[erikjgreen]

Hello -

After some time working on other projects this spring, I'm getting back to my replacement of the electronics in my Lagun 250 CNC mill. I'm keeping the servos, which have the following specs:

> Baldor Big Mho Permanent Magnet Motor
> Model # 4201990
> 500 oz-in
> 120 VDC 2500 RPM
> 10 AMPs Constant, 39 AMPs Peak
> Tach output 7.0 V/KRPM
> Volt Constant Kc=40 V/KRPM

...and after getting driven nuts by the whine from the old drives, I've purchased 3x DG2S-16035 servo drives and braking circuits:

CNC4PC


Now I'm building a linear power supply for them. I have my control method worked out, linux-cnc with Mesa cards generating step/direction, and the DG2S drivers handling the encoders (with virtual encoders in linux-cnc).

Here's what I've worked out for the power supply, I'd appreciate someone with experience looking it over and seeing if it makes sense:

I have a 5kva transformer (huge, I know, but it's what I have) with 110V secondaries. I'm going to rewind it to output about 91 volts AC instead of 110 volts.

I'll rectify the 91 volts AC into 129 volts DC (rounding) using a bridge rectifier attached to a big heat sink rated 35 amps and 600 volts.

I'll have a capacitor bank across the output from the rectifier consisting of 8x 4100uf 400v electrolytic capacitors, for 32,800uf.

The above size is a guess by me. Estimating the size of supply I need is where I'm fuzzy. I'm using 130 volts as a target because the drives can handle it and it'll help performance, it's more than the old drives ran at.

According to the sizing rules recommended in the DG2S manual, I should size the power supply for at least the continuous current drawn by each motor. It's not clear on whether that's one motor or three. Rules of thumb I've read elsewhere (in an article linked in another thread by Al the Man, for example, at Power-Supply Considerations For Servo Amplifiers - Page 3) say to size it for one axis and see if it gets hot in operation, increase if needed.

For simplicity, I'll assume it's three axes?

So that means at 10 amps continuous I need a supply capable of 30 amps, to run all three axes at continuous current at once? I realize a lot of the time the draw will be less than this, and sometimes more. Each servo can draw up to just under 40 amps if the drive lets it, but the DG2S drives will fault out above 35 amps, so I'll have to tune the mill to not do that.

So.. assuming my power supply needs to deliver 130 volts at 30 amps, that's a 3.9kva transformer, which is big. If I didn't have the oversize transformer already, I'd be worried about finding one that size.

I've read some rules of thumb that say I could probably size the supply for one axis only and increase as needed (in articles, etc). If I assume I need to size it for 1.5 axes, that's just under 2kva (15A * 130v), which is quite a bit smaller.

I guess my question is this: Does it buy me anything to just go ahead and build a 130 volt 35 amp supply with the parts I have already, or should I try to minimize the size of the supply (and maximize efficiency) using other parts?

The transformer that came with the mill and ran its drives previously I still have, it has a 78 volt open circuit voltage and looks to be about 2 kva. That would mean it output 110 volts dc (post rectification) to the old drives, so the servos were being driven under their voltage (and probably current) limits in velocity mode. I could possibly rewind that one for 91 volts instead and use it, if it makes sense.

Finally, is my cap bank a good size? If I read the DG2S manual right, I should size the ripple caps at about 2000uf for every amp over 80 volts. Since I'm planning on 130 volts thats between 20,000uf for a 10 amp (1 axis sized) supply or 60,000uf for a 30 amp (3 axis sized) supply. I picked 8x caps as a guess for size because I have them on hand, but obviously the number of caps is dependent on the size of the supply.


Guidance requested.

Thanks,
Erik

[erikjgreen]

One more question.. for simplicity, could I just use the big transformer as-is?

The rectified voltage would be 110*1.414 or 156 volts, but the DG2S manual says optimal voltage is between 10 and 20 percent above motor rated...motor rated is 120 volts, so 20 percent above that is 144 volts, which is close.

Of course, 110 is 'nominal' and I think the actual voltage is 117 or so, plus it varies by the hour of the day by a percent or so... which might put me over the limit even farther?

/sigh... never was really good at EE.

Erik

[Al_The_Man]

The Electronic Design article is a good one.
Just to put things in perspective, I have retro-fitted a few Excello Mills using the original DC servo motors, these motors are similar that were fitted to many mills this size, they are SEM 150vdc 13.2amp and 75a peak (before demagnetization), you normally would not reach the continuous torque value, let alone the peak, which would normally mean you have lost control of the servo!
These mills were fitted with a TXfmr that I would estimate to be between 1.5 and 2kva.
After a days use, these Txfmr's are barely hot, IOW, apparently working under their max. rating.
In most multiple servo applications, I would say it would be Extremely rare for all three motors to be all drawing their continuous current at any one given time, same goes for just one motor!
My take would be if you have a supply already used with these motors, as long as the drives will accept the present DC voltage you should be OK to go ahead and use as-is.
This is what I have done in my case, using either A-M-C or Copley drives.
If you require to lower it, you do not usually need to rewind, but take off turns as necessary, a little trickier with EI lamination's over a Toroidal, but I have done it on more than a few occasions.
Al.

[erikjgreen]

I believe I could use the existing transformer. At 78 AC volts and my estimated kva that would give me a supply of about 89 volts and 22 or so amps, over twice the servos' continuous rating. So that would work, I think. I don't have a complete supply since part of it went with the old drives' card cage... I'll add in my caps and bridge rectifier to the transformer.

Is it worth it to increase the voltage? I could peel turns off the big transformer to make any DC voltage up to the 160 the drives can handle... what would be the benefit of going to the optimal voltage (per the dg2s manual) of 144 volts? I'm going to run the drives in torque mode instead of the original velocity mode, if that matters.

Thanks,
Erik

[Al_The_Man]

I also set mine up in the Torque mode, the lower voltage usually means that you may not achieve the maximum rpm of the motors.
RPM relates to applied voltage and torque to current.
Al.

[erikjgreen]

Ah, the light bulb just went on. I'd heard that before some where.

Hm, the motor data doesn't quite make sense to me then... 40V/KRPM means that at 120 volts as on the sticker it should be rotating at 3000 rpm? Unless the efficiency of the motor is 83%?

Anyway, thanks for the reply. Now I just need to decide if it's worth it to mess with that transformer or not. Probably not, it would make more sense to use the old one for now and upgrade the servos to something more powerful in the future if I want faster rapids or acceleration, or more force for cutting something (what, I don't know).

Erik

[Al_The_Man]

It is easy to determine the max feed rate based on the applied voltage and the resolution between the motor and the final moving object, table, using the present gearing/reduction and B.S. resolution etc.
Al.