32 replies to this topic

[sasman]

Hi Is there any one who can help me with advice on making a CHEAP!! Low voltage 3v to 5 v 100 amp+ Dc PSU?.. i have a very basic knowledge of electrics.. at the moment i'm using several computer PSU's ..and to vary the AMPs i am using some resistance wire which glows red hot when passing 20 amps?...


Now i can use 5 or more PSU's to get my 100 amps but this is a bit of a bodge and would prefer a nice big 100 + amp PSU..Can any one help?..

I have read that you can use microwave transformers and rewind them to produce low voltage high amperage..I would also i'd like to do away with the resistance wire and use some sort of solid state device?...There is a powerful PSU on ebay but it going to cost nearly 200 quid and that only supplies 80 amps.

Also i have a large oil cooled AC electric welder that is 50v and 80v and 250+ amps..is there any way i can get the voltage down to 5 or 6 volt?..

Any advice?..or can someone point me in the right direction to webpage forums etc..

thanks

[alany]

Welders are generally constant current, the 50-80 V is open-circuit for starting the arc. Once it is established the voltage across it is only a few volts. If you use it to supply your (per)chlorate cell with a bit of caution it will probably work just fine.

Rewinding a MOT is quite practical. They are usually designed for 1 V/turn so 10-15 turns of *heavy* Copper or Aluminium busbar will do the trick, after you remove the old secondary. One word of warning though, there will be virtually no current limiting, a dead short across the output will pull hundreds of Amps and probably melt the primary, the core, and weld the short like a spot welder. Inductive or phase-angle control on the primary side would let you change the average current losslessly.

Your most expensive and difficult to get part would be the 100+ Amp rectifiers. For a continous 100 A supply you'll need at least 2 MOTs from a big microwave, 1.2 kW at least. Welding isn't 100% dutycycle and the transformer cores will get very hot in continous service.

This guy's site might be helpful:

http://www.dansworks...rc welder.shtml

If you aren't good with power electronics I seriously suggest you don't attempt building your own high-current supply. There are just so many ways to kill or mame yourself doing so that can't think of them all!

[sasman]

Thanks for that alany..

i have just purchased his ebook only cost a few $ and will have a look later..first glances looks very promising ..

Would it be possible to connect say for eg... 5 computer atx psu's each supplying 20 amps in parallel so as to produce 100 amps?.. i have a funny feeling doing that would end up with a big banG!

[alany]

Actually it depends on their construction. You'd have a better chance if they were all identical. Their regulation circuitry might fight each other a little, but it should work OK in parallel. The small resistance of the wiring from each supply will help equalize the voltages and share around the load in practice the same way the internal resistance of batteries does when they are paralleled.

The output side is DC isolated from the mains so in theory you could series them too, but what's the bet the -ve side is earthed to the case.

The welder with a variac (if its current control system isn't a variac already) might be the easiest supply to deal with. You might have to disable its HV striking circuit, if it has one. There are welders that are designed to double as car battery chargers with very little additional circuitry, just some voltage monitoring.

How much work have you done with smaller supplies? I am using a 25 A 3-15 V volt linear supply, but it isn't really suited. Still works though. When I scale up production I intend to either hack over a welder or parallel several chunky transformers.

[Andrew]

First of all what is the intended use?

Computer PSUs are a type of SMPS (Switch Mode Power Supply). As Alan said, you can connect them in series to obtain an addative potential difference. However SMPSs are not designed to opperate in parallel with each other. If you dont want to hear what happens skip the nesxt paragraph

In most cases SMPSs are neither matched significantly, or protceted enough. It is not good practice to do it, and often results a very big bang, and destroy at best, just the one that went bang. If they are regulated they will "Fight" against each other. Instead of then all providing their fair share of the current, One will provide it all until the next one starts pulling it's weight and so on. This all sounds very well, but there is a catch. The tollorances are enough to cause catastrophic failure. What happens is the four with a lower regulated outputs (PSU2, PSU3, PSU4 and PSU5 in order of highest regulated output first), sink all the current from the highest one (PSU1). PSU1's output voltage drops because it can't supply enough current. PSU2 now stops sinking current and starts to hold the voltage at it's regulated output. PSU's 3 4 & 5 sink all the current from 1 & 2. Now the voltage drops to the regulated output of PSU3 and PSU3 stops sinking current. Two things can now happen. The first is the happy ending, the output is that of PSU3. PSU3 stays cool and PSUs 1, 2, 4 and 5 get hot. When you require your 100 amps all the supplies source current and they all get hot, but some more than others. The second posibility is the one that actually happens. This is because there will rarely be PSUs that are that good. What happens is the perviously described trend carries on until PSU5 is trying to sink all the other sources. This is at least 80 Amps, more likely to be a couple hundred Amps. If big bang number one hasn't happened, it will now. Then PSU4 will try and sink the power from three souces (PSU1 PSU2 and PSU3). Big bang number two happens. A similar thing happens to PSU3 and you might even get big bang number three, but posibly just a little one. Then it's just PSU4 and PSU5. What happens here is pot luck, but usually it will smell bad and there is a real posibility of a fire starting. If you are very very very lucky you might just have one that still works afterwards. The above passage will usually happen within a second for your first three bangs and then the death of the best built PSUs will happen within 10 seconds. I even think that some computer PSUs say on the side "do not connect" to other PSUs.

