Sciencemadness Discussion Board - Schematic for Electrolysis Device

Schematic for Electrolysis Device

quicksilver - 24-1-2010 at 08:05

Prior to posting I searched quite a bit for a schematic for a Electrolysis design.
I have tried battery chargers but most today have an automatic shut-down feature that makes continued operation impossible.
Does anyone have such a schematic?

OR...has anyone found such a battery charger that would actually be used without spending a house payment?

We all know that the problem is that we are shorting the device. How do electro-plating devices function in that genre'?

I had thought of a charger going TO a battery and then to the solution but still th short makes most designs impracticable?

Any pointers, tips or actual designs would be deeply appreciated.

hissingnoise - 24-1-2010 at 08:17

I'm out of touch with modern chargers but I presume you mean a thermal cut-out.
If heat dissipation is the problem you could bolt on a heat-sink or use a cooling-fan.
All you need essentially, is a suitable transformer and bridge rectifier.
Bigger tranformers are better since heat build-up is lessened.

bbartlog - 24-1-2010 at 08:36

Battery charger I have has a little switch to toggle between 'auto' (stops when it thinks the 'battery' is charged) and 'manual' (you stop it when the needle tells you the battery is charged... or not, if you're running it for electrolysis). You might want to look for one of those.
Now of course that cutout happens when the resistance is too high... the opposite problem, shorting and overheating, are not something I've triggered. But it seems to me that if your cell is properly designed, the resistance should not be so low as to short or overheat a battery charger.

Contrabasso - 24-1-2010 at 09:56

It's a supply of about 6v at about 25 - 250 amps that you need not a battery charger!Ebay will help as will the youtube vid of how to canibalise an ATX PSU.

bbartlog - 24-1-2010 at 10:35

(car) battery charger and ATX PSU repurposing are pretty comparable in terms of capability, assuming you get a 25A or 50A battery charger. The ATX PSU is cheaper of course, but you need to do a little work. With the PSU you get 5/12V whereas the battery charger will usually be 6/12V. In both cases amps can be 20-50 with a decent sized unit.
Of course a real bench power supply would be far superior to either of these options.

entropy51 - 24-1-2010 at 10:40

I use a variac driving a stepdown transformer rated at 10 volts/10 amps. The transformer output is rectified by a Radioshack bridge rectifier rated 50 volts/25 amps. I have a voltmeter across the rectifier output and an ammeter in series with the electrolysis cell. Not particularly heavy duty, but simplicity itself.

quicksilver - 24-1-2010 at 11:01

Thanks you guys:

I REALLY tried on this one but I am just not an engineering type.

Let me explain further because you all obviously know a Hell of a lot more than me here. When we speak of a thermal cutout, I am making the leap that the switching transistors reach a point that they say "hey, we are getting hot from a full charge; let's stop". In this case, by making a more efficient heat sink - I see the logic, however it would APPEAR we'd need a fan too to spread the heat or localized heat would also give the same message...(?) no? Let's say that heat really does spread out... (logical) would that simply give me a bit more time for the same transistor to reach the same conception OR do a larger more efficient heat sink keep that thing just under the "magic" switch number?

Now with a modern ATX power supply; doesn't that reach the same point? OR are the I.C.'s in there say "Hey fans; speed up, the machine's getting hot"?

I looked for a manual battery charger (the one bbartlog mentioned) but those are VERY tough to find. I found one made cheaply and it simply opened a circuit breaker.

This project opened a whole host of things from pulling the HN4NO3 from CAN to making KCLO3 from water softener salt. The list is wonderful. But I sadly have to admit I just don't have the "chops" to rig this fellow up. I had thought of getting a 12vhigh amp transformer and heat sinking most everything + adding a fan. I DO have a electronics supply place in town that could hook me up with a 12v 20A, fuse it but I realized that the rectifier would also have to be a tough one.
I also believe that since three of you guys obviously know some strong, effective solutions; I am making a mistake (wouldn't be the 1st one). If we had to make a call on three differing methods: ATX supply, old fashioned battery charger, or a "hot-rod" charger (or similar) - which and why would be the best be for a guy household electric experience & minimal electronic engineering knowledge?

I honestly appreciate any input, because realistically; I'm at a loss at finding a method short of the electroplating professional machine (& I don't that much $ since I co-starred with Basil Rathbone in "The Woman in Green".

[Edited on 24-1-2010 by quicksilver]

Contrabasso - 24-1-2010 at 12:01

As a good method, get a 12v charger rated 25amps or more and drive it from a variac from your mains. That way you can adjust the voltage across the electrodes to control the current flowing. Keep the high current wires short and fat, keep the variac well away from the cell and the fumes and splashes. Ebay.com has several (hundred!) variacs some better and some nearer to you

It's the passage of electrons (ie current!) that does the work so though you may chose to monitor the voltage it's the current that matters

12AX7 - 24-1-2010 at 23:12

Quote: Originally posted by quicksilver

When we speak of a thermal cutout, I am making the leap that the switching transistors reach a point that they say "hey, we are getting hot from a full charge; let's stop".

That works in some cases, like a moderate overload. More amps = more voltage drop = even more power dissipated = gradual temp rise past limits = thermal cutout. OTOH, sudden overloads kill transistors instantaneously (~10us), for which I'm talking less than a 10x overload (for cheap computer supplies, this point comes at about 0.5 times rated output!). In this case, the five dollar transistor protects the ten cent fuse. Fuses (and thermal cutouts) are for protecting wiring, not electronics.

I'm considering building a 0-5V, 100A (adjustable volts/amps) power supply. Also considering power factor correction, universal input (80-268VAC with no 120/240 switch) and synchronous rectification, resulting in efficiency around 90% -- low enough that it could be sealed inside a nice stainless steel box (except for the small protruding aluminum heatsink) and operated in the same building as an electrolysis cell, with a long service life. I'm guessing the cost would be a few hundred bucks, but it'll never die.

I could also extend functionality with digital controls and a timer -- potentially, a very long duration timer, such as "run for a week", or "run for a million coulombs", both very useful for electrolysis. And I could add features like "stop if voltage or current exceeds xxx and note failure time". Hmm, brownouts over that time scale are a possible problem; I'm not sure if it would be effective to store "where I left off at" in nonvolatile memory, so it can remember previous settings, including interrupted run time. A brownout detector wouldn't work fast enough, but maybe a capacitor would be enough to shut down on. Alternately, it could operate out of NVRAM (battery backed), or use FeRAM, which is NV without the battery. Ah yes, they probably make little serial FeRAMs... that would be the ticket...

Tim

hissingnoise - 25-1-2010 at 04:20

Quote:

I'm considering building a 0-5V, 100A (adjustable volts/amps) power supply.

I made a low-tech version 12AX7.
It consists of a NST trans. sawn in half with the HT secondary removed.
If I want to, now, I can wind multiple secondaries.
The two halves fit back together nicely but it hums and vibrates something awful when running - but hey, it works; and it makes me work as well.
And of course I do a lot of guessing. . .

woelen - 25-1-2010 at 04:35

I have written a completely worked out recipe for converting an ATX power supply to a power supply which can be used for electrolysis. With these instructions even the really non-engineering type of people should be able to make a decent power supply for just a small amount of money. You might even be able to get your hands on an old ATX supply for free, e.g. when an old out-of-date PC is replaced.

http://woelen.homescience.net/science/chem/misc/psu.html

Using 12 V directly for electrolysis is too much, you will have excessive wear of electrodes and a lot of side reactions and a lot of heat. Putting two cells in series gives a double output, another option is to use series resistors for feedback control, allowing very constant and predictable behavior of your cell.

I am very happy with this type of power supply. I can easily draw 10 A at 12 V from my supply, but modern 500 W supplies can deliver well over 20 A at 12 V (possibly divided over two separate rails). In my experiments, however, I seldomly draw more than 3 A from the supply. At higher currents the cell may become very hot and you also can have excessive wear of electrodes.

quicksilver - 25-1-2010 at 07:50

@ woelen :
That was well done. I will give that one a shot ASAP --THOSE parts are free! I am actually impressed that so many folks have thought about this from so many different angles!

pip - 30-1-2010 at 07:59

woelen would you know how to boost the power output on a pc type power supply but still for pc use? I have ran out of power in my xbox and need to increase 12v power by an amp and the xbox uses a propritery power supply (I think)

I tried to figure it out myself and the best I can figure out is changing the tranformer, am I right and would anything else need upgrading?

