above this - Joules needed for vaporization of a 2 cm 28 AWG wire
or calculate it here yourself _
http://www.mantaro.com/resources/impedance_calculator.htm#sk...
also attached below _
True all of this is classic EBW technology and practice. The prevailing
school of thought is a minute EBW subjected to a very brief pulse, which
serves as a choke point to the pulse, thereby condensing the energy.
The problems with this is that it requires very careful engineering with
no ' wiggle ' room. If something is off by just a bit due to unforeseen
parasitic losses , either because of lowered voltage reducing the needed
energy, or stray inductance sapping it, then there is no bang.
In my view and the point of starting this thread is that everything
is wrong with this established approach and that it may work at all
is a wonder.
.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Quote: Originally posted by franklyn ![](images/xpblue/lastpost.gif) |
You stated
~ 4 Joules = {( 20 uf ) X ( 630 volts ) X ( 630 volts )} / 2
10 Joules at 1000 volts is not realistic , if the rating is
substantially exceeded the caps will fail , count on it. |
I know what my capacitors are rated for. From the data sheet,
DC test voltage -- 1.6 * V_R (2 sec.)
Operating voltage for short periods -- 1.25 * V_R (2000 hr.)
They are 630VDC (250VAC) units, so they are safe to use at about 1008V for short periods (probably with reduced capacitance as a result, because these
are self-healing type capacitors.) The datasheet also claims 60mohm typical ESR, which is around 3mohm for the 200 in parallel. (SRF is claimed as
around 5MHz, but the hardware connecting them drops that noticably, to about 0.1uH total and a bulk SRF of 114kHz.)
Quote: | Of course you can get around this by placing 2 caps in series
to withstand 2 X 630 volts. The energy will remain ~ 4 joules
for the reduced equivalent capacitance of 5 uf |
No, the energy doubles because capacitance halved and voltage doubled. Or even more obviously, energy doubled because stuff doubled.
Quote: | Discharge time will not change |
Actually, it will probably be longer, but how much depends on geometry.
Quote: | although the actual output certainly will be less. Read my third post above for why that is. |
Not at all certain. In fact, the two reasons capacitors are connected in series are these: one, to provide higher voltage capacity than otherwise
available; two, to increase dI/dt (that is, rise time, and peak power as a direct result).
Quote: | The ESR ( Equivalent Series Resistance ) of the capacitor(s) determines
what the discharge time will be. By your numbers the ' RC ' time constant
has to be (.003 ohm ) X (.00002 farad ) = 60 nanoseconds |
I would be quite impressed if you could suspend Ampere's law during a test fire of my capacitor bank. Sadly, Maxwell's equations come in fours, so
this cannot be.
As I clearly stated, my bank has a measured 100nH series inductance, thus forming a series resonant RLC circuit. As time constants go, the L
dominates over the R, which is why your RC figure is absurd.
Quote: | Extremely short ( a few millionths of a second ) discharge times
[unless there is some other need besides just exploding a wire] is
wholly uncalled for and unnecessary. |
It is my understanding that EBWs require sharp rise times in order to produce a shockwave per se. Is that correct?
A shockwave at 10km/s crosses a 28AWG wire (about 10 mils = 0.25mm diameter) in 25 nanoseconds, so it stands to reason that the wire must explode
about as fast in order to produce a complementary shockwave. That's something you certainly will not accomplish with an electrolytic.
If "shockwaving bridgewires" are, in fact, unnecessary, and mere "turning to an arc bridgewires" are sufficient, then yes, you could get away with an
electrolytic. This is something else I do not know.
Quote: | A very big problem which rears up with short discharge times is
the skin effect which overshadows any stray inductance by at
least another 2 orders of magnitude. ( 100 times or greater ) |
Non sequitur. Skin effect is due to self inductance. Skin effect also has no effect, given this kind of wire (you need significant energy beyond
10MHz to make 28AWG wire look futile; my cap bank will have little beyond 1MHz).
