DC bus overvoltage on VFD 1

[gsjhand]

I am working in a mill and have 5HP Powerflex 70 VFD connected to mat former rolls. Drives are running in sensorless vactor mode. Recently I started getting DC bus overvoltage faults on one of the drives. This fault is intermittent. What I have observed is that drive is constantly switching between regen and at speed (which is displayed on its HIM screen) and at same time dc bus voltage varies from 810 volts to 1000 volts. There is a DB resistor connected to the drive. According to my understanding all the energy from regen should go to the DB resistor but in this case this dosen’t seems to be happening quickly enough and I am thinking that this might be causing DC bus voltage to rise above limits. Does this mean that DB resistor is too small? Is there any calculation by which I can calculate required value of DB resistor? Supply voltage is 600 Volts 60 Hz. DB resistor connected to drive has rating 0-1000 V DC (PWM); 120 ohms, 260 watts
I would greatly appreciate any help provided in his regard
Thanks

 

[LionelHutz]

The speed or rate of energy transfer from the VFD DC buss to the resistor is dependent on the resistance. With the DC voltage the same, more energy can be transferred to a lower ohm resistor over a given period of time.

So, you need to install a resistor that has the minimum resistance as specified for that VFD. In your case, it could very well be 120ohms.

Once you have selected the resistance, then the resistor must be sized to be capable of dissipating most of the energy that is being regenerated into the VFD. Unfortunately, this is much more complex to figure out.

 

[blacksea]

More detailed info regarding Shunt operation you may find in Galil AN -

http://www.galilmc.com/support/appnotes/miscellaneous/note5448.pdf

If min Shunt resistance isn't help you, than deceleration time must be increased for lower DC bus overvoltage.

 

[gsjhand]

That was my first guess so I already increased decceleration time from 1 sec to 7 sec but results are same. Thats what stumped me.

 

[blacksea]

Do you see DC bus voltage on Regen resistor when motor starts decelerate? - Be careful - it must be up to 1000B DC!

Do you see deceleration time changes really?

 

[itsmoked]

Some meters top out at 1000V.

Also you didn't state whether this was a new problem on a drive that has been working well, or is this a start up issue?

Blacksea; You can use a shunt on an AC drive??

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[blacksea]

Keith: Why not?
Because modern AC drives have power stage with DC bus.

 

[LionelHutz]

It sounds like your application isn't really decelerating but rather it sounds like the load is just overdriving the motor. In this case, the decel time setting will not help fix the problem.

I think Keith is onto something asking about new installation or previously working installation. If it was working, you need to look at reasons it is acting up, such as someone changing the process and trying to run this VFD slower than before. Or possibly the analog speed signal has messed up and is either sporadically telling the drive to decel - basically, it is cutting in and out leaving the drive running slower than it should be.

 

[gsjhand]

As far as installation is considered this equipment came in operation in Dec. 2007 and this is the first time we are getting overvoltage fault. I checked into the process part and nothing seems to be changed. I agree with LionelHutz that something in field seems to be overdriving the load. But what I think is that DB resistor should be able to dissipate that extra energy.
As far as analog speed signal is considered this drive is getting signal from control logix PLC through control net and there are around 14 more drives connected on same network and they are running fine
I swapped the drive with a new one but the results are same. For parameters I just uploaded from old one and downloaded into new one.

 

[DickDV]

First, check to see if the drive is seeing a stable speed reference or one that is jumping up and down. What's going on with the others is not nearly as important as what is going on with the individual problem unit.

Second, if speed reference is stable, check to see if there is any voltage appearing across the resistor when the DC bus is jumping around. If no voltage is appearing on the resistor, the brake chopper is likely bad. If voltage is present, check to see if there is any current flowing. The resistor may well have burned open.

Third, as for resistor sizing, 260 watts is about 1/3 hp. That would be a continuous rating. If the resistor is cold when you hit it with power just one time, you can figure the wattage at ten times or about 3.3hp. That's probably big enough for a 5hp system with a little friction in it and a moderate decel ramp.

Fourth, if the system is cycling back and forth from motoring to braking and the speed reference is stable, either the load is actually cycling back and forth like that or it is a sensorless vector setup with the the speed loop tuning way too hot.

Advise more and we'll try to help you.

 

[itsmoked]

Have you pointed a temp gun at the resistor? That would be a quick tell as to it being open or not.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[sreid]

I'm with dickDV and itsmoked, Ohm the braking resistor and see if it has opened up. The pulse power in these resistors will find any weak spots.

 

[LionelHutz]

I'd think the resistor described would be one series resistance element so it would either be open or good. But, it may be some parallel elements and one of them has opened. That would raise the resistance but still allow some amount of energy dumping.

You described the DB buss voltage as varying from 810 to 1000V. To me, that sounds about right for a VFD that has a resistor connected to it. The VFD starts dumping energy once the buss reaches about 1000V so that's why it holds there. I was thinking the resistor may have opened but the operation you describe indicates it is working or still connected. It may have changed value but it seems there is something connected.

Just because the speed signal comes from a network doesn't mean it is good. The PID loop in the PLC could be messing up and oscillating the speed signal. The VFD should be displaying a fairly steady speed reference.

If we knew what the process was for these mat former rolls then we could maybe give some further suggestions.

 

[gsjhand]

Talked to AllenBradley about this problem and discussed some of the parameters. They suggested me to change the Stop/Brk Mode, parameter 155 and 156 to Ramp, option 1(It was adjusted to 4 which says fast brake). This setting will offer the fastest stopping time and will utilize the DB resistor. Fast Brake doesn't use the resistor; it monitors bus voltage and adjusts frequency. Also I adjusted Par. 161 to 3(It was adjusted to 1). This parameter is for bus regulation mode. After making these changes I did lot of trials and everything seems to be working fine.
I really appreciate help provided by all of you in solving this problem.

 

[itsmoked]

So checking the temperature of the DB resistor would've shown a problem of some sort. Interesting.

Thanks for the feedback.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[blacksea]

As you may see above, test of real (!) operation of Regen resistor by check voltage across it was suggested before.