Contrabasso - 30-1-2010 at 13:26

You will be very lucky if you modify a PC PSU for increased power, Better buy a higher power one.

dann2 - 30-1-2010 at 18:40

Hello,

There are some circuits here that will give a constant current from a computer supply. They are not too difficult to build. A constant current supply is the biz for what you are doing.

Dann2

quicksilver - 6-2-2010 at 10:17

I had completed woelen's design and was VERY impressed. Cheap, easy, flexible design. I HAD thought of using a MOT with the secondary's modified, etc but woelen 's design made me actually laugh it's such a flexible good idea. - You could build upon the basis of his idea to include other things. If I was to consider what to build I would do it again!

One of the issues with using HV parts is they are becoming very tough to get now. I have a little collection of NST's (nice old one like France) and I don't know if I could get that stuff anymore. BUT MOTS and computer PSU....that's going to be free for a LONG time!
What I had thought about was replacing a modified MOT into a charger and upping the strength of the switching transistors and most other parts that have a terminal level of current substantially lower than whatever I made from the MOT.

Simple, workable MOT design for 30 amp PSU:

http://www.youtube.com/watch?v=Yr8HKlX4VnU

[Edited on 7-2-2010 by quicksilver]

densest - 7-2-2010 at 13:55

Yes, woelen's design is very nice indeed.

A "quick and dirty" approach I have used is to salvage an appropriate connector from a discarded motherboard and attach all of the necessary resistors and other wires onto that connector. The connector and attached tangle of wires needs to be glued or tie-wrapped to a board to be stable, or (for someone less lazy than I) put into a box. It does save opening the PSU case and allows the PSU to be swapped out if necessary. All of the resistor values, etc., that woelen specifies are applicable.

For a low voltage cell (silver or gold electrolysis, for example) the 3.3V or 5V outputs are very high current and are closer to the cell voltage, thus needing to waste less power as heat to regulate the cell current.

For the electronics obsessed, here's an example of a switch-mode constant current source which could work well hooked to the 12V output: http://www.edn.com/contents/images/53002di.pdf It's limited to 2A by the LM2576; it would be easy to scale up to 20A or more with a different chip and an external transistor.

The example as shown uses a TL082 dual op amp which requires +- 15 volts to operate (at least the + supply must be more than 2V over the highest input voltage expected). More modern op amps such as the LM6132, LT1492, or LT1211 will work with V- tied to ground and V+ tied to the input supply. Other more exotic chips such as a current sense amplifier like the LT6106 would allow high input voltages and use many fewer parts - only 3 resistors, 2 or 3 capacitors, the switch chip, catch diode, inductor, and current sense chip. The biggest disadvantage is that the current sense amplifier chips mostly are not made in DIP packages so they are harder for a hobbyist to use.

Other interesting additions would be a voltage limit or monitor to alert the user or turn off the circuit if the cell voltage changes, for instance, when the chloride concentration of a chlorate cell gets too low. Computer-compatible inputs and outputs for control and monitoring would be reasonably simple as well.

@12AX7 - your "universal" PSU is interesting; maybe something like an expanded version of this as a post-regulator for a computer PSU would be a place to start? Designing an off-line PSU can be exciting, but this would fit in a little box and have few safety implications... If anyone wants something like this perhaps we could collaborate on something? Once this existed, adding the Heavy Duty front end is a separate project - for instance, a 48V telecom supply picked up from EBay would serve.

quicksilver - 8-2-2010 at 09:33

Well then there's always HARBOR FREIGHT. They have some serious Chinese & Indian welders and chargers for about $100 and IF the base parts are there and will stand up the idea of a very high amp PSU is possible.
There is a place near me that sell electronics surplus and has very serious resistors for x-ray, power station, etc application and those can be had for about $9, making the cheap stick welders Harbor Freight has at $100 very much in the running.
But the idea of using a MOT - re-wound is a very simple way to get amps with virtually NO $ invested. I've torn apart monitors and TV's looking for flyback transformers and other HV goods and found some 20watt resistors & since the standard CRT monitor is now a throw-a-way item the parts are free.

quicksilver - 16-2-2010 at 14:16

It seems there are two ways to go if an individual does not get lucky and find a high end switching power supply: woelen's design or an OLD battery charger that will maintain "manual" operation and not shut down. Those chargers are a lucky find in today's market.
I was looking at common battery chargers and opened a very common brand (Schumacher) but I could not determine the device that senses when the charge has completed.

Since both concepts of PSU's are generally similar in price & availability, it would be a good thing if there was a way to make a 10A+ charger maintain it's charging. I tried a parallel group of moderate wattage low amp resistors to allow the Charger to sense that "the battery" was still needing a charge, but to no avail. Is there a way to get Battery chargers to function as PSU? Because now most all common brands are "automatic" in their shutdown procedure to protect the battery from over charging; making their use on a continuum as challenging as an unaltered computer PSU.

woelen - 17-2-2010 at 08:20

Im also experimented somewhat with battery chargers and it seems that these devices have some voltage sense circuitry. These chargers do not have a stiff power output, but fancy wave forms are generated. A common mechanism is a pulse at fairly high voltage (e.g. 14 volts) and then the output goes low (high impedance) for a while and at that time the voltage, delivered by the battery is measured. If that voltage exceeds a certain value, then the battery is considered charged.

Some chargers also simply have a timer circuit, sometimes combined with some current sensing mechanism. Usually a simple controller chip in a SMD package controls the entire system and you cannot easily locate the subsystem which does the actual shutdown. The SMD device can do the voltage regulation, it can contain the timer, the wave form generator and the measuring circuitry at a high level of integration.

I quit further investigating a battery charger. Too much hassle because of the availability of another good alternative in the form of an ATX power supply.

quicksilver - 18-2-2010 at 12:17

I am glad you looked into it. I had a feeling that there would be too many variables in different chargers and working with SMD's just seems a pain.

Perhaps in the future some will find a simple solution to the sensing issue (aside from dealing with some dead motorcycle battery in parallel) :-)

Contrabasso - 18-2-2010 at 14:53

With the advent of sealed batteries in cars there is a need to prevent people overcharging them and gassing away the electrolyte. So battery chargers now sense something to determine a cut off time.

One of the hobby electronics shops in your country must supply a high current transformer with about 6 - 8 volts on a secondary with enough available amps to make it a good choice. Then simply a bridge rectifier and a chunky rheostat to limit current and an ammeter. Simple and effective.

Added 546-165V7 from Mouser will be a great start.

[Edited on 18-2-2010 by Contrabasso]

entropy51 - 18-2-2010 at 15:15

Quote: Originally posted by entropy51

I use a variac driving a stepdown transformer rated at 10 volts/10 amps. The transformer output is rectified by a Radioshack bridge rectifier rated 50 volts/25 amps. I have a voltmeter across the rectifier output and an ammeter in series with the electrolysis cell. Not particularly heavy duty, but simplicity itself.

I like your approach, Contrabasso!

Lambda - 19-2-2010 at 02:32

Rewinding a MOT's Secondary with about One Winding per Volt Output works just fine, and they are free of charge. The MOT Transformer is very rarely the culprit in a broken down Magnetron Oven, and at most, the secondary has shorted out, witch we don't need anymore. The Secondary winding is in anycase removed, just be careful not to damage the primary winding during this procedure. And use sufficiently thick wire to rewind the Secondary, better too thick than too thin. Car Audio Multi-Core Battery Power Wire is Thick and Flexible, or you can use Thick Single or Multicore Core and Isolated Copper Earthing Wire used in Homes and Buildings. Single Core Copper Wire used at 50-60 Hz is just fine, for skin effects are irrelevant at these very low frequencies. Just Don't rewind a Computer PSU with thick single stranded Mono-Core wire in this way, for skin effects then do play a significant role at the frequencies used in these Units.

