| Pages:
1
..
4
5
6 |
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
I already said that the Xe lamp needed controlled current, they are negative impedance devices as are many gas discharge/arc. Poor regulation of the
current will cause various problems, and can make it not too useful for your intended application (although it may be successful at quick
scarification)
You're proposing replacing a lamp manufactured by a knowledgeable outfit with a home fabbed device; and then instead of using a DC supply with current
control you want to drive a variable load with a high power RF amp - I suggest you go study the term "impedance matching" for awhile.
Low pressure Xe discharge lamps do not give the nice 'white' spectrum of the high arc lamps, they give the xenon line spectrum, the term to study is
"spectral line broadening". Xenon has a lot of lines, but at lower pressures they still are discrete lines. As you raise the pressure the
continuous emission rise in intensity, but around 1 atmosphere you'll still see strong lines; BTW a lamp that is at 1 bar when cold will be above that
when running, your homebrew lamp will need to handle that while being a rather toasty temperature.
Yes, you can buy RF power amps, for much more money than the DC supply would cost. A microwave oven is already putting out more than 100 watts,
wireless communications devices are piss poor sources of RF for boosting up for gas discharge drivers. Leaking microwaves are a fun way to fry bits of
yourself, in some cases without knowing it right off as the nerves can be killed before they report getting hot.
All in all you've shown enough lack of comprehension of this general topic that I suspect your attempting to mach such a device could make it as a
Darwin Awards nominee. You __really__ need to do more study of these things your wish to build, if you want to have a reasonable chance of success.
|
|
|
aliced25
Hazard to Others
Posts: 262
Registered: 31-7-2010
Member Is Offline
Mood: No Mood
|
|
Well I'm waiting to see the data on that, according to several Russian Papers, Excimer-type lamps with Xe/Kr with a halide are the strongest sources
available for the 200-400nm region. The problem they had was the attack of the halides on the electrodes (which were inside the vial), which made it
interesting I'm sure and short-lived.
Then there is this, they dismantle a 12V, 2.5A supply and use it as a 5kV, 20mA supply in order to produce an Excimer-type lamp by simply wrapping the wire
around the vial (which is the normal way of doing things now I believe?).
That said, if it makes Xenon give plasma I really don't give a rats, it shows it is possible from a regulated power supply, now the only question is
how to translate that to a better quality light, and I'm not talking homemade.
Oh yeah, while I'm here, here is a fucking incredible site I wandered across, Luhs Laser, it is bloody incredible work, check out the various educational kits, the drawings & the pictures. It really is worth a look.
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Your detector and dispersive elements, at least those you've discussed in the past, aren't going to work well below about 350 nm, from there down to
200 nm wouldn't seem to be of concern.
Indeed, they make some low intensity glow lamps; this does not mean that the method is suitable for what you want to do - you'll either need to try
it, or find more details.
I myself would be leery of basing too much planning on research articles that include statements such as
| Quote: |
Computer codes do not calculate the discharge stability. Therefore, it is possible that the discharge with a positive column radius r = 4.5 cm and/or
[Xe], [Kr] ~ 4 × 10<sup>19</sup> cm–3 does not exist. Nevertheless, we present the results of calculations for these values.
|
It is also possible that there are fairies in the bottom of the garden, but I'd not take that to the bank.
You've now jumped from using HID lamps to homebrew (and yes you seemed to be talking about homemade back there) Xe lamps with microwave excitation to
rare gas-halogen lamps with lash-up glow discharge drive, all in as many posts. You've also gone from light sources that give fairly flat output from
the deep red down into the quartz UV region, to light sources intended for in a fairly narrow ban in the VUV. Again, I suggest picking one choice and
studying it so you really truly understand it, remembering that the same overarching term may apply to a fair wide range of devices, conditions, and
results. Example, nuclear magnetic resonance is used to measure degree of cure of a glue or the liquid water content of ice, for imaging living
things, and for doing molecular structure determination, however those classes of applications differ by about an order of magnitude of needed
accuracy for each step as you go from the first 2 to the mid to the last. A device can work wonderfully for the 1st two applications, and fail
miserably for the last.
BTW - the reversed transformer is current limited, I'd bet, although not very well. And a current controlling supply is as much a regulated power
supply as a voltage controlling one.
You may find this of interest http://www.opus.ub.uni-erlangen.de/opus/volltexte/2007/711/p...
|
|
|
aliced25
Hazard to Others
Posts: 262
Registered: 31-7-2010
Member Is Offline
Mood: No Mood
|
|
This is one of the papers on the microdischarge arrays, pretty much the same principle, though much neater. I'm reading as much as I can, very
quickly, but the glow-discharge lowers the spark-voltage of the Short-Arc by /2. It also produces light by itself, though mainly in the VUV - although
with either sapphire/F-Doped Fiber (both commercially available, though not cheap), that might yet be overcome (Lumogen is the answer to the first
question, really doesn't need discussing does it?). That microdischarge-microhollow plasma - given it does the 200-400 UV quite well, might be "AN"
answer, there are plenty of lights that will do the remainder of the spectrum on fuck all power. In fact, those Microhollow plasma emmitters, anode
and cathode about 1/2 mm apart, with a dielectric in the middle, with 1/2mm dia hole, enough of them should give pretty much the same spectra at STP (see p17/28) and high kV/ low mA as a D2 Lamp. I suspect the price would be lower though...