If stray inductance is a problem (for instance, it is entirely the reason my cap bank has 100nH ESL -- although that's much better than the 500nH to
1uH a more naive layout would yield!), it can be shielded to some extent. Transmission lines, ground planes and coaxial fixtures are all worth
considering in this regard. This is precisely how the atom bomb makers delivered nanosecond pulses to the slappers.
Quote: |
You cite 2.3 usec ( I'll humor you for the moment )
frequency is the reciprocal of time , f = 1/ t , then the 2.3 millionths of a second
translates to a frequency of 435 kHz. You will need at the very
least a 32 AWG wire or much preferably smaller to lessen attenuation
of the pulse. |
Again, as clearly stated in my post, 2.3us is the quarter wave time. This is the time it takes to transfer all the energy from one store to the
other, i.e., from capacitance to inductance or back again. As you can readily calculate from circuit values, actual series resonant frequency
(complete cycles of energy transfer) is around 110kHz. Please read my posts in their entirety -- it's not at all a chore, as my posts are generally
short and concise (although I am making an exception in writing this one).
Quote: | In my view and the point of starting this thread is that everything
is wrong with this established approach and that it may work at all
is a wonder.
|
The only thing that I gather from your reply is that your established approach is wrong. I am sorry to say this makes four people* I have met on this
forum who inextricably seem to know more electronics than myself, despite showing no proof of their talents, meanwhile attempting to proclaim their
dangerously lacking knowledge as incontrovertible testament.
I do not know much about exploding bridgewires, but if you are attempting to apply your present electronics knowledge, it does not surprise me in the
least that you are drawing conclusions like "everything is wrong with the established approach". Instead of attempting to invalidate someone or
something with much more experience than yourself, perhaps you should check yourself first.
Tim
* Four screen names, that is. Has anyone else wondered how you type exactly the same as Rosco? Oh, and it's still really badly
formatted.
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
@ 12AX7
I was going to add this opening bit to the item you're replying to
by all means don't take my word for it.
" Electrolytic? Pffifle. Too much inductance."
http://www.bychoice.com/capacitor_DF.pdf
Bottom of page 6
In the case of a capacitor, particularly in the low frequency range
(30Khz and below), the XL term is extremely small compared to XC
and can be ignored for computation purposes.
http://www.avx.com/docs/techinfo/eqtant.pdf
Read first 3 paragraphs page 1 , ending with the following _
" ESL is partly associated with the body of the capacitor
and partly with the leads: the value of the latter part is
proportional to the length of lead left on the capacitor
when it is mounted in its circuit location. Provided that
these leads are kept short, the effect of inductance can
be ignored at frequencies below about 100kHz."
From this we see that inductance is mainly a property
of the connecting circuit rather than contribution from
the electrolytic capacitor itself.
General atomics has perhaps the best overview of ESR
http://www.gaep.com/tech-bulletins/capacitor-engineering-bulletins....
_____________________________________
As to your rambling response - I have absolutely no idea what you
are going on about. My take is that you are getting lost in math
(which you do not disclose ) that has no bearing on pulse discharge
physics, as you say " resonant RLC circuit " , is not really applicable.
For example: capacitors in parallel are figured thus
10uf + 10uf = 20 uf
capacitors in series thus :
10uf X 10uf = 5 uf
10uf + 10uf
100 pairs of 2 in series gets you to the same result. You didn't state it
but I'm guessing 100nf caps , ( 20uf / 200 ). If you want 10 uf equivalent
capacitance you would need 400 capacitors 200 pairs 2 in series. Yes
double energy for double stuff. The original 20 uf bank would need to be
800 caps 400 pairs 2 in series.
_____
Induction is the direct result of current, the more there is the greater
the induction. Skin effect is just the result of rapid change in current.