Dans Homebuilt arc welder:
http://www.dansworkshop.com/electricity-and-electronics/home...

How I Built My 70 Amp Arc Welder:
http://aaawelder.com/70amp.html

Mega Buzz Box 100 Amps:
http://aaawelder.com/indextoo.html

How-to: Build your own spot welder:
http://hackaday.com/2009/06/23/how-to-build-your-own-spot-we...

Micro Wave ovens are a source of treasures for the tinkerer:
http://www.home-workshop.com/MOTWelder.htm

Destructive Testing, 7V Rewound MOT:
http://www.youtube.com/watch?v=_AqItwvYq7s

I am fortunate to have Bridge Rectifiers of hundreds of Amps, but if you are not, then eBay may be the solution. Or, you can wheedle them out of a Car (12 Volts) or Lory (24 Volts) Alternator/Dynamo. You guys are creative enough to find "Free" sources like I do, from old Lift Motors (or what ever, even old Crane or Industrial Motors) and old Mainframe Computer Supplies (MOT not needed anymore) found in scrapyards. Or just become a Dumpster Diver like me, it's fun and rewarding too. Money lies on the streets for those who seek, and find ...

Lily Allen - The Fear:
http://www.youtube.com/watch?v=q-wGMlSuX_c

Lambda (Always Diving where the Sun does not Shine, ... and my Girlfriends love it

)

Lambda - 20-2-2010 at 12:27

40 Amp Power Supply Unit

http://www.qsl.net/vu2upx/Projects/40apsu.htm

*Input voltage: 240
*Output voltage: 13.2V DC
*Output current: 40A DC
*Over current protection: current limiting al 40 Amps
*Short circuit protection: Regulator shut off.
*Over voltage protection: Shuts off DC input and discharges input stage reservoir.
*Over temperature protection: Fan automatically operates as heat sink temperature of 65°C.
*Indicators: Power ON LED and Short circuit Protection active LED.

This Power Supply concept is based on the popular 723 Chip, and can easily be re-engineered to fit our needs. Schematic and PCB included, to give you an impression of the simplicity of this design. It also uses remote sensing, to compensate for voltage drop over the Supply cables. However, constant current would fit our needs better, if used in an electrolysis setup. Regulation down to about 0.7 Volts would not work at 40 Amps in this setup, for too much heat would then be dissipated via the Output Transistors. Lowering the input Voltage via a variable Transformer Junction Tap, the use of an Variac Transformer pre-regulation (Tricky Concept), Triac or FET pre-regulation would work out. The 723 Chip would then need a separate small Power Supply Transformer to feed it. You still won't get the High Efficiency of a SMP, but you will have a virtuously indestructible Lab Power Supply which can also be used for other purposes, even to Charge your Lead Batteries with.

Carpark North - Shall we be Grateful (HD):
http://www.youtube.com/watch?v=-R1RQniPohI

Ladyhawke - My Delerium (Fan Death Remix):
http://www.youtube.com/watch?v=XlNIlnC6Hsk

Veronica Maggio - Måndagsbarn:
http://www.youtube.com/watch?v=NtJxy0zc7nU

Lambda

quicksilver - 20-2-2010 at 12:44

That looks like a nice design. In fact, the collection of designs in that site are quite nicely laid out.
However "It is a substantial project that is not for the novice constructor." That gentleman was not kidding. But then most of the materials or not too tough to get and the nice thing is how he laid out the PCB.

[Edited on 20-2-2010 by quicksilver]

quicksilver - 20-2-2010 at 13:15

It hit me at the same time you typed it. Thank you, however.

Lambda - 20-2-2010 at 14:01

@quicksilver, you had already Edited your Post, but taking the Power Supplies Efficiency into account, @entropy51 has given the correct answer:

If you go from 240 Volt to 120 Volt Primary Supply Voltage, then the current is doubled for:

P = I x V

The Power (P) uptake remains the same, and because only the Voltage (V) changes, the Current (I) will vary accordingly to match up to the same requested Power consumption. In this case, Fuses used will then have twice the Current rating on the Primary. Nothing has to change on the secondary though. Going from 50 Hz to 60 Hz won't make much difference, except that the Bridge capacitors may Theoretically be slightly smaller due to the Higher frequency. But, because these capacitors are often chosen by a rule of thumb, it really isn't a big deal here at all. What I like about these Analog Designs, based on a 723 Chip (or what ever), is that they are far less tricky to design and build for the layman in Electronics, and above all, Indestructible when designed properly. SMP are far to tricky to build, and when components explode, then the mains Voltage Power Supply is often involved, pumping Hundreds of Amps through the(se) Component(s). These explosions can be so violent, that fragments can literary fly right through the skull (read penetrate) of an unfortunate builder. Using a Transformer, also gives far more protection against electrocution while assembling and measuring on these Projects, and less isolation hassle and problems too. If you are not an old grunt in electronics (and they also get electrocuted and killed), then it's best to stay far away from SMP's. Personally, for reasons of safety and reliability, I think we should focus on a Robust Analog Design like the one presented in my previous Post. They have a much higher rate of reproducibility, component availability, safety, and in the end, ... Satisfaction !!

Jazmine Sullivan - Bust Your Windows:
http://www.youtube.com/watch?v=RqQBz3BzHLM

Lambda

12AX7 - 20-2-2010 at 15:06

Quote: Originally posted by Lambda

What I like about these Analog Designs, based on a 723 Chip (or what ever), is that they are far less tricky to design and build for the layman in Electronics, and above all, Indestructible when designed properly. SMP are far to tricky to build, and when components explode, then the mains Voltage Power Supply is often involved, pumping Hundreds of Amps through the(se) Component(s). These explosions can be so violent, that fragments can literary fly right through the skull (read penetrate) of an unfortunate builder.

Wow, I would be very impressed if electronic shrapnel could penetrate bone.

Fortunately, it can not.

In fact, because your favored analog circuit stores several orders of magnitude more energy than a switching supply, peak currents of thousands of amperes will flow when a device fails. And in a circuit like that, a device can fail quite easily, taking much of the circuit with it.

Switching power supply circuits are getting simpler. All the time. They are more important than ever, and a great incentive for the novice to learn a new technology.

Tim

[Edited on 2-20-2010 by 12AX7]

Lambda - 21-2-2010 at 06:56

Power Supply Design

Power Supply Safety - By Jerrold Foutz (16 January 2005):
http://www.smpstech.com/mtblog/power_supply_safety.html

These Components are all connected directly to the Mains Power Supply:

And many Novice Builders will try and attempt to Fabricate a similar Design, mistakes made by these Components D21-D24, C5-C6, Q1-Q2, D1-D2, can all lead to Violent Explosions. I have known people to have had pieces of exploding components logged in there skull. A member here has known somebody to have had a Tantalum Capacitor blown through his hand, and many people have lost there Eyes by similar incidents. Even more people have been electrocuted, and even Killed by working on similar designs. I often feel very uncomfortable, when working on new Designs of SMP's, but then thats Me, and You are You !

But on the other hand, this still dose not prevent me from working on New SMP's Designs, and for those of You who are also interested, including Me, here You go ...

Curiosity Killed the Cat !

Build a 13.8V, 40A Switching Power Supply:
http://ludens.cl/Electron/PS40/PS40.html

13.8V, 40A Switching Power Supply Schematic:

Depeche Mode - Wrong (Live at Echo Awards, Berlin) (HQ HD):
http://www.youtube.com/watch?v=L2GaCnAiuvo

Lambda

woelen - 21-2-2010 at 07:06

Every kind of power supply has its own merits. I do like switched power supplies. They are really compact, are safe to operate and are robust with respect to short circuit and overloading. That's why I made such a lab powersupply from an ATX power supply.

