VFD's & Line Reactors

[Besoeker]

Just to clear up a misconception though. That drive does NOT include a "line reactor", it includes a DC bus choke (inductor). Functionally they both provide a small amount of harmonic mitigation,

The DC choke does a better job in that respect.

 

[jumper]

Totally cool thread.:thumbsup:

Out of my league, but really interesting and fascinating to read.

You guys really know your stuff.

 

[Besoeker]

Out of my league, but really interesting and fascinating to read.

That may be more profound than you realise.
"The beginning of understanding is the acknowledgement of ignorance."

 

[ghostbuster]

I too think this sounds like a DC bus regen problem, only indirectly related to your power glitches. That drive does include a Dynamic Braking chopper, so you could set it up to be used as a way to dump off any excess DC bus energy from regen by simply hooking up a resistor.

You might also want to look for a feature called (something like) "Flying Restart" which provides the drive with the ability to detect that the motor is spinning and catch it on the fly to re-accelerate it. With that, you might avoid some of the regen in the first place. If you find it and it is not enabled, enable it and try it before buying the resistor.

Just to clear up a misconception though. That drive does NOT include a "line reactor", it includes a DC bus choke (inductor). Functionally they both provide a small amount of harmonic mitigation, so from that standpoint you don't need to add another line reactor. But that is not the ONLY reason to have a Line Reactor, in fact it's not even the best reason. A real Line Reactor ahead of the drive is like cheap insurance for the DRIVE. It slows down the rise time of line transients caused by OTHER external events and helps the drive rectifier components survive better. A DC bus choke is ZERO help in that regard, because it is BEHIND the rectifier, not ahead of it. So your vendor was an ignoramus.

I agree with Jraef.A fast rising voltage transient (in the nanosecond to millisecond range from utility switching) could be entering this drive.Adding blocking inductance on the front end will slow down the voltage rise time.Also you may consider adding a TVSS unit.
Adding power factor capacitors near this drive will have the exact opposite affect refer to:

http://forums.mikeholt.com/showthread.php?t=156975&highlight=dc+drives

Drives can be purchased with 3 to 5% impedance input drive reactors to "dampen" a utility voltage transient rise problem.
Good Luck

 

[Besoeker]

I agree with Jraef.A fast rising voltage transient (in the nanosecond to millisecond range from utility switching) could be entering this drive.Adding blocking inductance on the front end will slow down the voltage rise time.Also you may consider adding a TVSS unit.
Adding power factor capacitors near this drive will have the exact opposite affect refer to:

A plain (uncontrolled) diode bridge with a blooming great big capacitor bank on its output is a pretty good way to clamp transients.............

 

[don_resqcapt19]

I don't see this as a restart issue...it appears that either the supply voltage rises enough to raise the DC bus voltage to the trip point or the over running load causes the rise in the DC bus voltage.
You need to know what is causing the high DC bus voltage before you can look at solutions to the problem.

 

[Besoeker]

I don't see this as a restart issue...it appears that either the supply voltage rises enough to raise the DC bus voltage to the trip point or the over running load causes the rise in the DC bus voltage.
You need to know what is causing the high DC bus voltage before you can look at solutions to the problem.

I agree.
I tend to think it is the latter. The DC bus has a large capacitor bank. That would generally swallow supply transients without appreciable voltage rise.

 

[kwired]

I don't see this as a restart issue...it appears that either the supply voltage rises enough to raise the DC bus voltage to the trip point or the over running load causes the rise in the DC bus voltage.
You need to know what is causing the high DC bus voltage before you can look at solutions to the problem.

Isn't the overrunning load and the restarting issue kind of both the same issue if the drive is not configured to restart on the fly? Or at least is somehow a combination of two different issues?

 

[jim dungar]

Isn't the overrunning load and the restarting issue kind of both the same issue...

The OP never mentioned that the drives are actually trying to restart. The OP simply said that after a power system 'blip' the drives trip on DC bus overvoltage.
Everything since then has been a guess at what items could cause this type of overvoltage, without ranking them based on probability.

 

[shoon]

Many thanks to everyone for sharing their knowledge! Gives me quite a few things to try out. I'll try and provide a little more info on what I've tried and what our setup is.

The OP never mentioned that the drives are actually trying to restart. The OP simply said that after a power system 'blip' the drives trip on DC bus overvoltage.
Everything since then has been a guess at what items could cause this type of overvoltage, without ranking them based on probability.

Nailed it.
We have a 4160MCC that supplies power to our 4 main drive motors, as well as 2x 4160 - 480v transformers that feed 2x 480v MCC's.

These drives seem to be the most sensitive equipment we have as far as tripping. There are a few things that cause them to trip, but for the most part its shoddy power coming from the utility that causes it.
-Anytime we are starting any of our main drives (4160V compressor motors: 2x 1750HP and 2x 4157HP) we utilize softstarts on all, and a 6410KVAR capacitor bank for the 4157HP motors. Not as big a deal, although I imagine its not too good for the VFD's.
-If there is ANY slight blip in power. Usually occurs when its windy, these VFD's are the first to trip but everything else stays running.

