Siemens VFD Strange Overvoltage fault

[Rem1061]

Hi Folks,
I am (I thought) fairly good with VFD's but have run into an issue I cannot resolve.
We have a large number of both large and small Siemens VFD's in a fairly new unit, and the small ones keep tripping when our main PF cap banks switch.
As this is a bit annoying, we are installing LINE (not load) reactors. Issue is on one of these units, which normally runs fine, when we install the line reactor it will instantly trip on DC Bus overvoltage every time. Within 1/2 a second. The drive hasn't even begun to increase it's output from 0Hz. Without the line reactor, runs great. Add it and instant trip. Tried the same reactor on a different drive and everything was fine. Tried hooking up a same size (1 hp) motor directly at the drive and it runs just fine with the reactor. I could maybe see reflection on a long lead length, except we are talking about a LINE reactor, and we don't have any problems with DC over voltage until we install the reactor. Any ideas?

Thanks

 

[TwoBlocked]

Sound like the problem is in the drive. A screwed up parameter? Is there another drive with the same or very similar configuration that you could transfer to the troubled vfd? Is there a way to reboot the brain in the vfd?

 

[synchro]

Have you measured the DC bus voltage with an external meter to see what it's doing? Also compare those readings to what the drive will display for the bus voltage. What does the bus voltage do when the drive is started?

Perhaps some resonance phenomena is happening that involves both the input and output circuits to the drive. If so it might not be causing a real increase in bus voltage, but instead could be interfering with the bus voltage detection circuitry and causing it to false. This is just speculation but I'm throwing it out there.
Perhaps you could change the carrier frequency and see if it has any effect.

 

[Rem1061]

This is the only drive that this happens to so far, we just started installing the line reactors when this happened. I have changed out the front end which did not help, and the one other drive that does work, has the same parameters. And as I pointed out, with a motor about 3 feet from the drive, it runs fine. Just having a hard time believing that we are getting reflectance that is getting past the output IGBT's and cranking the bus voltage up to +800 VDC even with the motor not even having time to begin turning. Output never even has a chance to increase, the drive trips instantly. As far as the measuring the bus voltage goes, Siemens does not make things easy to access the board when installed, not sure if I can get to any test points but may try. Really underwhelmed with these Siemens drives and their software really su**s.

 

[winnie]

Do a web search on 'line reactor vfd resonance', and then hope that @Jraef chimes in.

Just to confirm: the drive faults on overvoltage when the _line_ reactor is added. If the line reactor is not present, the system runs correctly. If the motor is right next to the drive (with the line reactor), then the system runs correctly. The fault occurs after the drive starts to modulate, but the output is still at 0Hz.

-Jon

 

[garbo]

Several times I had VFD'S trip out on DC buss over voltage. Every time it was on larger AHCU'S return fans( 60 to 125 HP ) that the supply fan was causing a stopped return fan to turn several hundred RPM. Had to place the 1 or 2 supply fans on hand a bring speed down to 20 to 25% then restart the return fan. For 480 volt 3 phase VFD'S the DC buss voltage is usually 655 to 670 volts DC and forget the voltage that drive will trip out on DC buss over voltage but think its at least 100 volts over the normal 655 to 670 DC buss voltage.

 

[kwired]

Not sure your knowledge and experience but DC bus overvoltage often is because of too rapid of deceleration parameters an/or decelerating a high inertia load, though you are trying to start the motor when this is happening.

Depending on the driven load, by chance is it "windmilling" at time of startup? If spinning backwards this might make conditions even worse. Many drives have parameter that can enable starting while motor is turning, which makes the drive check backfeed to determine what kind of output signal to produce, vs assuming normal ramp from stopped rotor.

 

[Jraef]

I’ve been on vacation in the mountains, no cell service. It was heaven… 🫥

PFC capacitors and VFDs are a match made in hell to be quite honest. There are really complex circuit issues that can combine to cause a resonance, pumping up the DC bus regardless of load. Adding a line reactor is sometimes the last component necessary to make this happen. It’s rare, but I’ve seen it maybe 5 or 6 times in my 30+ years messing with VFDs.

The solution has been to remove one or the other; remove the PFC cap system because it is unnecessary with VFDs (unless the VFDs are only a small part of the inductive load), or remove the line reactor. If you are concerned about “raw-dogging” the VFD, you can try using a Drive Isolation Transformer instead. It doesn’t reduce the line harmonics the same way as the reactor, but on a small VFD that’s usually not a problem. It then has most of the same protective effects of slowing down transients coming into the drive while the isolation helps to diminish the resonance. The problem is, you won’t know if it worked until you try it so it will mean investing a few hundred dollars (plus time) with no guaranteed outcome. I HAVE seen it work however.

Good luck with this. I have wasted an inordinate amount of time chasing this exact issue on some Eaton drives once, before I found a guy who explained the above to me. In that case we convinced the user that the PFC caps were a relic of the past and removed them, problem solved.

 

[kwired]

I’ve been on vacation in the mountains, no cell service. It was heaven… 🫥

PFC capacitors and VFDs are a match made in hell to be quite honest. There are really complex circuit issues that can combine to cause a resonance, pumping up the DC bus regardless of load. Adding a line reactor is sometimes the last component necessary to make this happen. It’s rare, but I’ve seen it maybe 5 or 6 times in my 30+ years messing with VFDs.

The solution has been to remove one or the other; remove the PFC cap system because it is unnecessary with VFDs (unless the VFDs are only a small part of the inductive load), or remove the line reactor. If you are concerned about “raw-dogging” the VFD, you can try using a Drive Isolation Transformer instead. It doesn’t reduce the line harmonics the same way as the reactor, but on a small VFD that’s usually not a problem. It then has most of the same protective effects of slowing down transients coming into the drive while the isolation helps to diminish the resonance. The problem is, you won’t know if it worked until you try it so it will mean investing a few hundred dollars (plus time) with no guaranteed outcome. I HAVE seen it work however.

Good luck with this. I have wasted an inordinate amount of time chasing this exact issue on some Eaton drives once, before I found a guy who explained the above to me. In that case we convinced the user that the PFC caps were a relic of the past and removed them, problem solved.

I was assuming OP had other motors not VFD driven that need power factor correction and it is apparently interfering with these drives when switching said capacitors. I also understood he installed the line reactors and must have resolved most the issues other than one that only goes into overvolt fault the moment they try to start that motor.

 

[garbo]

Not sure your knowledge and experience but DC bus overvoltage often is because of too rapid of deceleration parameters an/or decelerating a high inertia load, though you are trying to start the motor when this is happening.

Depending on the driven load, by chance is it "windmilling" at time of startup? If spinning backwards this might make conditions even worse. Many drives have parameter that can enable starting while motor is turning, which makes the drive check backfeed to determine what kind of output signal to produce, vs assuming normal ramp from stopped rotor.

Believe Danfoss calls it flying start. Only capable of starting a backward running motor maybe 10 to 15% of full motor RPM. On faster running return fans it would try to overcome the fast speed but trip out. I contacted Danfoss to see if I could purchase braking device on some 40 to 75 HP VFD'S but told have to order it with a new replacement drive. I wanted to install a speed switch and a full wave rectifier to put 480 volts to one set of motor windings while drive was off to stop motor. We used full wave rectifier & a timer set for the lowest setting that would stop a three phase motor at another company.