There is one exception where switched power supplies should not be used and that is when your power supply is used to power high voltage devices. I killed one ATX power suppy by connecting a 12 V --> 15 kV converter to it. Suddenly, while I disconnected the 15 kV device from the 12 V wires, the power supply quit. Most likely one of the SMD CMOS devices was killed by means of an electrostatic discharge. So, since then I use a transformer-based simple power suppy with a 25000 uF filter capacitor for high voltage experiments and I use the ATX PSU for electrolysis purposes (it is much lighter and more convenient to handle).

Lambda - 21-2-2010 at 10:33

Once things are working, SMP's are Modern Marvels, Efficient, Light, and Compact. But we are (at least I am) referring to Building one from scratch, and with all the associated dangers involved for the Novice Builder. And I don't mean Converting an existing SMP. But on the other hand, maybe this would indeed be the way to go, for Computer SMP's are now available up to, and above 1000 Watts, and they are dead cheap too. Maybe we should focus on Converting an existing SMP, rewinding the output transformer (maybe not even), and introducing Constant Current Regulation into an existing design.

Ladyhawke - Paris Is Burning:
http://www.youtube.com/watch?v=F1HDZNR9cY4

Lambda

quicksilver - 21-2-2010 at 12:00

I went through all sorts of hassle to find a good deal on a PSU that pumps 30A .....then I find that for $100 I could have gotten a 1000W computer PSU that easily could duplicate that:

http://www.tigerdirect.com/applications/searchtools/item-det...

There are a lot of "you fix it" Lamda and Sorensen professional PS floating around that a novice like myself can occasionally fix. But this 1000w stuff is good to go out of the box. The idea of getting a female main board plug and doing the work outside the PSU is attractive because then you still have your nice 1000w supply if you need it.

What's more if you like sparks and arcs a high frequency supply allows a simpler driver for things like a flyback transformer for a Tesla coil

Contrabasso - 21-2-2010 at 14:18

While agreeing that SMPS is the way to go for good DC and good voltage stability and with access to current monitoring as well as great low weight and high efficiency. Getting the transformer exactly right for a one off production isn't an easy matter. A simple mains frequency transformer rectifier unit will get volts on tap limited only by the availability of components.

Lambda - 21-2-2010 at 17:39

Quote: Originally posted by quicksilver

http://www.tigerdirect.com/applications/searchtools/item-det...
... The idea of getting a female main board plug and doing the work outside the PSU is attractive because then you still have your nice 1000w supply if you need it.

And on the same WebSite for ~ $50 more:

Ultra X3 ULT40311 1000-Watt Power Supply for $159.99:
http://www.tigerdirect.com/applications/searchtools/item-det...

Specs:

Input Voltage: 115 ~ 230
+3.3V: 2 4 A
+5V: 28 A
12V1: 70 A
-12V: 0.8 A
5VSB: 3 A

It's Modular, so you can chose the desired Voltage, and then plug in your own cables without Modifying anything on the PSU itself. The Connectors used on this PSU, are no big deal in respect to availability. You can also sacrifice and use one or more of the Cables that come with the package, for there are usually more (extra) than you need for the average Computer. And Split Junction Cables are also plentiful available, and very cheap too.

Quote: Originally posted by Contrabasso

... A simple mains frequency transformer rectifier unit will get volts on tap limited only by the availability of components.

Here you have an Ultra Simple Design, with overly abundant Components !

13.8V 20A Linear Power Supply:
http://ludens.cl/Electron/Ps20/Ps20.html

13.8V 20A Linear Power Supply Schematic:

Yello - Till Tomorrow:
http://www.youtube.com/watch?v=QQducSCD4YM

Lambda

densest - 22-2-2010 at 12:09

I've got a couple of questions: one is electrochemical, one for the users of these power supplies.

(a) Does the voltage waveform matter? A simple rectified sine wave goes from zero to 1.4 x RMS AC voltage and back at 2 X line frequency, so the reaction(s) start and stop 100-120 times a second. I don't know if there are chemical reasons this would be bad or good in any particular application. There could easily be time efficiency or cell maintenance implications. For instance, if the peak voltage is 5V and the reaction only starts at 4V, the reaction only runs 40% of the time. If there's a current threshold above which electrodes are damaged, the peaks could cause either unexpected damage or require running a cell at very low time efficiency.

(b) A lot of maximum voltages and currents have been mentioned in this discussion. The range is enough that one design can't cover the whole range while remaining efficient, inexpensive, easy to build (if necessary), easy to apply, and (relatively) safe. Are there sets of applications which require different supply capabilities? I'll throw out some guesses:

A single cell needs current:
test tube size - 0 - 0.5A
bench size - 0-5A
prototype size - 3-30A
lone scientist pilot plant - 10-100A
truly mad scientist basement factory - 25-400A :cool:

truly mad scientist downs cells - 200-2000A

A single cell needs voltage:
precious metals, delicate reactions - 0.5 - 5V
serious oxidations - 5-9V
wide electrode spacings or high resistance cells - 7-24V
are there any cells needing more than 24V? At that voltage, even a small current can boil things.

Series connected cells multiply the voltage.
Parallel connected cells multiply the current, but are unwise unless current is monitored per cell.

I'm asking this question in order to understand exactly how people want to use these supplies. All of the major design classes of power supplies have been mentioned. Each have their good and bad points. Choosing the best one for electrolytic service in an amateur lab has unique challenges.

And a suggestion for current regulation: In the "bad old days", people put incandescent lights in series with their cells and changed the wattage depending on how much current they needed. Such a bulb acts as a crude constant current source. The tungsten filament increases in resistance as it heats so that tends to limit current. A big advantage was that the bulbs are already designed to run white hot so hot resistors are not a problem. Put a lamp socket or parallel some sockets in series with your supply (automobile headlights run at 5-8A each, 100W 120V is .8A or so, etc.)

Voltage limiting: mechanical relays can act as voltage sensors.

overvoltage.gif - 3kB

quicksilver - 22-2-2010 at 12:56

THAT is (IMO) worth buying if you have "mad money" because you can always use it for a SERIOUS computer.
If a guy [put out some $ for some quality anodes, etc - 70a, that's a talking a real 5 gallon bucket cell! But you KNOW that in about 6 months it will be $98 !!!

@ Densest:

I do believe that waveform matters to a degree (if you are talking about what is immediately going into the cell) because what will pass to the anode will be altered in it's usage. More informed individuals will provide more in- depth material. I am not a 'scope user, so I imagine that you get what you get. Either the wave form plays well with what it's going into or not.
To use the above relay oriented setup may only result in a short from your mains (if that's your supply). If your supply is is separate, & NOT switchable my belief is that it will overheat.
I have a switchable a Sorensen,a Lamda & computer PSU. When I use a straight Radio Shack or home made I am fighting a heat battle that really needs the switching transistors (TO-3's) and several resistors "beefed up"; able to handle much more than the rated current or I will harm the supply. That's why one often sees home-made supplies with towering heat sinks and fans. I would bet that if you peeked at the transistors, resistors and rectifier, they would be rated much higher than the stock wattage, etc. This is the difference between a $19 12v 3A from Radio Shack & a switchable rack mounted supply from Raython for $400. This is why the computer PSU is a FANTASTIC thing and the alterations - a true gift to hobbyists.

[Edited on 22-2-2010 by quicksilver]

woelen - 22-2-2010 at 13:47

The wave form does matter in electrolysis. But just adding a filter capacitor of sufficient capacity solves the problem without reducing efficiency very much. When currents of tens of amperes are needed, then you need a big capacitor, e.g. 100000 uF. But these do not need to have large physical size, because the rated voltage for this capacitor does not need to be high. A good thing is to buy multiple capacitors, e.g. 5 pieces of 22000 uF/25 volts. These are not really expensive and using a parallel circuit of 5 of these allows big currents to flow through the total capacitor.