Taking an RC time constant as the reciprocal of frequency as I do is
more of a rule of thumb approximation since the discharge is a spike
and not a repeating waveform as such ( another reason circuit math
doesn't mean much applied to this )
_____
I've been involved In electronics manufacturing all my adult life.
I never got my EE, personal circumstances compelled me to drop out.
I did get as far as partial differential equations and Fourier transforms
all of which I have since forgotten. Maybe that's why.
I'm reminded of Dorothy Parker, member of the Algonquin Round Table
here in New York, who playing a parlor word game was once asked to
make a sentence with the word horticulture and quipped,
" You can lead a whore to culture but you can't make her drink " , alas
.
|
|
Hennig Brand
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1284
Registered: 7-6-2009
Member Is Offline
Mood: No Mood
|
|
My ESR meter(which I just dug out), says the cap I have(25uF@2KV,oil filled cylindrical can) is very close to zero esr. My meter is not very good for
this at it is mostly used for testing to see if caps are fried or not. It looks to be around 0.05 ohms, which would give me a time constant(for the
cap alone) of 1.25 microseconds right? Assuming everything else to be zero in the "perfect circuit", then this cap could potentially deliver 33 joules
in that first time constant, since 2/3 roughly of the charge is delivered in the first time constant right? Assuming the ESD curves do look the same
proportionately, more or less for the different capacitors(even electrolytics), would it be possible with the super fast, HV capacitor banks to build
up more energy in the metal wire? Higher voltage caps with smaller capacitance will have a smaller time constant, since esr is usually smaller and
capacitance is smaller. My feeling from common electrical laws is that the higher voltage will give less speed and power losses in the hook-up lines
if done right(I think). This could maybe cause a more violent explosion of the wire from more energy being built up before the inertia of the wire is
overcome(before explosive vaporization)? The higher tension caps could maybe also provide more umpf were it counts directly following, and during
explosive vaporization? Maybe most of the overall speed differences are caused by the blast machine circuit and not the capacitors though usually? I
will read more since I have the feeling you will point out something, about how the differences can be dealt with by using thicker bridge wire,
different cable and circuit etc(I think you may have already once or twice).
One thing I have noticed though is that electrolytics usually won' t cut the mustard if one wants the most tough reliable bank(they won' t take the
abuse the others will). The rise time in the discharge curve is much longer, and can be more unpredictable, largely because of the more delayed and
drawn out rise(greater error potential). They are just often very cheaply made to, even for what they are. One of the biggest reasons(commercially)
for using these devices(EBW) is for multiple point, simultaneous, accurate detonations. There is not many fractions of a usec to spare many times
before best results are not obtained and the system impractical. It does seem though that the high voltages of commercial units is not necessarily
advantageous for single shots(single point), for the hobbiest as you have shown, which I didn' t understand before(took it for granted HV was
neccessary).
[Edited on 9-7-2009 by Hennig Brand]
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
@ Hennig Brand
The .05 ohm ESR you measured is an expected value for paper oil caps.
Don't dwell on charge, a capacitor's energy is in its voltage, work from that.
I explained above that after 1.5 ( R X C ) time constant, 95 % of the energy
has been discharged, and this can be considered the useful pulse width.
Yes there remains another 3.5 time constants until the capacitor has completely
discharged, but the remaining 5 % is very low grade energy, similar to the hot
exhaust from an engine after most of the work has already been done.
The energy is not proportional to the voltage , it is a square function of voltage.
After 1 ( R X C ) time constant the remaining voltage is 37 % , but energy just 14 %.
The 63 % drop in voltage accounts for 86 % of the energy.
by ( C x V x V )/ 2 , = (.000025 uf ) X ( 2000 volts) X ( 2000 volts) / 2 , = 50 Joules
so 50 X 0.86 = 43 Joules , expended after the first 1.25 microsecond.
The physics of this are , ESR will determine the time constant , and to a lesser
extent manipulating the capacitance and voltage relation to have the energy
needed to explode a particular wire. Because , to have a short time constant
you really also need a small capacitance , ( remember R C is R and also C )
In order to have meaningful energy this then has to be charged to very high
voltage. You could instead work at half that voltage with four capacitors of
equal value in parallel instead. But why stop there ?