The barrier discharge and the microhollow-microdischarge all seem to be utilizing the tiny plasma formed in the presence of a high voltage - low
current electrical source (or RF/Electrical). I'm actually wondering why a short arc wouldn't work with, essentially, a surface spark plug-type
cathode in the center of the anode? It arcs in cylinder heads, so why wouldn't it arc there? It would certainly simplify the design issues of how to
get the cathode in the vicinity of the anode without blocking the light-path.
[Edited on 29-11-2010 by aliced25]
[Edited on 29-11-2010 by aliced25]
[Edited on 29-11-2010 by aliced25]
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
There aren't many lights that give reasonably uniform 'white' light across the quartz-UV through red. Nor are there common analysers, prism or
grating or fixed FT or whatever, that covers from the short QUV into VNIR; that's more than 2 octaves and SFAIK gets tricky. Lumogen is only a
solution, really part of a solution, if other factors are taken into consideration, show me your proposed design and then it can be commented on if it
is a solution.
Your speculation | Quote: | | . I'm actually wondering why a short arc wouldn't work with, essentially, a surface spark plug-type cathode in the center of the anode? It arcs in
cylinder heads, so why wouldn't it arc there? |
is to a degree answered in the slides you referenced:
| Quote: | For stable operation the product of pressure, p, times distance between cathodes, D, needs to be in the range between approximately 0.1 Torr cm and 10
Torr cm (depending on gas, electrodes, and geometry).
...
Electrode materials: molybdenum, semi-insulating silicon
Dielectrics: mica, ceramics
Dimensions (hole diameter): 30 µm to several 100 µm
|
The distance between electrodes appears to be on the same order of magnitude as the hole size.
Dimensions are important, the semi-insulating Si works as a ballast and to suppress the formation of current-hogging arcs, more evenly distributing
the discharge across the electrode surface.
| Quote: | | should give pretty much the same spectra at STP (see p17/28) and high kV/ low mA as a D2 Lamp |
Whatever. Xe Microdischarge, Deutarium, conventional Xe short arc :
|
|
|
aliced25
Hazard to Others
Posts: 262
Registered: 31-7-2010
Member Is Offline
Mood: No Mood
|
|
Sorry, lost my reply
I'm actually contemplating utilizing an almost 3K pixel long sensor (monochromatic cmos), to collect the spectra - I'm wondering, there are a number
of options regarding the light source, the DUV-VIS Fiber, etc. But with the width of the spectral bandwidth, it might be worthwhile looking at using 3
separate small gratings, having a 3 channel instrument, 150-450nm (I'm hoping the phosphor can cope), 400-800nm & 800-1150nm. There could be some
overlap, but I'll try and work out a way to minimize it, realistically though, the NIR blocking optics should do the job for me @~800nm and the
insensitivity of the sensor should do it for me @~400, so there is a bright side.
If high l/mm gratings were used, it should be possible to get resolution down to almost the tenth of a nm in some areas, which would be pretty fucking
spectacular.
With a grating optimized for 400-800 it might be worthwhile going for the blue/green lasers after all. The main problem with them originally was the
resolution, if the resolution is not a problem and the spectrometer could be realised for dual use, then a highly accurate Raman + UV-VIS-NIR Spectrum
might actually be worthwhile taking.
I'm just re-reading on the optics, I'm thinking maybe going with aberration corrected, flat-field gratings (cheaper than you'd think) specifically
optimized for each of the three regions. If it could also be used for Raman, the central channel with either the blue/green laser (or stick with the
red and go for the Anti-Stokes?), I'm thinking the DPSS is the best choice by far, it is way cheaper than the respective high-power blue's...
3 gratings + 3 sensors, 2 or more light sources and the associated optics. It won't be cheap, but it will still be a fucking lot less than purchasing
a basic spectrometer from the major players.
Damn, I'm sitting here going cross-eyed trying to work out the relative etenue of the spectrometer and the fiber. Might have another go in the
morning.
|
|
|
aliced25
Hazard to Others
Posts: 262
Registered: 31-7-2010
Member Is Offline
Mood: No Mood
|
|
The design will be based upon "hopefully" something like Ocean Optics is currently using - a three tier arrangement, which allows the full spectra to
be broken into its constituent parts 200-500nm, 500-800nm & 800-1,100nm, or 300nm/grating with the blaze at the central wavelength. As the
majority of reflective gratings are based upon aluminium coated epoxy, it should "technically" be possible to design a concave, aberration corrected,
3-level grating substrate that can be mounted so that the only light incident upon each section of the grating is the designed for part of the spectra
(ie. physical isolation - separate chambers within the same spectrometer). The first one would be a design challenge and a half, the replicas not so
much.*
The concave, aberration corrected grating is the key here, without it getting a flat-field image is going to be a fucking bitch. However, I'm not too
concerned, given the price of gratings (if one is willing to try and separate the wheat from the chaff amongst the manufacturer's directly - no easy
fucking task), the prices fall dramatically. Even to the point that it 'might' be feasible to actually utilize 3 separate gratings & 3 sensors to
do the job, instead of trying to compromise. In any event, I'll still be sticking to the CAC-Holographic Grating. It goes without saying that
designing the optics to cope with the etendue caused by using fiber optics to carry the light are going to be a major fuck around.