Lambda - 22-2-2010 at 15:14

Power Supply Design eBook Collection:

Power Supply Cookbook (EDN Series for Design Engineers) - 2nd Edition - By Marty Brown (Newnes & Butterworth-Heinemann - 2001) 278s - Including Bookmarks.pdf

Switching Power Supply Design - 3rd Edition - By Abraham I. Pressman, Keith Billings & Taylor Morey (McGraw-Hill - 2009) 880s - Including Bookmarks.pdf

Practical Switching Power Supply Design - By Martin C. Brown (Academic Press - 1990) 258s - Including Bookmarks.pdf

Regulated Power Supplies - 4th Edition - By Irving M. Gottlieb (TAB Books - 1992) 476s (d).pdf

PCB Layout Guidelines for Power Supply Engineers.rar

as

Power Supply.rar

iFile.it Download Link (34.13 MB):
http://ifile.it/fpmaz2u

No Password Required !

Note:

@quicksilver, you may find the following Excellent eBook of particular interest, for it's chockablock full of Schematics, and Theory (Included in the Download):

Regulated Power Supplies - 4th Edition - By Irving M. Gottlieb (TAB Books - 1992) 476s (d).pdf

Yello - Out of Dawn (HQ):
http://www.youtube.com/watch?v=zMoJWcCCLDI

Enjoy !

Lambda

dann2 - 22-2-2010 at 15:29

Hello,

I looked at some waveforms from crude supplies accross a Chlorate cell.

http://oxidizing.110mb.com/chlorate/raw.html

With no Caps. the current drops to zero between Voltage 'humps', with Caps. it stays flowing all the time with ripple. The exact implications of ripple would be hard to ascertain. Best to use lots of Caps. if you think it may have negative effect on what you are doing. With a Chlorate cell it does not matter a bit IMO.

Platinum Anodes do NOT like ripple at 50Hz or less at it wears them but that is not a problem for us as the worst we are goint to have is 100Hz ripple. If you were to start with a supply with a single diode doing the rectification when it would be a problem but no one would be doing that.

Dann2

Lambda - 22-2-2010 at 21:50

At a given Mains Frequency of 50 Hz or 60 Hz, once you know the Load Current, then the Filter Capacitor can be Calculated to give Voltage Waveform X.

DC Power Supply Design:
http://www.zen22142.zen.co.uk/Design/dcpsu.htm

But as @dann2 had remarked, if the exact implications of ripple would be hard to ascertain, then Calculating the Capacitor becomes a little problematic. Introducing a Coil between the Diode Bridge, and Capacitor(s), will make the Ripple even smaller. But then again, will this extra addition be of any benefit, unless used as an Adjustable Coil and Current Soak. Introducing an Adjustable Constant Current Soak into the system, would seem to me the way to go. Unless other Design Topologies are used.

Air - You make it Easy:
http://www.youtube.com/watch?v=0ulxUG8w2VU

Lambda

Contrabasso - 23-2-2010 at 00:31

Having worked in commercial electro plating, the power supplies we used were three phase transformer rectifier units off the mains. Three phase means that the ripple is less but as several electrodes passed 2 - 3 KA and the conductors looked like copper girders there was NO attempt at further smoothing. Smaller units ran off single phase with no smoothing.

For people who want good DC then a motor/car alternator system may give poly phase DC from an AC supply and if you can rework the regulator then all is possible.

Lambda - 23-2-2010 at 10:23

Introducing Soft-Start into our Power Supply !

“Soft-Start” is a device for Limiting High Short-Lasting Current in the moment of attaching the Transformer onto the Network. It is suggested for the Transformers on a Power range of 300 – 3000VA (Definitely >500VA) and with Primary Voltage of 230V (and also 110V). A device is to be installed in an appropriate box, or directly inside a device.

Soft-Start of about 05-1 Second is advisable for 300 – 3000VA Transformers. When Paralleling many MOT's for enhancing Serial Higher output Voltages at Very High Current (You can only do this if the Secondary is Rewound to a Low Voltage, and NOT in it's original configuration as used in a Microwave Oven). Advantages of Putting Transformers in Series are; Higher Output Power, and the ease of Introducing Variable Output Voltages WITHOUT having to Switch the Output Voltage (via Junction Taps) at Very High Currents. However, the Transformer(s) left Turned OFF, will then act as an Impedance Inductor (Coil), and Soak the Current (Limiting Maximum Current of the Transformer(s) Turned ON).

Soft-Start Circuit for Power Amps:
http://sound.westhost.com/project39.htm

Soft-Start Resistors and Relay Contacts:

Auxiliary Transformer Control Circuit:

Using Soft-Start Resistors in Series may be advisable with a primary voltage of 230V:

Soft-Start Circuit for Power Amps:
http://www.electronicsinfoline.com/Projects/Electronics/Audi...

Soft-Start Circuit for Power Amps Schematic (similar to the above Schematic, but now with a 12 Volt (D1) Voltage limiting Zener Diode):

Sarah Blasko - All I Want:
http://www.youtube.com/watch?v=QX1RgyCl1Xs

Lambda

quicksilver - 23-2-2010 at 13:02

@ Lambda: Thank you, that stuff is truly appreciated.

woelen's illustration of parallel capacitors is a great way to save money. It will certainly smooth out waveform and is commonly used in small drivers for getting sparks from flyback transformers which have to have high frequency smooth supply. One look at Mouser's catalog will show the monitary difference between a big old Coke can cap and several 20000uf. This stuff doesn't HAVE to cost big $.
One little issue I discovered is to not go chintzy on the traces or wire connections on a project that carries fairly strong current: make your connections thick!

Lambda - 23-2-2010 at 14:18

Quote: Originally posted by quicksilver

@ Lambda: Thank you, that stuff is truly appreciated.

@quicksilver, You are Welcome !

High Quality Siemens Capacitor; In Deutschland ist alles immer Besser ! :cool:

High Quality Mallory Capacitor:

High Quality Sprague Capacitor; I would Dump the Coke Can anyday for this Capacitor !

A Scrapyard is an Excellent place to find Large Capacitors etc., especially in Linear Power Supply Appliances:

Car Audio Shops Sell Hugh Capacitors of 1 Farad !!, and even Higher Values:

Remark:

Electrolytic Capacitors typically go bad in Older Equipment with age, and lack of use. The Aluminum Oxide Past tends to Crystallize, and instead of acting like a Capacitor, it acts like a Conductor, taking out Transformers and Rectifiers !!

Bruce Springsteen - Missing (Soundtrack from the Brilliant Movie; The Crossing Guard - 1995):
http://www.youtube.com/watch?v=CUOB_kopu_o

Lambda

Lambda - 25-2-2010 at 14:13

Pulse your Capacitor Bank by Opto-Isolated Zero Cross Triac Firing Transformer Pre-Regulation

This is an Idea that I have for a Pulse Charged Capacitor Power Supply with Constant Current Pre-Regulation, while minimalizing Power Loss in the Regulatory Network.

By keeping the voltage low over a Constant Current Source, you will enjoy the satisfaction of having enough room for good and stable Current Control Regulation, while at the same time minimalizing power loss in this Control Circuit. This may also be referred to as Transformer Pre-Regulation. By doing this at the Zero Crossing Point, High Turn-On Currents and Inductive Spikes are avoided. In Theory, the Transformer can be Fired 100 Times per Second at a Mains Frequency of 50Hz, and 120 Time per Second at a mains Frequency of 60Hz. In this way it's possible to Charge the capacitor Bank according to your needs of Power Uptake of the Electrolysis Cell. The Transformer will be Triac Driven via an Optically Isolated/Separated Zero Crossing Triac Driver Chip.