The quarrel I have with this approach apart from the ancillary headaches of
dealing with high voltage and skin effect of hyper rapid current rise , is a short
pulse only matters if you need high precision in timing a pulse. Any ordinary
application of pulsed power requires no more precision timing than can be had
with a doorbell push button, so why bother if you have no application for it.
The question then too is do you have a Krytron available and the resources of
a NIST certified calibration lab to take full advantage of this potential precision.
This is power switching not small signal electronics , any ordinary gating scheme
will trigger variably plus or minus in tens of microseconds , this is known as jitter.
A few nanosecond pulse can occur anywhere along within that time frame.
It is as if you use an atomic cesium clock to time the boiling of eggs, or drive
a world land speed record holding jet car to the local convenience store.
.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
And inductance simply doesn't matter? I ask again, are you able to suspend Ampere's law at will?
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by 12AX7 ![](images/xpblue/lastpost.gif) | And inductance simply doesn't matter? I ask again, are you able to suspend Ampere's law at will? | Apparently
so, even if no one else is.
This discussion, though, brings up a couple of questions for me.
What's the proper equivalent circuit for the system during discharge? (Repeat: "during discharge") I assume there's a
capacitor on the left, with its parasitic elements. What else? At the time scales that are being bandied about, I can't tell whether you should model
the extension wires as a transmission line, or whether their bulk inductance is adequate. Does the inductance of a spiral-wound filament matter? How
does the circuit change as the filament heats, melts, and vaporizes?
What's the actual goal at the load? I presume you don't get explosion unless you vaporize the filament, which means paying for two phase
transition energies. What's the mass of an appropriate filament? Is vaporization the actual goal? Or is something lesser acceptable?
[Edit] Some answers:
Vaporization is indeed the goal. The Wikipedia page on EBW has further details, including this: "The wire used in the bridge tends to be highly pure gold or platinum, 0.02–-0.05 mm in diameter, and 1
mm long." Now it's clear they're using gold for low resistance.
The same page also mentions "Low-impedance capacitors and low-impedance coaxial cables are required to achieve the necessary current rise
rate." Given that coax is common, I'd say it's proper to model an extension cable as a transmission line.
Also "the peak current required ranges between 500 and 1000 amperes". That's a solid engineering goal for the load.
[Edited on 11-7-2009 by watson.fawkes]
|
|
Hennig Brand
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1284
Registered: 7-6-2009
Member Is Offline
Mood: No Mood
|
|
I did make a math boo-boo, it does change things for sure. I can' t remember what I did exactly, but I may have been calculating for coulombs then
said joules(thats what it looks like I did), tricky, tricky! Will try and avoid charge and stay with joules and volts from now on. I am starting to
think that it is pretty difficult to tell exactly what is taking place in this EBW senario, even with a good command of the electronics laws. Even
normal electronics, at slower speeds etc, is studied mostly by looking at the effects of things changing that we can see, or using relatively simple
instruments(except maybe by high level people). I sense, that to get a good feel, or any degree of accuracy would take some special doing. The easiest
thing is to probably test, by trial and error, as is the case with a lot of hobby science experimenting(If the goal is infact to initiate explosives).
The really small things probably come into play, and make huge differences. I am not trying to say we should abandon trying to understand as much as
we can theoretically however.
I think you may be right about the over-emphasis on the super short pulses, and it may not be as necessary as many tend to feel it is. We may be
drawing some false conclusions, based on information regarding specific uses differing from ours for EBW caps. The amount of energy that goes into the
before, and during explosive vaporization phase must be different for the different speed pulses though. Maybe this is not significant, but I am
interested, and others may be able to provide some insite that I cannot. There must be a very limited portion of that first 1.5 TC that has the most
impact on initiating ability. I would think as with chemical explosions, that not all are created equally, but it may be different in some ways for
EBW. Maybe the differences aren' t that great, it would be fun to test maybe.
