G-Rex/Servo Drive overvoltage???

[jabuffi]

Has anyone found out (for better or worse) what happens when the 80V max is exceeded on a G-Rex or a gecko servo drive? I have a 3 axis mill that needs new controls, but the existing motors appear to be rated for 100V... I don't want to sacrifice performance by undervolting the servos, but I would REALLY like to stick with Geckos...

Thoughts?

[Mariss Freimanis]

The G320 (like all of our drives) is rated at 80VDC, designed for 100VDC and comes apart at 114-116VDC. Another well-known drive is rated at 100VDC even though it uses the same MOSFETs as we do. Let's see what happens to each in a real-life situations:

1) You have an unregulated 80VDC power supply and a 100VDC supply for each. Your power company is feeling a little frisky and the AC supply it provides goes to the +10% tolerance limit. 80VDC becomes 88VDC and 100VDC becomes 110VDC. So far, so good.

2) You decelerate very rapidly and cause a 10VDC voltage "bump" on the supply. 80VDC becomes 98VDC and 100VDC becomes 120VDC.

3) Simultaneously you have a large AC line load dump. A compressor shuts off or something similar that causes the shop lights to brighten for half a second. Add another 10% voltage "bump" to your supply voltage. 80VDC becomes 107VDC and 100VDC becomes 131VDC.

The G320 shrugs off all 3 simultaneous events like they never even happened. It was still 7 to 9VDC short of its 114 to 116VDC destruct voltage. The 100VDC rated drive dies when 2 events are simultaneous.

So, how lucky do you feel? Murphy's Law says all 3 will happen simultaneously sooner or later.

How do we know the destruct voltage? We test 2 drives from every production batch of 1,000 to destruction. Why not use higher voltage MOSFETs? Because commodity MOSFETs only come in 100V and 200V flavors. We use 100V 33A rated parts. A 200V part would be 8A rated for the same silicon die size because MOSFET "on" resistance goes up with the square of its rated voltage.

Mariss

[jabuffi]

Excellent explanation!!!

So,IFF I had a supply that was VERY tightly regulated at, say 100V, I MIGHT be able to fend off Murphy and run my existing servos without constantly expecting to "let the smoke out".

The kicker, of course, would be buying/building a supply that could take 10%-20% variances on the line supply voltage... Am I thinking correctly here?

[Mariss Freimanis]

Mr. Murphy is a very clever and devious person. I steer clear of trying to out-wit him.

We rate the drives at 80VDC.

Mariss

[jabuffi]

Exactly what happens when a servo motor is supplied a voltage less than its rated input? I seem to recall that it will run at a proportionally lower rpm...

[Mariss Freimanis]

That is correct. A 100V rated motor will reach 80% of its rated RPM when run with 80VDC. The motor's continuous rated torque will be unaffected.

Mariss

[Runner4404spd]

excuse my ignorance, but can you use a line conditioner similar to one used for a computer to regulate the voltage to the power supply?

[jabuffi]

That's what I'm thinking... and an overvolt load protector on the output as well...

Runner- just so we're on the same page, what do you mean as a "line conditioner"? You're not talking about what most everyone calls a surgeprotector, are you?

[Runner4404spd]

a line conditioner is kinda like a big capacitor. its not a surge protector. a few volts won't be caught by a surge protector and might make it worse since it will trip and power won't be able to go anywhere except into the gecko components. a line conditioner usually has some kind of battery back up as well. they can use the battery power balance out the incoming line voltage so its stable. i just don't know if a large enough one will be economical for home use.

[cjmerlincnc]

Hi, I run 2 Uniterruptable power supply's at home, made by APC and at 1kw output each, they also provide under and over voltage protection and at power fail youv'e got 20 mins before they shut down.

Something like this could be useful to you. I got mine on Ebay cheap.

Regards