Application Notes:

AN-3003 - Applications of Random Phase Crossing Triac Drivers (Fairchild Semiconductor, Rev.4.07, 2006).pdf
Attachment: AN-3003 - Applications of Random Phase Crossing Triac Drivers (Fairchild Semiconductor, Rev.4.07, 2006).pdf (462kB)
This file has been downloaded 692 times

AN-3004 - Applications of Zero Voltage Crossing Optically Isolated Triac Drivers (Fairchild Semiconductor, Rev.4.00, 2002).pdf
Attachment: AN-3004 - Applications of Zero Voltage Crossing Optically Isolated Triac Drivers (Fairchild Semiconductor, Rev.4.00, 200 (368kB)
This file has been downloaded 775 times

Examples of Chips with a Zero Crossing Triac Driver Output with Optocoupler separation of the Driver Stage:

DataSheetCatalog.com:
http://www.datasheetcatalog.com/fairchildsemiconductor/258/

Part Name, Description, Manufacturer

MOC3162-M 6-Pin DIP Zero-Cross Optoisolators Triac Driver Output - Fairchild Semiconductor
MOC3162FM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162FR2M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162FR2VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162FVM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162SM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162SR2M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162SR2VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162SVM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162TM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162TVM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3162VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163 6-PIN DIP ZERO-CROSS PHOTOTRIAC DRIVER OPTOCOUPLER - Fairchild Semiconductor
MOC3163 6-PIN DIP ZERO-CROSS PHOTOTRIAC DRIVER OPTOCOUPLER - Fairchild Semiconductor
MOC3163-M 6-Pin DIP Zero-Cross Optoisolators Triac Driver Output - Fairchild Semiconductor
MOC3163FM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163FR2M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163FR2VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163FVM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler Fairchild Semiconductor
MOC3163SM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163SR2M 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163SR2VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163SVM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163TM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor
MOC3163VM 6-Pin 600V Zero Crossing Triac Driver Output Optocoupler - Fairchild Semiconductor

By choosing a region of about 5 Volts (maybe less if Current Fluctuation is not too High, and the capacitor bank has sufficient Farad Capacity) above the Voltage for the Current Source to operate stable in, witch is constantly measured by an OpAmp (as Differential Amplifier), and fed to a VCO (Voltage Controlled Oscillator) witch then fires the Triac. The smaller the Voltage becomes in this predetermined region of 5 Volts, the faster the Triac is Fired a the Zero Crossing Point to charge up the Capacitor Bank. And the Higher the Voltage becomes above this predetermined 5 Volt point, then the Firing rate is slowed down accordingly to counteract too high a deviation.

Using Zero Crossing Solid State Relays is an other alternative. I have managed to find about 50 of these Solid State Relays ranging between 25 and 100 amps in a Scrap Yard. They have a Trigger Voltage of 3-32 Volts at a few Milli Amps.

Solid State ReIays SSR.pdf
Attachment: Solid State ReIays SSR.pdf (106kB)
This file has been downloaded 748 times

Solid State Relays:
http://www.elco-italy.com/eng/index.php?option=com_content&a...

Triac Solid State Relays:
http://www.elco-italy.com/eng/index.php?option=com_content&a...

Portishead - Glory Box:
http://www.youtube.com/watch?v=GxsopQLZpCI
http://www.youtube.com/watch?v=JQyUdMdLZ18

Lambda

12AX7 - 26-2-2010 at 00:36

Ewww, phase control. I thought that died in the 70s.

Tim

quicksilver - 26-2-2010 at 08:28

Mouser has SS relays - they are NOT cheap. A BIG triac is often found in Microwave ovens.

Lambda - 26-2-2010 at 11:32

@quicksilver,

The only Mouser that I would like to have is a Gun !

But if You buy an SSR from Mouser, then a Wheelbarrow of Cash is required !

But on eBay ...

SSR 40A 24~250VAC Input 5~32VDC NEW for €13.50 from the Netherlands, but NOT Me:

SSR 25A SOLID STATE RELAY. OUTPUT OPERATING 24~380V AC (NEW) for £9.40 or €10.60 from the UK, but NOT my Brother:

CRYDOM D2440 SOLID STATE RELAY NEW for $12.99 from the United States, but also NOT my other Brother:

SOLID STATE RELAY, OPTO22, 240D25, 25AMP, 3-32VDC CONTR for£4.99 from the UK, my Sister is also NOT Involved:

If You are also Manic Depressive like I am, and hesitant to Jump, then David Sylvian will shore as hell push You right over the Cliff in "Flying" Colors ...

David Sylvian -- Nostalgia:
http://www.youtube.com/watch?v=CBupS5PMMsc

Lambda (I am still alive, ... but barely ...

)

quicksilver - 28-2-2010 at 13:43

I attempted an experiment wherein I used a 2 L cell with MMO anode Ti cathode and super saturated solution of Cl @70C. I started the cell at a level of current much higher than I normally would at 30A. The salt did NOT appear to settle out. It stayed at 60C for approx 2 days and at that point I backed off to 15A as the level of Cl was quite high (both from colour and gas venting. A small bit got on a shirt and bleached it SO fast and viciously, it became white and the threads become near to dropping apart. Super Bleach!

I allowed this solution to maintain for 8 days. At the end of which I emptied the cell & recrystallized the material within. It was KCLO3 & appeared to be pure by both crystal structure, tests from colour, H2SO4 and the making of a few differing coloured stars. Additionally the "window-pane crystals fell out at an increasing rate due to taking the heated solution and placing it in a refrigerated environment; once the bottom of the flask was covered, the formation was extremely fast. I only wish I has a decent movie camera because the formation was unique in speed.
I was very surprised that a percentage of that was not plain salt. but the level of current was very high from such a small cell and the speed in which current was applied was immediate.

Upon completion I attempted this at a lower level of current (3A) and the cathode (perforated type) would clog with material and the original salt DID remain at a level of approx 50gr. It appear that a super saturated solution CAN be utilized IF the level of current is high enough and applied quickly enough prior to the hot solution to cool down and higher levels of potassium chloride to drop out. With 70-80C water would digest 420 gr, cooled down 330 would remain in solution.

I did finally find a scrapyard with electronics and was overjoyed to find that I now can buy some things without working with eBay. And you all are absolutely correct as the above ss relays did sell fo about $9USA. The best thing is to go when the proprietor is at lunch as I got another powerfully switchable (this time a company called QSI that was bought out by Lamda) that pumped 150A. So I am busy setting up a 5 gallon cell and seeing what can be done with graphite gouging rods. I don't want to risk my MMO on current of that level.
IF anyone wants I have the address and phone. They may ship overseas. They do have a a small stack of rack mounted PSU (Lamda's mostly) but they are, your luck. I got one that needed a new filter cap (that they had) for $30! plus a new cap at $15. PM me if you want details.

[Edited on 1-3-2010 by quicksilver]

Lambda - 1-3-2010 at 18:06

Quote: Originally posted by quicksilver

... I did finally find a scrapyard with electronics and was overjoyed to find that I now can buy some things without working with eBay. And you all are absolutely correct as the above ss relays did sell fo about $9USA. The best thing is to go when the proprietor is at lunch as I got another powerfully switchable (this time a company called QSI that was bought out by Lamda) that pumped 150A. ...

@quicksilver,

I am pleased to read about you successes at the Scrapyard !

Working with eBay can be a Pain in the Rear, for not only are You dependent on the Reliability of the Dealer, but the Guys working for the Postal Services in My Country Holland, Steal like Raves !! And the Postal Services refuse to refund Stolen Goods, for they don't employ Thieves. I have lost a substantial $ and € amount of Goods in this way :mad:

And when dealing with the Scrapyard, by leaving the Proprietor out of the Equation, indeed can lead to a substantial reduction in Price. A Crate of Beer also works pretty well too

Taking a Beautiful Girlfriend along will also Impress them !

A LAMBDA Power Supply is indeed a Top Notch Modern Marvel, ... how could anyone expect anything less ? And where have I heard that Wonderful Name before ?

More Info on Power Supplies and Triacs:

DC Power Supply Handbook

Introduction:

This handbook is designed to aid that understanding by providing complete information on the operation, performance, and connection of regulated power supplies. The handbook is divided into six main sections: Definitions, Principles of Operation, AC and Load Connections, Remote Programming, Output Voltage and Current Ratings, and Performance Measurements. Each section contains answers to many of the questions commonly asked by users, like:

- What is meant by auto-parallel operation?
- What are the advantages and disadvantages of switching regulated supplies?
- When should remote sensing at the load be used?
- How can ground loops in multiple loads be avoided?
- What factors affect programming speed?
- What are the techniques for measuring power supply performance?

In summary, this is a book written not for the theorist, but for the user attempting to solve both traditional and unusual application problems with regulated power supplies.