In a lot of blasting manuals it describes EBW technology as being not too uncommon as a way of initiating multiple point source explosions(I don' t
know what kind of accuracy and precision they get, or how much is needed for their purposes, but it is apparently used). The Germans apparently used
this technology as well during the last World War, for mining and demolition(I think I read this awhile back). I would love to find some detailed
information about their machines and how they were used in blasting.
[Edited on 10-7-2009 by Hennig Brand]
|
|
dann2
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1523
Registered: 31-1-2007
Member Is Offline
Mood: No Mood
|
|
Hello,
Paper here (my favourite communication tool as it gives the impression I actually know WTF they are saying ).
What is the length (approx.) time period involved when we declare that an EBW is actually able to detonate secondary explosives? (assuming we have
enough Joules in that time period). There must be some (approx. ballpark) time period where time greater that this maximum period are considered
simple too slow no matter that the pulse energy.
From the paper (it's about making nano powders and has pulse times in the order of 2 to 4 micro seconds) there is a graph below and some text. It is a
worthwhile read.
_________________________________________
The copper wires used in our experiment are 196 micro m in
diameter and 85 mm in length, which yields an equivalent
inductance of 0.11 micro H calculated using Eq. (1). Figure 2 presents
the waveforms of the discharge current and voltage
measured in the experiment in which the energy storage
capacitor was charged to a voltage of 20 kV. It shows a
typical picture of exploding wire. The voltage begins to
rise strongly when the current reaches its maximum of
about 10 kA and then falls down, which means that the
vaporization of the wire begins, leading to a rapid increase
of the wire resistance. After the current falls down to a
value of 3 kA, it rises up again, which represents an arc
breakdown through the wire vapor, a shunting of the
current by this low resistance arc.
It should be noted that the experimentally measured
voltage shown in Figure 2 is not the voltage of the wire resistance
but rather the voltage across L2 and R2 in addition to the
wire voltage. The voltage of the wire resistance is important
for us to make an estimation of the energy deposition in the
wire before the explosion.
__________________________________________
Fig 2 below. L2 and R2 are the resistance and Inductance of leads inside vacuum chamber (unavoidable).
![wire_exp.jpg - 19kB](https://www.sciencemadness.org/whisper/files.php?pid=157260&aid=8142)
[Edited on 11-7-2009 by dann2]
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
Now that I understand more about how
these things works, I can say with confidence that this is just wrong. The details have to do with the phase transitions. When the bridgewire turns
molten, it will start melting out whatever its in contact with. In the (absurd) long limit, if the pulse is slow enough it will melt the wire, which
will run out after melting through its confinement.
This leads to the vaporization transition. If this transition is slow enough, escaping vapor will similarly escape through whatever gas channel is
available. But when, in the correct option, vaporization is fast enough, simple gas kinetics retard outflow well enough to create a shock wave, which
is the whole point.
The upshot is that the basic operation of this device requires a short pulse.
|
|
Hennig Brand
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1284
Registered: 7-6-2009
Member Is Offline
Mood: No Mood
|
|
I don' t know exactly were the limits are with this, because I haven' t studied it enough, but this is mostly my feeling as well. I am going to try
and find some more information, or try and possibly do a little testing. With explosives anyway, not all shockwaves are equal, and the more brisant
make a much more dense, fast and powerful shockwave(if I have said this right). What different secondaries require from the EBW could be wildly
different as well. For the sake of arguement I guess PETN is normally the standard though in EBW technology(thought I should say so, so as not to
confuse the issue).
[Edited on 11-7-2009 by Hennig Brand]
|
|
dann2
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1523
Registered: 31-1-2007
Member Is Offline
Mood: No Mood
|
|
Hello,
Doing some reading myself on time frames (my own link!) at this url
http://www.teledynerisi.com/0products/8td/page03.html (this was posted before BTW) it seems to give the ball park figure of needing a rate of
current rise of 1000 Amps per micro second.
