Best way to regulate a high current PS?

[ButtonPuncher]

Hi,
I am in the design stages of building a high current, low voltage power supply. I just got my April issue of N&V and right on page 3 is an ad for a 14VAC, 1000Watt transformer. I would like to parallel two of these transformers. My goal is to build a 13.8V 140A power supply.<p>I already have a 200A, 400V rectifier assembly out of an old UPS. The only thing that I would need is a giant cap for filtering. I figured that a 2 Farad cap at max load would have about 170mV of ripple. Not too bad.<p>What is the best way to regulate this supply?<p>I was thinking of using a variable transformer on the input side. Either that or a high capacity dimmer. Both are kind of kludgy though. The dimmer idea would take up alot less space, but I'm guessing that the harmonics caused by it, would cause siginificant transformer heating.<p>Any info yould be greatly appreciated.<p>Thanks,
Ben

[rshayes]

You may have to scale things down a little bit.<p>If you are using a 110 volt outlet as a power source, you should limit the total current to under 15 amps (RMS). The wall receptacle may not be rated for any more than this. In older houses, the wiring is probably rated at 15 amps with a 15 amp circuit breaker or fuse. Newer houses may have 20 amp circuits with 20 amp fuses, but no higher. Total power into a resistive load would be 1800 watts.<p>Wiring the secondaries in parallel may be tricky, since voltage mismatches between the transformers will result in circulating currents which will cause some additional loss. Wiring the secondaries in series will give you 28 volts at a maximum of 64.3 amps. The actual current available will be less due to transformer losses.<p>The rectifier will probably need to be a bridge rectifier. This will drop about 2 volts, possibly more. The rectifier loss will be about 128 watts, so the rectifier output will be less than 1673 watts, with an average voltage around 20 VDC.<p>This assumes a choke input filter. A capacitor input filter draws current from the source in short pulses, resulting in an RMS current much higher than the average current. The RMS current will probably be enough to exceed the 20 amp circuit breaker rating long before you get full output from the power supply. The input current of a choke input filter is continuous, with some AC ripple.<p>If the 20 volts is reduced to 13.8 volts by switching techniques, the duty cycle will be about 69%. The RMS current will be about 1.2 times the average current. If the primary current is 15 amps, the secondary current will be about 64 amps RMS or 53 amps average. It would be reasonable to expect an output of less than 53 amps at 13.8 volts (732 watts), since there are other losses which have not been considered.<p>Voltage control could be done with a variac in the primary circuit, an SCR control in the primary or secondary circuit, or a magnetic amplifier in the secondary circuit. The variac will be expensive and heavy. The SCR circuit is more practical (several pounds of copper and iron are hard to melt). The magnetic amplifier would be the most rugged, but would require the design and construction of saturable inductors.<p>You might be better off using one transformer and rewinding the secondary for higher voltage, since the power source will be more of a limit than the transformer.

[Mike]

<blockquote><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><hr>Originally posted by ButtonPuncher:
Hi,
I am in the design stages of building a high current, low voltage power supply. I just got my April issue of N&V and right on page 3 is an ad for a 14VAC, 1000Watt transformer. I would like to parallel two of these transformers. My goal is to build a 13.8V 140A power supply.<p>I already have a 200A, 400V rectifier assembly out of an old UPS. The only thing that I would need is a giant cap for filtering. I figured that a 2 Farad cap at max load would have about 170mV of ripple. Not too bad.<p>What is the best way to regulate this supply?<p>I was thinking of using a variable transformer on the input side. Either that or a high capacity dimmer. Both are kind of kludgy though. The dimmer idea would take up alot less space, but I'm guessing that the harmonics caused by it, would cause siginificant transformer heating.<p>Any info yould be greatly appreciated.<p>Thanks,
Ben<hr></blockquote><p>
might i ask why you need such a low voltage high current supply? the only thing i could think of is a high power amp or somethin but 13.8V isn't going to do too much.<p>i dont think there is anything that can do that.

[Externet]

Hello Ben.
As expressed above, 140 Amperes at 13V is not a simple task.
The bulkiness of the finished supply may well be the same size as using an automobile battery instead, as a buffer fed by a plain battery charger. If you want more running time, get a bulldozer battery.
For beefier chargers, there is commercial starting boosters/chargers capable of aiding the battery for such high demands.
The result will be less costly and the capacitor equivalent of an automobile battery is about 150 Farad
Miguel<p>[ March 28, 2004: Message edited by: Externet ]</p>

[toejam]

I would look into a dc. arc welder and a 20 amp variac.
tj.