DC Power Supply Handbook (Agilent Technologies AN 90B - 1978, Rev. 2000) 126s.pdf
Attachment: DC Power Supply Handbook (Agilent Technologies AN 90B - 1978, Rev. 2000) 126s.pdf (1.1MB)
This file has been downloaded 824 times

Triac Control by Pulse Transformer

Introduction:

Among the many ways to drive a triac the pulse transformer is one of the easiest. By applying some simple rules it can be used to design an efficient triac triggering circuit without reduction of the commutation capability of the triac.

Why use a Pulse Transformer?

The use of pulse transformers in triac triggering circuits offers many advantages:

- Galvanic insulation between the power and gate drive circuit (a few kV).
- Gate drive circuit with a few components.
- Choice of the gate current polarity (triggering in the 2nd and 3rd quadrants for SNUBBERLESS triacs).
- Optimization of gate signal (single pulse or train of pulses).
- Possibility to drive several triacs with only one drive circuit

Triac Control by Pulse Transformer (STMicroelectronics AN436 - Rev.D2A-3575 - April 2004) 9s.pdf
Attachment: Triac Control by Pulse Transformer (STMicroelectronics AN436 - Rev.D2A-3575 - April 2004) 9s.pdf (66kB)
This file has been downloaded 673 times

Thyristor Theory and Design Considerations Handbook

Thyristor Theory and Design Considerations Handbook (ON Semiconductor - HBD855-D - Rev.1 - November 2006) 240s.pdf

iFile.it Download Link (2.48 MB):
http://ifile.it/zmie1la

No Password Required !

Pink Floyd - Money ("Pulse" Tour):
http://www.youtube.com/watch?v=6RqwLWR42nE&feature=relat...

Enjoy !

Lambda

quicksilver - 2-3-2010 at 10:23

Thanks for posting those!

The only place I found pulse transformers that were sold a a decent price was:

HTTP://amazing1.com

I DID find a Neon sign shop that was going out of business awhile back when I was making Tesla coils and got about 300lbs of HV transformers. Here we have a "club" of people that experiment with high voltage stuff but it's in the city and I have to drive about an hour to get there. but many of these guys have their doctorate in electronic engineering and have made stuff I can only dream of putting together.

I also trip into this forum, which has a wealth of info.

http://4hv.org/news.php

[Edited on 2-3-2010 by quicksilver]

densest - 2-3-2010 at 15:37

The amazing1 transformers are not what you want. You want "gate drive" or "SCR/thyristor trigger" transformers with a turns ratio of 1:1. "Photoflash" trigger transformers have a high turns ratio and deliver low current at very high voltage.

This is one case where it is usually cheaper, easier, and safer to get new parts from a distributor. You can get guaranteed 1500VRMS isolation for $4 or less and guaranteed characteristics.
For large, heavy, relatively low tech items like big iron transformers surplus is definitely the way to go, but this isn't the part to go "unknown specs" on. If you don't get it right you will get random flickery outputs and/or shorted driver circuitry, and for these pulse transformers new parts are cheaper than any surplus supplier I've found!

The biggest problem with buying these from anyone is that they're rarely classified correctly in online shopping sites. You almost have to start at a manufacturer (google gate OR scr OR thyristor transformers) and find their distributors.

One part that looks good is Pulse Engineering P0584NL. It is $4 each from www.mouser.com. Murata has a line, etc. Newark/Farnell also stocks some pulse/gate drive transformers but their search engine is pretty useless. If anyone wants some help picking the right transformer send a PM.

I know people have been trashing Mouser as a supplier. For small items like resistors, capacitors, transistors, etc. they are often the cheapest distributor - I've bought 100s of parts there after comparison shopping. They will ship any way you choose and are quick - usually 1 day from order to shipping. They also gave me a credit account

You can wind your own if you don't care about safety or you're willing to put 2 layers of transformer insulating tape and 10mm spacings between wires, etc... a small toroid (1cm) and 2 x 10 turns of #30 wire should do it. Don't use RFI/interference suppressing material because that would kill the pulse! There's an article on how to design one http://powerelectronics.com/passive_components_packaging_int... for the masochistic.

For a 25A SCR a pulse 3 microseconds long at 5V should trigger it reliably as long as the SCR is above 0C. It might take 4-6us if it's -40C. A huge (300A) SCR will take 7V for 30microseconds.

Using an excessively long pulse or high drive voltage will not help. The transformer will "saturate" i.e. cannot hold any more energy and suddenly look like a short circuit to the driving electronics and its output will drop to zero. Pulse transformers are rated in "volt-seconds" or "volt-microseconds": the voltage applied times the length of the pulse. Small transformers cost about $0.10 per volt-microsecond, so a $3 or $4 unit should be plenty if the drive circuit provides a short (2-4 microsecond) , high current (100-200mA) pulse at 5V.

[Edited on 2-3-2010 by densest]

quicksilver - 2-3-2010 at 15:57

NO, I know what you're saying. Mouser IS fine for some stuff. I just wouldn't buy their variacs :-)

Thanks for the tip!

watson.fawkes - 2-3-2010 at 19:37

Quote: Originally posted by densest

Masochistic? That article's well written. The comment on bifilar/trifilar winding is particularly useful, as it would tend to lower the maximum fields in the core somewhat.

I'm confused about what you said, though. How do you get 10mm spacing on a 10mm (1cm) toroid? I suspect typo. Can't you get adequate isolation between primary and secondary with a high-quality insulation on the wires? If you need 10 kV isolation, I suppose it gets easier to use an air gap, but where's the trade-off point?

densest - 2-3-2010 at 23:11

@watson.fawkes - yes, I confused "safety" with "basic" insulation requirements: I think 1.6mm is enough for 1500VRMS - reasonably safe if the triac/SCRs are directly connected to the mains.

Bifilar/trifilar windings are really useful in many cases. If you want to isolate control circuits from nasty mains voltages or switching spikes, the increased interwinding capacitance has to be taken into consideration. Enough could leak backwards to upset CMOS logic or microcontrollers, etc. You always have to trade isolation for parasitic capacitance and efficiency

Personal safety using the equipment can usually be ensured with a GFCI in the mains.

Keeping the logic happy is important. I learned many years ago that while most ICs have good protection on their inputs, they are quite vulnerable to high energy transients on their outputs which far too often exit through the ICs inputs to the previous stages. :mad:

So I worry a lot about the design where the 3V/5V logic meets the power line.

I have a piece of a power line inverter (probably from something like a 50-100 KV DC system) comprised of stacked high voltage SCRs. The trigger transformer is unique. Each SCR has a secondary wound on some toroids spaced for high voltage. The primary is a straight wire in a thick insulating tube passed through the center of the toroids. Lousy coupling, but at 20KV/mm, the system is set up for 100KV. I use the assembly as a hat rack. :cool:

12AX7 - 2-3-2010 at 23:56

Cute.

SCRs are an awful lot harder to handle than MOSFETs (or IGBTs). Lossy and slow, you can get a lot more power and efficiency from FETs, with far smaller transformers because the operating frequency is higher. The same voltage and fault conditions exist (less, actually, since MOSFETs don't stay on once you trigger them).

Dollars per weber is a curious measure for a pulse transformer. I have some which should be very expensive indeed, then; 2000 turns will easily take several webers (not uWb!). Seems to me, a more useful measure is wattage, since the current that a given winding can be manufactured for depends on the available winding window, which is a matter of core size.

Tim

densest - 3-3-2010 at 05:11

@12AX7 - quite true, SCRs and triacs don't want to move in a hurry, have relatively high "on" voltage, and require substantial trigger power. They're still made in huge quantities because they're cheap, cheap, cheap, rugged, and are fast enough to use in dimmers, small variable speed motors, etc. For those applications, no transformers are used at all - all isolation is done with insulation or open space and the drive circuits are very very simple. Did I mention cheap? and easy to make for relatively high voltage operation and surge tolerant. All good for consumer items not needing to be superbly efficient. Cheap. And easily made on old semiconductor manufacturing equipment in a basement somewhere.