The capacitor is close to the bridge wire using a short transmission cable with a spark gap between the capacitor and bridge wire. The cap is charged
relatively slowly and when the spark gap fires (at some kilo Volts) the charge is dumped into the bridge wire.
You would need a heavy transmission line. Perhaps two flat Copper conducting planes seperated by the appropriate insulation layer.
Multiple firing points could be set off at the same time with a control line to the 'spark gap'. The spark gap being replaced with a solid state
trigger of some sort. (I guess).
Regarding using Gold or Pt for the bridge wire. Perhaps it is these metals ability to avoid corrosion etc that they are used. Silver or Copper is more
conductive but much less inert. The (variable depth depending on age etc) oxide layer that may form on them may be considered unacceptable.
edit:
According the manufacturers they use Gold as it is very stable for years. (its in one of the pdf attached).
Attached file(s) (in zip) 0298.pdf gives inductance values etc for detonators and cables used in firing and books.
0792.pdf using longer cables
@ Hinnig Brand
0592.pdf for delayed firings (sound weird to me but I guess it must work)
see page5.pdf for some stuff on times and what happens as the pulse progresses.
If you Google
technical site:http://www.teledynerisi.com
you will gets lots of PDF on technical stuff on the systems that they sell.
I Google and downloaded and attached the lot.
Its a bit of a hod podge but good reading IMHO.
Dann2
Dann2
[Edited on 11-7-2009 by dann2]
Attachment: ewb.zip (873kB) This file has been downloaded 944 times
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
Quote: Originally posted by Hennig Brand ![](images/xpblue/lastpost.gif) |
I am starting to think that it is pretty difficult to tell exactly what is taking place in this EBW senario, even with a good command of the
electronics laws. Even normal electronics, at slower speeds etc, is studied mostly by looking at the effects of things changing that we can see, or
using relatively simple instruments(except maybe by high level people). |
Referring, perhaps, to an oscilloscope? Those are fairly standard equipment, and a must-have even for mere repair techs. I wouldn't at all consider
a 'scope "high level". ![:)](./images/smilies/smile.gif)
You can get a fair base model DSO (that's Digital Storage Oscilloscope) for $500 or so new, or a better, older one for $100 on eBay pretty easy. You
pretty well need a DSO for this, since you only get one shot of waveform, kind of hard to observe on a live analog scope (although they did make
analog storage scopes!).
Quote: | I sense, that to get a good feel, or any degree of accuracy would take some special doing. The easiest thing is to probably test, by trial and error,
as is the case with a lot of hobby science experimenting(If the goal is infact to initiate explosives). The really small things probably come into
play, and make huge differences. I am not trying to say we should abandon trying to understand as much as we can theoretically however.
|
I don't think it's nearly as imposing. Now, trial and error is nice for a test, but if you're making more error than success, you don't get much from
it. The good old combination of theory, test and verify is much more effective, and produces a lot more data a lot quicker. Of course, that requires
theory and instrumentation, but these are not hard to come by either: even the basic RLC circuit tells one a lot about the general circuit parameters
to expect.
Tim
|
|
franklyn
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 3026
Registered: 30-5-2006
Location: Da Big Apple
Member Is Offline
Mood: No Mood
|
|
@ 1 2 A X 7
" And inductance simply doesn't matter? I ask again, are you able to suspend Ampere's law at will? "
Inductance matters to what ? The time constant chart above _
http://i41.tinypic.com/35mgob5.jpg
is the same for RC as it is for L/R. You either have one or the other
not both together. If the energy is in capacitance being discharged
it cannot also be in inductance , though this may acquire an induction
from the others discharge much the same as in a parallel tank circuit.
Inductive reactance will only serve to delay the onset of the pulse at
the terminals , not affect the discharge time , as only resistance can.