As to the use of Lumogen, there are other luminophor's/phosphor's that can and will allow for the detection of UV light - I'm actually wondering
whether it is possible to mount these on/in a thin film, directly in front of the sensor (AR Coating on both) to eleminate stray light. If the
phosphor is emitting above 500nm even stray light would not be a problem, the relevant grating is incapable of using it.
* I'm still considering how exactly to construct something like this, a single concave surface, with three replicated gratings molded onto it in
strips, separated by flat surfaces allowing for the optical separation of the three channels... Given that replica gratings aren't precisely new,
surely it is possible (if somewhat out of the square) to form multiple gratings on a single surface which is engineered to allow for 0-stray light
between the chambers/channels. The concave grating spectrometer, with its minimalistic optics and flat-field output seems well suited to this design.
PS With those separations, the options for a combination Raman-UV/VIS/NIR Unit would require either a 780nm Stokes-Shift setup, with the Raman
Spectrum being in the >800nm region, or an 808nm Anti-Stokes setup, with the Raman spectra in the <800nm region. In either event, the relevant
diode lasers are fairly cheap.
[Edited on 3-1-2011 by aliced25]
|
|
|
aliced25
Hazard to Others
Posts: 262
Registered: 31-7-2010
Member Is Offline
Mood: No Mood
|
|
Ok, having a look around - there are long, narrow, back illuminated Hamatsu photonics image sensors - they are fairly flat from 200-1,000nm. Now, if
we used a quartz xenon flash-lamp, and a reflective grating, that should allow us to collect the entire UV-VIS-NIR spectrum in one hit, then run it
100 times and average the lot. I'm also thinking that with a decent filtered laser diode, it should also be possible to use the same optoelectronics
(and grating) to grab the Raman Spectra using the same instrument & the same sample.
I'm putting some thought into this as I think there is a need for home-instrumentation that is cheaper than that currently available but which will do
the job. That said, the Hamatsu sensors aren't cheap.
From a Knight of the Realm: "Animated movies are not just for kids, they're also for adults who do a lot of drugs." Sir Paul McCartney
|
|
|
bquirky
Hazard to Others
Posts: 316
Registered: 22-10-2008
Location: Perth Western Australia
Member Is Offline
Mood: No Mood
|
|
if your not chasing high speed you can rotate your grating and use a single photo detector (then you can use several photo detectors with different
spectral responses for different wavelength ranges) and dont have to worry about exotic hamatsu sensors (which often have a low pixel count) as an
added bonus it gets you off the hook somewhat asfar as putting a flat field on your sensor. achieving that can be a real PITA
alot of high end OSA (optical spectrum analisers) do exactly this and can acquire very wide spectrums the only drawback is speed.
|
|
|
bquirky
Hazard to Others
Posts: 316
Registered: 22-10-2008
Location: Perth Western Australia
Member Is Offline
Mood: No Mood
|
|
oh and you can allso replace the photo detector with a photo multiplier tube and do single photon counting..
im sure you can find some reason to want to do that 
|
|
|
not_important
International Hazard
Posts: 3873
Registered: 21-7-2006
Member Is Offline
Mood: No Mood
|
|
Quote: Originally posted by bquirky  | | if your not chasing high speed you can rotate your grating and use a single photo detector (then you can use several photo detectors with different
spectral responses for different wavelength ranges) and dont have to worry about exotic hamatsu sensors (which often have a low pixel count) as an
added bonus it gets you off the hook somewhat as far as putting a flat field on your sensor. ... |
The original idea was to use a monolithic interferometer based on designs mentioned in the thread http://www.sciencemadness.org/talk/viewthread.php?tid=13554
They have no moving parts, not even a piezo-driven mirror, and if properly designed can handle quite wide wavelength ranges. They produce an
interference pattern on the detector array, which is resolved using Fourier transform processing. The Hamamatsu back thinned NMOS sensors would
theoretically be able to resolve around 1/2 to 1 nm in this application.
Note that Xe flash lamps run at roughly a tenth or less the pressure as do continuous Xe arc lamps, as a result the flash lamps spectrum has
noticeably more prominent lines. The energy distribution of the output also drifts towards the red end over the course of a single flash, not
important unless you're taking really fast samples.
Also you should look at the specs on gratings. Most don't cover all that wide of a wavelength band very well. Some examples here http://www.coseti.org/specgrat.htm
|
|
|
ScienceSquirrel
|
Thread Pruned
13-3-2012 at 05:39 |
| Pages:
1
..
4
5
6 |
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