I was using $/wb because (simplified probably too far) a trigger/gate transformer is normally used to drive a moderately high impedance (mostly capacitive) with a defined waveform and power isn't a limiting factor. Yes, driving a FET/IGBT can require multi-ampere spikes at turn-on and turn-off, but in between the current approaches zero. So the required performance is often limited by how long the drive can be applied before saturation and the charge/discharge current, not by power transferred. And most manufacturers specify volt-microseconds, inductance, winding resistance, and turns ratio as the primary characteristics of the devices. A matter of convenience, really.

quicksilver - 15-3-2010 at 16:40

I have been have OUTSTANDING luck with Power Supplies, High Voltage goodies and bits at this site (which I highly recommend):

http://www.surplussales.com/index.html

Best thing is to talk w/ them on the phone. Their "hook" is free shipping after $100 or equivalent in Euros. They have a GREAT deal of Lamda stuff, - they seem like they enjoy "dealing"; so many prices may come down a bit.
Los Alamos stuff, & military gear. Some of the high amperage SWITCHING supplies are at $30-40 for a 5Vdc / 30A and great savings on HIGH VOLTAGE (hard to find) components.... They also have damn good Rack mounted supplies. Some are tested - others are "as is" but all are priced accordingly.

PS - an "as is" was simply a complex multi-switching supply (5vDC @ 150a) that had no manual and in a few hours I learned to set it up (the trim pots are wax sealed from the lab to match the material printed on the component.
That one was perfect for BIG tanks. Today I cleaned a 5 gal bucket cell which netted me 6 lbs dry KCLO3 with graphite rods and 10days but used perhaps 60a. If I wanted to deal with the graphite - I think that supply could handle a horse water feeder at 200 gal or a serious high volume plating.

[Edited on 16-3-2010 by quicksilver]

Lambda - 15-3-2010 at 18:04

Quote: Originally posted by quicksilver

... PS - an "as is" was simply a complex multi-switching supply (5vDC @ 150a) that had no manual and in a few hours I learned to set it up (the trim pots are wax sealed from the lab to match the material printed on the component. ...

@quicksilver,

You may be referring to this Power Supply:

Lambda Power Supply - Model (PS) LGS-EEA-5-OVR

Lambda regulated power supply. Over voltage protection. Original price = $1,176.00.

• Input: 105 - 132 vac @ 47 - 440 Hz
• Output: 5v @ 150a @ 40ºC reg. 0.1%, 96a @ 71ºC
• Dimensions: 4-15/16" x 7-1/2" x 16-1/2"
• $795 each

But, ... have You seen this One on the same Web Page which can Pump a Massive 300 Amps at 5 Volts !!

It does have a different input Voltage of 230 VAC, instead of 105 - 132 VAC mentioned on the previous Model.
http://www.surplussales.com/PowerSupplies/PowerS-3.html

Jeta Power Supply - Model (PS) CP137

Jeta Model CP137 power supply. No manual, no cords. Not rack mountable. Mfg. P/N: 725-3308-1.

• Input: 230 vac
• Outputs: 1850 watts, V1: 5v @ 300 amps, V2: 12v @ 20 amps V3: 5v @ 10 amps, V4: 12v @ 5 amps
• Dimensions: 12" x 18-1/2" x 5"H
• $125 each

Fine Young Cannibals - The Flame (VHS Remaster):
http://www.youtube.com/watch?v=GHYdaorAKY4

Lambda

quicksilver - 16-3-2010 at 08:51

I don't have a phone camera with me or I'd post it. There's a "QSI" (sold to Lamda) w/ fans that 150a @ $40 It is a "as is". no manual, not material at all. It was strictly a gamble. But I'LL do that;. I really know little about electronics but I worked as an electrician through the construction industry and have a very effective way of problem solving.
I got it going and SLOWLY figured out all the idiosyncrasies. $40 for a 150amp switching supply that is MADE for 24/7 use! I was stoked. They also have quite a few excellent 5Vdc 15a NEW supplies that simply need an enclosure, some are 5v and much higher. An enclosure is not only easy but you can put more than one fan to create a "flow-through", that's very effective!

On another (somewhat off topic) subject I have a warning for some of you who use Gouging Rods for anodes/cathodes. IF you have prepared your rods carefully with dilute acids to knock out the remains of copper and iron you may sill get a batch that is off color(beige). That material should unquestionably be re-crystallized. Unless your batch is white as snow; re-crystallize. With this you take out an insurance policy. An untreated rod will not last very long where a well treated rod may last more than 2 weeks! But introduced into the brine might be a fuel (especially if that same rod is used for a long time).
There is a possibility that some other material was left within the the chlorate (or perchlorate) crystals. this may constitute the same as material as a fuel, intimately mixed in the chlorate! Just like BP, the mix would be extremely intimate and the result could be very serious IF the material was exposed to any stimulus. Re-crystallized until you have a snow white crystal. Linseed oil, shellac, carbon, etc are all fuels! The mix would be very intimate similar to a PB mill. I strongly suggest to recrystallize, as ANY fuel with a chlorate is a serious problem. The off white color may be an indication of some type of fuel. It may also be an indication of a simple color change from filter paper or inks; but why take a chance? This issue generally takes place in a large cell but regardless, if you believe your resultant material may have something "extra" - do something about it.
If the material is off color when still wet, let it dry and see what takes place. A crystal can "throw off" a dye of sub micron carbon just as it can retain it. Take a very small sample and test. PURE chlorate with no fuel should not take fire in the same manner as one with a fuel. Friction & impact will differ also. If it acts quite sensitive - clean it further.

[Edited on 16-3-2010 by quicksilver]

dann2 - 16-3-2010 at 11:13

Hello Quicksilver,

Quote: Originally posted by quicksilver

.......... An untreated rod will not last very long where a well treated rod may last more than 2 weekks!........
[Edited on 16-3-2010 by quicksilver]

Are you using pH controll with these rods?

Dann2

quicksilver - 16-3-2010 at 12:19

I only used untreated graphite once (those were"bars" of graphite 2"x1/2"x10") and did control ph but they broke down pretty quickly. So I shut down and from then only used treated rods and with those -Yes; by that I mean I'll maintain 6.5-7 ph. I test at the beginning and at 2 day intervals. However I'm using a lot of current as compared to what I had previously used. Using the formula you described previously I was pretty low before. I don't have the numbers in front of me but I had used 3a with 2.5 L which should have been higher. I know that I'm low because with another cell with Ti/MMO mesh I would get formation within the perforation (of crystals). Kick it up a slight bit and not crystal formation occurs in the perforations.

I believe it helps the rod's longevity (stringent ph monitoring) but I'm not convinced that the rods are consistent in porosity or density. Opinion-wise I think that anytime enough current is moving through there to maintain 50+ C, the rods' going to "shed". I get enormous yields of chlorate with graphite and the cost is low but the mess and lack of snow-white crystals (even from a rolling boil to re-crystallize) makes me think that the extra work is the trade-off.
I'd be curious to try an experiment wherein a dye is placed in the tank prior to cultivation and see whether the crystal forms without it from a single recrystallization. I pulled one after 10 days and the crystals -appear- beige, but are still moist and I am going to dry these out as the water from graphite is SO stained that it may be I am looking at dyed water. However I have seen some that unquestionably had not been purified by a single recrystallization. Those that were stained within the crystal were hyper-sensitive.

edit 3/17/10
well I tried 3 re-crystallization and ALMOST got that yield water-white but it still had some beige in it. I believe that it may be from the linseed oil. I'm disappointed in it's lack of purity wherein the MMO/Ti material is about as clean as it gets from a single re-crystallization. Of course I'm going to keep it separate and do a heck of a lot more testing to determine if it's safe. However I am pretty disappointed that such a high yield may not be the ideal method. I'm going to try another but with a great deal of attention to ph AND I'm going t remove any outer layer of linseed oil even if it's not obvious (this was not obvious to a greater degree) but better safe then sorry.

[Edited on 17-3-2010 by quicksilver]