DC power supply for CNC?

use it for what? the application will determine the required purity of the DC

for running servo motors via amplifiers.

Only if you want to fry them. The amplifiers need a well filtered power source. Regulated is even better. What you have is not much more than a battery charger.

Iggy, electrolytic caps are really cheap. just add some. This web site has an explanation of how big a unit to get.

Are you using the existing amplifiers? If so, do they have a power supply already?

On my EMC2 conversion I was able to use the existing amplifiers, servos, encoders, power supply, I just changed the control, saved some money. Looks like you have some hopefully good encoders on the way, especially if they fit where the old encoders were.

RogerN

I do not yet know. I have existing amplifiers, and also three AMC amplifiers that I bought inexpensively.

I am not sure how I wuold figure out the wiring for existing amplifiers, but there is a chance that I can do so.

The hope that I have in gutting the existing stuff, is that I can also switch to single phase and end up with a system that I understand.

I hope so.

I got two 10,000 uF capacitors from digikey, 100v max, 18000 hours life. When charging the caps niitially, would the diodes be stressed?

Oversize the diodes; large rectifier bridges are dirt cheap. I've buit many supplies for AMC and Copley servo amps with just a transformer, rectifier bridge and cap and never had a problem. Oh, and don't forget a bleed resistor for the cap.

It may work great, but if you didn't remove any resistors from it you may find that it has poor regulation under a varying load. Most battery chargers need some sort of current limiting; you can do this in the transformer by designing in a healthy amount of leakage inductance (microwave oven power transformers usually have a slug of transformer material wedged or spot-welded into the core between primary and secondary for this purpose). So you may find that the transformer it inherently incapable of good regulation.

And poor regulation to your servo amplifiers could cause all sorts of weird problems.

I'd try it, but I wouldn't be surprised if it wasn't good enough.

And yes, put some filter caps in there.

Tim, how can I test it? Can I, say, make it produce, say, 15 amps of current, and I would watch the voltmeter to see if it sags a lot?

Yes. That's the test for voltage sag. Testing it without the filter caps will measure the sag from the transformer. With filter caps you have an effect that can either be described as "ripple causing sag" or "ripple that looks like sag" depending on your point of view, but which is, in either case, a ripple whose high-voltage point doesn't change much but whose low-voltage point sags more and more with increasing current.

You'll be measuring a pulsating DC voltage; some (really cheap) digital voltmeters will jump around a lot measuring that, others will be steady (analog voltmeters will be rock steady). If you voltmeter jumps around a lot then try another one, or post again and I'll show you a low-pass filter that'll steady it out.

The servo amplifiers will be able to tolerate some sag -- they're designed to work with power supplies that have considerable ripple, and to them ripple and sag are the same thing. But they'll also have some input voltage limit below which they're just not going to work right, and that input voltage is going to be reached when you're asking for the most torque from your motors.

Basically, you want the lowest instantaneous voltage out of the supply to be above whatever the amplifiers needs, without letting the highest instantaneous voltage out of the supply exceed whatever the amplifiers can stand. You can compensate for some amount of sag by putting in honking big filter caps -- that'll reduce the ripple, which gives you more room for sag. The best part is that the more current limiting you have in the transformer, the more you can just load in the filter caps without worrying about the power factor of the current, because the transformer will be taking care of that.

Ah, come on. You're spoiling Iggy's fun. I got zapped by an electrolytic cap once. Gets your attention, and then you learn all about bleed resistors and never forget them.

Karl

The OmniTurn CNC controls have a big toroid transformer, a moderate sized bridge rectifier and a decent sized single cap to run the servos.

Works just fine with both AMC and Copely amps. Also the Glentec amps are the current amps supplied. Very versitile amp.

Gunner

One could not be a successful Leftwinger without realizing that, in contrast to the popular conception supported by newspapers and mothers of Leftwingers, a goodly number of Leftwingers are not only narrow-minded and dull, but also just stupid. Gunner Asch

I actually already know (though I was not zapped). I learned many things the hard way, but not this one.

If the transformer really does have poor voltage regulation as theorized in other parts of this thread, then the oversized rectifiers would be unnecessary.

Bleed resistors take much of the excitement out of life.

After a bit of thinking, I began to realize that the whole setup is limited by the transformer, and so, putting lathe caps on output would not be likely to ruin the existing diodes. I will try turning it off and on a few times. If the diodes fail I will oversize them.

Tim, OK, that makes sense to me. Now, what is the easiest way to consume 15 amps of DC at 80 volts for a few seconds? I may try some big wirewound resistors that I have

On Tue, 8 Jun 2010 11:17:31 -0500, "Karl Townsend" wrote: SNIP

Hey Karl,

Ha ha ha. I can't tell you how many fairly decent "injuries" loaded caps have caused. Not from the voltage you catch, but from the mechanic reaching into a controller cabinet, getting a "zap" and then REALLY banging or scraping or even slicing an elbow or their head or something else when they "jerk" that body part out and hit the cabinet or a bracket or a screw or some other hard/sharp part of the controller cabinet !!! Stitches required a number of times.

Lot of caps are used as timers on elevators, so don't have the bleeder resistor, and when the power is pulled the normal discharge path is opened, so they stay charged. I think maybe the fact that the power is supposedly "off" is what causes the violent reaction....makes you think, in that less than a split-second, that maybe you hadn't turned everything off, and that maybe you got into the 575 volts or something.

Brian Lawson, Bothwell, Ontario.

ps....of course, it's never happened to me. Yeah!! Riiiigghhhtttt !!