Refitting a welder is a lot of work, and the ebay idea sounds out.

I would suggest buying a 500VA toroidal transformer and wind in a new secondary coil. The new secondary needs to be made from wire that is at least rated to carry 100amps, although you could use a lower rated wire and just ensure it does not get too hot. before you wind on your new secondary, unwind the original and count the number of turns. The secondary will be the thicker wires. Say that you count 100 turns, and the output was 50volts, to get just 5volts you will need only wind 10 turns. This output will be AC and you will then need to rectify this to get DC. You could do this with diodes or perhaps bridge rectifier modules. The transformer will cost around ?30 and the rectification can be done on the cheap as well. You could put in four (five to be safe) 25AMP bridge rectifiers with all the same tabs connected (parallel connection) and this would cost ?8. Alternatevely you could put in 100 1amp rectifier diodes for each part of the bridge costing ?4, this would also look very cool. All in all you could build a 5volt supply for ?40 which is less that five computer PSUs. You could also built a lower voltage version for even less, ie. a 1volt version could be built for ?20.

hope helps

[sasman]

Thanks for that info Andrew .. mentioning the 40 pound price tag made me smile!!...The transformer is going to be used for continous use 5 to 7 v at hopefully 100 + amps..it's for a electrolytic cell..

Just a few questions..Is it easy to rewind a new secondary?.. i have just looked on ebay and found a few torridal transformers...they look like big doughnuts...were is the secondary winding>?..
If i were to buy a 500VA would this produce 100 amps? without getting to hot?..i understand the bridge rectification bit . its just this transformer bit im not to sure of?
..Also what does 500Voltamps mean in a transformer? it cant mean 500 volts at 500 amps can it?..

thanks..

[ProfHawking]

as far as i know, volt amps or Va is a measure of the power it can supply. So 500Va would be 500volts at 1 amp, or 100a at 5 volts, as determined by P=IV
power=Current X Voltage
I have found some nice transformers in the past out of UPS (uninteruptable power supplies) for computers. They have many hundred watt capacities, serving normally 24V at 50-100 amps, depending on the capacity of the UPS. Have a look on ebay for UPS spares/repairs. the transformer part rarely fails, so you should be able to rewind them to give any configuration you require.

[Andrew]

Yes, the unit VA is another way of describing the power rating. You can pick up a 500VA transformer for ?27+VAT.

Toroidal transformers are wired in a what can seem confusing way. They almost always have two primary and two secondary coils. They are wound continuously around a donought shapes ferite. All the flying leads come out of the same place. The Primaries and Secondaries are easy to tell apart. The primary or mains leads are thinner than the secondary leads (, but only when it is a step down transformer, which they nearly always are). Once you have your two sets of 4 wires you need to pair them off. This is done with a continuity tester or resistance meter. Then you need to assess their polarity, (done with a low level sine wave signal and a few conbinations, or visual inspection), and then join the relivant wires. You should have the two primaries joined in series with the whole coil wound in the same direction, this will give you a 230V primary.

The secondaries are not really a problem because you need to either remove them or just ignore them. I would unwind them and count the number of truns on one of them, the other will have the same number of turns. Say that you have two 30v 8.3A secondaries. (In normal operation you would have them in paralell to give 30v upto 16.6Amps, or in series to give 60v upto 8.3Amps). Say that you counted 120 turns, your 5volt coil will only need 20 turns. Rewind your secondary using either a new peice/peices of wire rated to 100amps, or the old secondary coil. To use the old secondary wire wind 20 turns but with 12 paralell lengths. The good thing about the VA rating, is that if you use all the wire in the secondary, you know that it is rated to the original rating regardless of the output voltage.

You said it would be for an electrolytic cell, what exactly do you mean by this?
I'm sure that you don't need to be told that mains voltage is dangerous, also, If your unsure about anyting ask or read up about it before attempting it.

Hope this helps

[Loci]

Hehe, could have sold you my Tesla Coil sasman 1.5 million volts at very high amapges

[Arthur Brown]

The transformer from a microwave is a good start. Keep the mains side intact and box it to keep it dry but well ventilated --fan? remember that to generate 5v at 100a is 500va continuous allow some headroom and lots for circuit losses rectifiers drop 1.2v per hot diode (less for schotkys) and good crimped or soldered connections are vital as is adequate wire sizing 100a -- ca 25mm2 thats big and doesnt crimp or solder with hand tools.

Scale the secondary to give about 8 - 0 - 8 as a centre tapped winding use two diodes prob stud mount onto a fan cooled live heatsink say 0.5 deg C/watt or better ( smaller number). secondary winding likely to be in 15mm x 2mm copper bar with mylar tape insulation, you may need to dismantle a 1 or 1.5 KVA transformer to have the room to bend the secondary into place.