Capacitive reactance cancels Inductive reactance at the self resonance
frequency of the capacitor given by f = .159 / √ LC this is where only
R , remains , but all this is moot since the manufacturer has thoughtfully
provided this data in a lump sum value typically as dissipation factor.
I outlined here
http://www.sciencemadness.org/talk/viewthread.php?tid=5064#p...
how ESR may be obtained from DF thus E S R = DF / ( 6.28 f C )
This is the 3rd time I have given this reference - twice to you
http://www.gaep.com/tech-bulletins/capacitor-engineering-bul...
Read on page 7 "How the ESR is applied" on into page 8
Depending on method , the contrived value for ESR can vary some
but not so much that the calculated discharge will be far off from
the actual discharge time that would be measured.
@ watson.fawkes
"Now that I understand more about how these things works, I can say with confidence that this is just wrong.
The upshot is that the basic operation of this device requires a short pulse."
Pulse width is irrelevant to whether explosion occurs or not.
Pulse amplitude is what matters since the area under the
power curve is the energy delivered. Knowing what energy
is necessary to make vapor of a known amount of metal ,
it is then a matter of delivering that in excess. If you just
send the exact amount success is uncertain. Inertia and
theta pinch guarantees the current converts the metal into
a supercritical fluid at which point the mechanics of the
gas laws predominate.
I'm continuously baffled as to the insistence that power
supplied must cut off abruptly in order for the wire to be
rendered into gas. Which is what is done by making a pulse
short.
.
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by franklyn ![](images/xpblue/lastpost.gif) | Pulse width is irrelevant to whether explosion occurs or not. Pulse amplitude is what matters since the area under the power curve is the energy
delivered. Knowing what energy is necessary to make vapor of a known amount of metal , it is then a matter of delivering that in excess.
| Repeating a false statement doesn't make it true. Next time, please engage the argument and address kinetic
self-confinement, which is the magic principle behind how these things work.
According to your argument, it doesn't matter at all how long it takes to deliver the excess of energy. So I'll build a circuit to deliver it over a
day. Hell, I'll pump in 1000 times the total energy-to-vapor over that day. I most certainly will not get an explosion. If, on the other hand, I
deliver that same amount of energy in 1 μs, I certainly will.
|
|
watson.fawkes
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 2793
Registered: 16-8-2008
Member Is Offline
Mood: No Mood
|
|
It seems that the ordinary mode of
use for these things is to press them right into the initiator. If so, you want a chemically non-reactive metal, no so much for oxides as such, but
for all the other things metals and explosives make when in contact with each other. So this is apparently the notion of stability they need.
|
|
12AX7
Post Harlot
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 4803
Registered: 8-3-2005
Location: oscillating
Member Is Offline
Mood: informative
|
|
That is because you get an RC or L/R time constant precisely when the other element does not show up in the circuit. Both are most certainly present
in this circuit! What must I demonstrate to you to prove that this is the case? Wire has inductance, resistors have inductance, capacitors have
inductance, it is most certainly a mandatory part of modelling this circuit's response!
Quote: | Inductive reactance will only serve to delay the onset of the pulse at
the terminals , not affect the discharge time , as only resistance can. |
Blatantly wrong! PLEASE read up on RLC series resonance!
Quote: | Capacitive reactance cancels Inductive reactance at the self resonance
frequency of the capacitor given by f = .159 / √ LC this is where only
R , remains , but all this is moot since the manufacturer has thoughtfully
provided this data in a lump sum value typically as dissipation factor. |
Dissipation factor is another representation of ESR and has nothing to do with ESL.
It is precisely because inductive reactance "cancels" with capacitive reactance that the complete RLC circuit must be considered.
