Power supply for DC brushed servos

[kmagyoyo]

Here is the build I'm working on

So, I thought I could use the existing DC power supply, but I have found that the output of the PS is 100 VDC, rather than the 20-80 VDC needed by the MachDrive BRB's that I bought. I had thought the motors were using 30V max, so I assumed the PS would be roughly the same.

Do I scrap the big heavy-duty power supply that came with the machine and replace it with the correct voltage for the drives? What do I use?

Or do I figure out how to reduce the output voltage of the existing PS?

If I do either of these, will the output voltage of the drives be too low to move the servos at their max speed? I can't return the drives at this point.

Similar Threads:

• Need Help!- Power Supply/Power Cord Sizing for Four 850 oz-in Servos, One is Slaved
• Need Help!- Power supply for servos
• Suitable power supply for these servos
• OT: (Larger) machine Rutex drives, 42in/LB DC brushed servos, power supply for sale
• power supply for my servos

[Jim Dawson]

If the original power supply is dual voltage input maybe like 120/240V then you could set it for 240V but actually feed it with 120V, that way the output would be 50VDC. The only other option is to remove it and replace it with the proper voltage PS for your drives.

My motors are 140 VDC, the original PS was 100 VDC, and I am now running them at 75 VDC. I can easily get 300 IPM, but I have never tried to go faster.

[kmagyoyo]

The original power supply is a massive beast, the transformer appears to have 3 sets of windings. 3-phase in on the primary and 3 outputs on the secondary. All three outputs are 60VAC. Those feed a separate board that held the Axor MiniSpeed 140 drives. How it gets from 3 legs of 60VAC to 100VDC is a mystery to me, I only see two big rectifiers and two big electrolytic caps.

I must be missing something, because the servos appear to be 10V/1000 RPM. If they have a 3000 RPM max and volt/rpm is linear??, then they should only be using 30V.

These are the servos:

[Jim Dawson]

It looks like the feedback Tach is 10V/1000 rpm. I think the motor is 16.5V/1000 rpm.

I'd have to look at a diagram to understand your existing transformer system.

[Al_The_Man]

The generated BEMF is 16.5v/1krpm, the maximum rpm is 3k.
Therefor the suitable P.S. would be 50vdc plus around 10%, minimum.
If this is a 3ph transformer in, 3ph out, then you may have a 3phase bridge rectifier for the DC.
Al.

[kmagyoyo]

The generated BEMF is 16.5v/1krpm, the maximum rpm is 3k.
Therefor the suitable P.S. would be 50vdc plus around 10%, minimum.
If this is a 3ph transformer in, 3ph out, then you may have a 3phase bridge rectifier for the DC.
Al.

I do know I need to get an Italian involved as much as anything...LOL. But this tells me there should be 50-60VDC on the board there somewhere, because these servos and PS were matched from the factory. Maybe I can upload some photos of the board that would be enlightening.

[joeavaerage]

Hi,
it may not matter that the power supply voltage is higher than is strictly required. It depends on your servo
drives, if they can handle 100V input then you will be OK with the supply you have got.

When you set the max rpm is the servo drive you are in effect setting the maximum voltage to the servo motor.
It would be nothing unusual that a servo drive reduces the input voltage to the volt age actually required by the
servo.

Craig

[Al_The_Man]

Generally you should have a PS that supplies the motor rated voltage plus 10%.
See PDF.
Al.

[Bob La Londe]

If its an old school transformer and bridge rectifier (with caps) type power supply you could just unwind the transformer. I did that on the my KMB1 when I was running GeckoDrives. With a switching power supply you might see if it has a trim pot, and see what you get when you crank it all the way down. Probably not low enough, but....

(I dumped the GeckoDrives and went to Dungong drives and have never been happier.)

[Al_The_Man]

If its an old school transformer and bridge rectifier (with caps) type power supply you could just unwind the transformer.

Very true.
This is why if building my own supplies I use the Toroidal transformer type, very easy to modify and also add a small overwind if a aux supply is required.
Al.

[kmagyoyo]

It has crossed my mind that a servo drive is probably similar to an audio amplifier? (my rudimentary electronics experience is in the AV field) So the PS has a ground, and + and - rails.

Between the + and - I have a clean and steady 98 VDC (what I had been rounding to 100V). From ground I have something closer to +50V and -50V, although it doesn't look very flat on the o-scope.

Is this actually the correct voltage? Should I be concerned about the ~3V ripple?

[Jim Dawson]

A 3 volt ripple sounds about right for a bruteforce PS. If you have +/- 50V to ground, do you have a center tap ground on the transformer?

The only +/- voltage used by the drive is the +/- 10V analog signal, and the onboard (maybe) +/- 15V supply.

The motor is driven by an H bridge circuit in the drive.

[kmagyoyo]

Ok, I’m going to ask Machdrive what they consider an acceptable ripple.

Unless someone has an idea to modify it, I’ll have to curb this PS. Each phase of the secondary windings reads 40vac to ground. If I were to start unwinding the xfmr how do I know when to stop?

A simpler solution may be just to buy a properly rated PS, any recommendations?

[Al_The_Man]

It has crossed my mind that a servo drive is probably similar to an audio amplifier? (my rudimentary electronics experience is in the AV field) So the PS has a ground, and + and - rails.

Between the + and - I have a clean and steady 98 VDC (what I had been rounding to 100V). From ground I have something closer to +50V and -50V, although it doesn't look very flat on the o-scope.

Is this actually the correct voltage? Should I be concerned about the ~3V ripple?

My take would be that 3v ripple on a 100v supply is trivial.
The maximum ripple occurs at periods off high or full current.
As per the PDF, you never use a servo supply where the voltage only meets the motor rating, it should be 10% higher, minimum.
Al.