Servo drives exploding after changing EMI filter

[jurgis1991]

Hi,we have a CNC machine that uses 7 servo drives (3 phase drives, Yaskawa 200V). Recently, someone destroyed (shorted phases together) one of the motor wires, which caused the drive to explode. The drive and wire were replaced,but we noticed that the EMI filter had one phase burned out, so we replaced the EMI filter (a no-nema chines 3-phase filter) with a 3-123-506 mains filter, a 3-phase 64A 760V 900uOhm Schurter filter. After the replacement, when we powered up the CNC, one of the drives exploded. The safety capacitors (from phases to ground) and MOV had exploded. We checked all the phases and neutral ground wires and found one that seemed to be lost. After we powered again and a different drive was expeloded (MOVS that goes between phases),
So now I really don't know what to do. Can the problem be that the new EMI filter does not filter the noise as well as the old one, which causes big voltage spikes that destroy MOVS?

Thanks for the help!

[Jeroen3]

Are you confident the power feed is still solid? No missing or high impedance neutral or earth?

[jurgis1991]

Yes, I verified every connection and attempted to load a natural wire while measuring the voltage to ground.

[Jeroen3]

The thing that bothers me most of your story is that you say the Y-capacitor to earth and a MOV exploded.
Suggesting an overvoltage between two phases and phase and earth.

Could there still be a short present somewhere?

[jurgis1991]

Wouldn't the main fuse blow rather than the MOVs if there was still a short?

[janoc]

Wouldn't the main fuse blow rather than the MOVs if there was still a short?

No, you could have a fault that a phase is shorted to the neutral somewhere, with the original neutral connection broken. Then your MOV gets the higher phase to phase voltage than the usual phase to ground one -> KABOOM!
And the main breaker/fuse won't blow because the neutral/ground is broken, so there is no current flowing because of that short. This has happened at my parents' house after a storm (and some really dodgy house wiring), leaving them with 400V in a few outlets and lighting fixtures ...

In short, don't mess with it, call an electrician to check all the wiring. Something likely got damaged in that first mishap.

Things exploding on high power/energy circuits is never good and trying to fix that yourself without knowing what you are doing is a good way to end up hurt (or dead). Your filter replacement likely didn't have anything to do with it at all - if the filter was bad it would either blow up/short/blow fuses or not work at all, breaking the connection, not cause stuff after the filter to explode.

[jurgis1991]

I forgot to mention that the machine worked(for 2 days) ,okay with the broken filter because I didn't notice right away that a phase was missing.
So thats why I was thinking that filter had something to do with it. Also, these drives don't have a nutural connection (they don't need it; it's a 3-phase system), only a ground connection through a ground wire, and the drives heatsink (body) is fixed to the electrical panel's back plane, which is also grounded.

[coppercone2]

why would a filter cause a mOV to break. The mov requires a significant amount of current to be destroyed.

The only thing there to have enough energy to blow up a MOV is the mains.

I think you had lightning or something bad happen that compromised all the drives and they were gonna fail no matter what, or there is a active problem on your power. Are you sure a person did this damage and not a surge did this damage and a person got blamed?

[jurgis1991]

They drove over the cable witch caused a motor phase to short to ground inside the cable and blow up the drivers IGBT package and fuse.

[BeBuLamar]

The drive may have a shorted rectifier.

[Circlotron]

Is the EMI filter between the mains supply and the drive, or between the drive and the motor? If it is between the drive and the motor then definitely a specific type of EMI filter is needed. With a standard of the shelf one placed between the drive and the motor the PWM voltages will give the X and Y caps something to think about.

[coppercone2]

He only replaced one filter, but he said another drive exploded. In that case every drive would have a different filter


I think if a load (motor + filter) on drive 1 was causing drive 2 to explode that is a very bizarre situation. are they really that shitty?

[jurgis1991]

Ther is only one filter on supply side for all drives

[jurgis1991]

Here is the picture (old filter) of all the drives

[Jeroen3]

I do not know about the mains system. But a short on a phase can move the neutral voltage, or even earth, up causing all sorts of strange problems.
Especially if the plant is on a transformer, or the machine is on a transformer or generator.

It can also be bad luck.

MOV's do not require a lot of energy, they blow if you inject a few watts. Their purpose is to short spikes of microsecond duration to protect the silicon diodes, they will burn if you give them millisecond duration overvoltage, such as one or more mains cycles.

You could do some Line and Ground impedance tests.

[Silenos]

You have a bigger or lesser shorcut somewhere which causes things in series explode, does not matter if it a filter or mov or rectifier bridge or igbt. Or 400 V where it should not be. Do drives run only on 3p, no neutral/PEN?
Anyway, I would start with disconnecting the equipment and check the power side - maybe something in breakers box got damaged, like some botched connection vaporized from the surge, and you don't have a neutral anymore. Or you have a phase contact to the grounded/PENed element somewhere. Even check the power with naked asynchronous engine if you don't have the equipment and do not fancy the proper electrician.
Many things may be wrong, hence check piece by piece. And listen to the German.

[WattsThat]

The filter could have capacitors where the original did not. Add the loss of a neutral connection or a phase to ground short and the additional caps and diodes of the drive front end can form an unintended voltage doubler - causing voltage to exceed the MOV rating and poof blown MOV’s.

Happens often in the US with corner grounded delta supplies when you add emc filters.

[coppercone2]

Would those not need to be big capacitors in the filter?