Quote: | Depending on method , the contrived value for ESR can vary some
but not so much that the calculated discharge will be far off from
the actual discharge time that would be measured. |
Ah yes, then how do you explain my capacitor bank discharging in microseconds? It is an empirically measured, indisputable fact that it discharges in
microseconds. How, then, are you able to construct a figure of 60ns, a figure which is blatantly wrong by two orders of magnitude? The only
explanation is that your method is wrong.
Tim
|
|
Hennig Brand
International Hazard
![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif) ![*](images/xpblue/star.gif)
Posts: 1284
Registered: 7-6-2009
Member Is Offline
Mood: No Mood
|
|
@franklyn
I see what you are saying, it is not that the kinetic self confinement is not important, or that pulse width is not important, and that the power
under the power curve is all that matters. It is that they are the same thing essentially, or describing the same phenomina(shape of curve and the
timing). The first power spike before the dip, tells use exactly how much energy is going into the bridge wire, and/or bridge wire circuit(depending
on where taking readings), before the explosive vaporization. This is so because(as has been stated and verified by reputable sources), the first
current rise is during the time the wire heats up(resistance becomes more on heating), then current drops(resistance becomes greater
still,vaporization takes place, but vapor held in place by inertia), then after current goes higher again after explosive vaporization(shockwave)
since resistance goes down(plasma very conductive). Observing these waveformes, will give a good picture of what takes place at the bridge wire during
the explosive vaporization. It also tells how much energy goes into each stage(pre vaporization, vaporization, post vaporization), by using area under
the curve(curves), and calculus normally, or computer(I think I said this approx right).
I would think that for normal purposes, even if you had to use a laquer, shellac or some kind of plastic coating on your bridge wire, even if a little
more power and/or a thicker BW had to be used it should still be workable I would think.
I just took some 40gauge wire(0.0799mm diameter) and made a bridge of approx 5mm length(arbitrarily). I used one of those $2-3 racket bug zappers to
charge up my 25uF(oil filled) cap to 600 volts, giving me about 4joules in first 1.5 time constant(1.25usec). I have done it 6 or more times now, and
it is a pretty healthy bang for sure, even with that small amount of power. It seems like it could initiate(just opinion), similar to silver fulminate
or something in effect(just opinion based on observation and sound).
By the way I believe when elctricity is studied at higher levels they have a little more gear than just a multimeter and an oscilloscope. Even when
just discussing oscilloscopes, there is horrendous differences in there sophistication.
It also says in that Wiki article under "mechanism of operation", that a current rise rate of 100 amperes per micro second is required, to develope a
shockwave(at least under their conditions). If I can do this correctly now. charge=CV, so 600V times 0.000025F, so 0.015 coulombs. In first time
contant(1.25usec) I will get about 63% of this, so 0.00945 coulombs. 1/1.25=0.8, 0.8x0.00945=0.00756 coulombs per usec.
0.00756coulombs/0.000001sec=7560 amperes for first microsec. It would seem that I have lots, especially if one doesn' t count the losses from the
circuit and switch, etc. It still seems that I have lots to play with, even at only 600V, with this small 0.0799mm diameter bridgewire.
I noticed in the Wiki article about EBW it states that all it takes is a current rise rate of 100A per usec, in order to get a shockwave. This doesn'
t say anything about that shockwaves ability to initiate secondaries though. In the Teledynerisi article it shows graphs with 5x this amount and
calling it a marginal firing, and thats with their (perfect) blasting cap. That article also talks of inductance, and says that if the inductance is
too great, because of too long a line or improper line or configuration, that the pulse will be drawn out so much as to not explode the bridge wire
properly"reducing the magnitude of the shock wave". It gives a flatter more gradual curve(not what we desire).Doesn' t a bigger time constant give us
a much more gradual curve as well in general. Those high capacitance caps can' t perform like the small capacitance caps, for the same energy. I
wonder what the difference in losses would be like between the high voltage(2000+volts) and the lower voltage caps(several hundred volts) in circuit?
[Edited on 11-7-2009 by Hennig Brand]
|
|
Pages:
1
2
3
4 |
|