[ButtonPuncher]

Thanks for your input everyone.<p>I want to build this supply to use it for testing high power car stereo systems for an extended period of time. I have looked around and a off-the-shelf 2kW 13.8VDC supply is around $800. If I can build one with the parts that I already have, plus a $125 Variac and a $125 2F capacitor, I'd like to give it a try.<p>Size really isn't an issue. In the past I have used three Optima Yelow Top batteries in parallel. If I can make it smaller and lighter than that, I'll be happy.<p>Also, input circuit capacity isn't an issue. I currently have a 10AWG, 20ft run from my breaker panel to my bench with a 20A breaker and 20A outlets. That should handle 2400 watts without a problem.<p>Right now I am leading towards the Variac. It will give me a variable, true sinewave source. That would keep transformer heating to a minimum.<p>How do DC welders regulate their voltage? I'm guessing that they use SCRs because ripple isn't of much concern for something as brute force as welding. Also, isn't the duty cycle on a DC welder something like 10-20%?<p>I have also looked into high efficiency Schottkey rectifiers. A 300A, 80V rectifier only has a 0.72V drop at full load. That would mean a 200W loss with a 144A load. Also, not too bad.<p>I realize now that 2kW would be only peak. I can probably only expect 1500W continuous.<p>Has anyone seen any designs for reverse phase dimmers? That sounds like it might be a good alternative to the Variac. I came across a blurb on this site.<p>Thanks,
Ben

[dyarker]

Stay with the Variac.<p>With any kind of AC duty cycle controller into filtered DC, the output voltage will always try to approach peak at anything more than 45% (with current pulses getting narrower and higher). At less than 45% the output voltage will adjust, but will vary depending on load (bad regulation).
((unless there is some kind of feedback from DC output. In which case you've built a low speed switching supply))<p>With the Variac the peak voltage getting to rectifier is being adjusted. The load regulation is better.<p>Use a full wave bridge rectifier. The rectifier voltage drop is then 0.72 * 2 = 1.44V. No problem though because with 115VAC to the 14V transformers the filtered voltage out will be about 14V * 1.414 = 19.7 minus rectifier, minus half the ripple, minus transfomer resistance loss; call it 17VDC at full current and Variac at 100%.<p>see ya,<p>[ March 29, 2004: Message edited by: Dale Y ]</p>

[rshayes]

With 140 amp output, the ripple due to the 2 F filter capacitor would be about .6 volts, This doesn't allow for the ESR of the capacitor, which could contribute another .2 volts or more. An ESR of .0015 ohms is not negligable at 140 amps.<p>The capacitor will discharge from 14 V for about 1/120 sec. At this time, its voltage will be about 13.4 V. It can only recharge when the input voltage is higher than this. This is about 10% of the time. During this time, the average current must be about 1400 amps to replace the charge on the capacitor. The RMS value of this current is about 443 amps. The transformer is only rated at 140 amps, so the windings will probably overheat. This will also reflect back through the transformer as 52 amps RMS to the AC line. The average line current would be about 17 amps. Unfortunately, the circuit breaker is probably a thermal breaker, which will sense the RMS value of 52 amps and not the average value.<p>What will actually happen is hard to predict. That 1400 amp pulse combined with the .0015 ohm ESR of the capacitor (about 2 volts) may prevent the capacitor from charging to the peak value at all. It will also increase the rectifier drop. There may also be enough stray resistance and inductance in the circuit to limit the current, but this will also limit the output voltage and current.<p>Incidently, Hewlett-Packard made a series of high power lab supplies in the 1960's. These used phase-controlled SCRs with a choke input filter. Some of these might be on the surplus market at a reasonable price.

[haklesup]

Putting a heavy duty automotive battery across the output of this supply will act like a very large capacitor for surges and may reduce the overall power requirements for the supply. Afterall, they cost less than that 2F cap. <p>This may be a good balance between power, size/weight and cost and will emulate a car's electrical system pretty well.<p>I havent recently looked but a regulator module designed for a car or truck might be useful.