Control is typically by variac or switched high capacity resistors in the primary as is FUSING the transformer should be fused with a T class fuse scaled to allow the required power through so a T3.15A fuse in the primary should be ok but ensure that the plug has a 5 or 10 amp fuse also.

Ensure there is NO chance of the LV leads or the electrodes touching as the flying UV and molten metal will damage your face and sight.

Torque the bolts well and plan the current flow to be through copper not steel.

[Andrew]

Another very important point when using toriodal transformers, is that you can never under any circumstances have a "short" going through the toriod. The top of the fixing should never be conductively coupled to the bottom. If it is you will induce a massive current to the bolt and melt it, and cause a lot of grief, for the bolt, the transformer and for you if you are anywhere near.

As Arthur Brown quite rightly said, circuit loss also needs to be considered, if you used componants to their rated values and not exceed them, they will not get too hot. The transformers are designed to maintain a moderate temperature if they are not used above 100% duty. Should there arise any need for cooling, once done you will drop less than 0.7volts per diode, that's 1.4 volts if using a bridge. But the voltage drop is not your real problem, the heating loss over the diodes is. You would need 14Volts to drive 100amps through 100x 1amp diodes, this is because of the diodes forward resistance. A centre tap gives you the ability to achieve full rectification using half the number of diodes. But if you wanted to apease the heating loss situation, you should use a more active solution. A bridge of thryistors or triacs would help but a switching configuration of MOSFETs would do nicely, this would require a small amount of circuitry though.

[sasman]

The transformer will be used in a perchlorate/chlorate cell with the intention that it will be on 24/7 for months at a time..Voltage 4v to 7 v...I would use the 5 x 25 amp bridge rectifers.. ..Is there any easy way to vary the current through the cell ,
At the moment im using electric heater element resistance wire but this glows white hot!! which i dont like...

Also i have just found a 400 amp 3.3 v power supply it came out of a cray super computer... water cooled this would be perfect if the voltage was higher?.. is there any easy way i could up the voltage?...to say 5v or 7 v?,,,

thanks

[Andrew]

The answer to both your question is yes, To vary the current through the electrolyte, you can vary the distance between the electrodes, vary the surface area of the electrodes or impead the current resistively, i.e. a resistor. To step a voltage up is easy. It involves a pair of diodes and some capacitors, or in this case a few diodes. I'll message you a circuit diagram of the voltage steppper. the 3.3v would be doubled minus the Vf of the diodes. That is 6.6-(1.2 to 2.2) or 4.4 in the worst case. You can add another diode and capacitor to tripple the voltage minus three times the Vf of the diodes, or 6.6volts in the worst case, . I'd use 25 or 35 amp bridge rectifiers from rapid electronics, and a few of them at that, probably in an array, to minimise the forward resistance. This design will step the voltage up but reduce the current it is capable of producing. You would still get your 100amps though. Your problems might be keeping the diodes cool and the rate at which you would produce perchlorate. If you actually manage to drive 100A through the electrolyte it would deplete the starting materials very fast indeed, or at least reach equlibrium very fast.

[sasman]

Ok i have just bought the 400amp SMPS Power supply...here is some info

Specifications
Output voltage: 3.3V
Output current: 400A
Slew rate 200A/uS (80% step)
Step response <70mV for 320A step
Auxiliary output: 12V 0.6A
Input voltage: 200-240V, 50-60Hz, PFC
Input current 12A
Remote sensing
Remote enable
Water cooled base plate fitted
weight 8kg
If im lucky there appears to be some variable pots on top left hand corner its says 3.3 ADJ? this maybe a voltage adjuster?.. there are a few more adjusters but not sure what any of them do...

I have a friend who is into ham radio..and mentioned that you say you can up the voltage with diodes? he said that you can do it with AC but he wasnt to sure if you can up voltage of DC?..he would be intrested in seeing the circuit diagram?..

[Arthur Brown]

Voltage doubler circuits only work AC and that badly. If you use SMPS the control is low voltage low power. If you use transformers the control is at mains voltage and moderate current.

As for altering that smps you may be lucky. Test it for safety, then test the op volts and range of adjustment Then check if you can alter the sensing circuit - if you can find it! Not an easy task. Talk to the manufacturers there may be a way they know but I doubt it.

Remember to insulate everything - the availability of 400 amps means that it could spray molten copper all over your face and hands if an arc occurred. If you do increase the volts the power will remain the same so the current will drop massively 400x3.5 is 200x7 ......

Is it really worth all this when perc is trade available if you have use and need. Also is it still available from Sweeden? The purification of perc is possibly more of a problem than the crude creation, Having a factory to make what could be called a terrorist ingredient could be bad. Please consider the HSAW implications of a power electrolysis cell. -- PPE, chemical proof suit, mask, breathing aparatus, plus where are you going to dispose of your waste pH bio and chem oxidiser neutralised