We had installed a power quality analyzer to see what was going on with our 480v BUS during a start of our 4160 motors.
I recorded the highest and lowest values in the trend:
Spike peak: 517v
Spike dip: 407v

The manufacturer provided a little bit of help, and we tried playing with the flying start settings. As its setup now, the drive will still trip on DCBUS overvoltage (this is the main issue), and then attempts to restart with the flying start settings tweeked (it restarts without issue). My concern is the drive is tripping for a reason, and to continually trip and program it to auto-reset doesnt seem like the proper way to handle it. I am thinking the best way to address the problem is figure out whats causing the trip (as the trip is implemented to protect the drive) and find a remedy to reduce its reoccurance.



I think the bottom line is our utility feed isnt the greatest and this is whats likely causing our issues. Due to the fact its usually quite windy when we have power blips, you guys have got me wondering if it might also mess with the inertia of the fans and the VFDs ability to control them. There are so many settings and menus on the drive, its hard to tell where I need to start looking!

Edit:
Scratch that. There are a total of 6 VFD's, 3 are fed from one of the two 480 MCC's respectively. For all 6 of them to trip simultaniously, I am now thinking it must most defintely be an issue with the power supplying them that is causing the DCBUS overvoltage.

 

[shoon]

Some more info: the DCBUS voltage thresholds are configured to 400v low, and 800v high.

 

[winnie]

Any chance of getting a scope on your 480V bus, to look at the traces of all three phases during a spike?

I am wondering if one of the line-line instantaneous voltages is actually exceeding 800V for a time. 517V as an _RMS_ value doesn't tell you what the peak looks like.

-Jon

 

[GoldDigger]

View attachment 9660

Some more info: the DCBUS voltage thresholds are configured to 400v low, and 800v high.

If you are getting unregulated rectified 480 nominal feeding the DC bus, the 517V would produce a peak DC of ~730V. Close to the 800 trip point, but would imply that the recorded 517V is not completely accurate (may be peak 517V RMS AC, but with waveform distortion, etc.?)


How high a voltage can the DC bus withstand (i.e. can you set the trip point higher?)

 

[powersupplyguy@netzero.ne]

This is interesting post. I have spent most of my engineering life working for drive companies.

Do you have auto restart enabled? I did not see that in your post. If in fact there are random overvoltage trips then a braking unit may help. If your process can withstand the overvoltage events then auto restart might be best choice. If the inverters are located in the same area you may not need brakes on every unit. One or two dynamic brakes may be all you need to clamp the voltage.

If you are trying to catch a spinning motor then I would expect you to see some over current events. There are a few things you could try. increasing the acceleration rate may help as well as reducing the software current limit might help. There are simple algorithms that some manufacturers use to catch a spinning motor. Generally they start the frequency at 60Hz and the voltage at zero. The frequency is ramped down while the voltage is ramped up until motor has been caught. If you have this software option available then you may give that a try.

Good luck. I would like to hear what final solution is.

 

[powersupplyguy@netzero.ne]

I am sorry. I am new to group and did not see pages 2,3&4.

 

[qcroanoke]

I am sorry. I am new to group and did not see pages 2,3&4.

Don't sweat that. I am very interested in this thread but I would not know a vfd if I was looking at it. My company doesn't do that kind of work.

 

[kwired]

The OP never mentioned that the drives are actually trying to restart. The OP simply said that after a power system 'blip' the drives trip on DC bus overvoltage.
Everything since then has been a guess at what items could cause this type of overvoltage, without ranking them based on probability.

Correct, but the fact that they could be trying to restart while on the fly hadn't crossed his mind either until it was brought up here from what I can tell, and that sounds like a good possibility that is what is happening. I do agree that it does need some further research to determine just what is causing the overvoltage on the bus.

He may want to also check out just what the output of the drive is doing when this blip happens, and what is it programmed to do when receiving a stop command. If it is programmed to decelerate instead of coasting when a stop command is issued, it may have a better chance of recovering without an overvoltage bus error if it should be receiving a stop command for only a brief time.

 

[kwired]

Don't sweat that. I am very interested in this thread but I would not know a vfd if I was looking at it. My company doesn't do that kind of work.

They are starting to appear in more than just industrial environments, I even have one at my house that runs the well pump, kind of marketed as a "constant pressure controller" but it is essentially a VFD.

 

[GoldDigger]

They are starting to appear in more than just industrial environments, I even have one at my house that runs the well pump, kind of marketed as a "constant pressure controller" but it is essentially a VFD.

Inverter type compressors in AC, especially mini-splits, are becoming more common. But their operation is transparent to the users.
Some inverter type well pumps are particularly attractive for off grid people because they will run off DC input over a wide range of voltages and can efficiently throttle back their output to match the available power.

 

[Besoeker]

Any chance of getting a scope on your 480V bus, to look at the traces of all three phases during a spike?

I am wondering if one of the line-line instantaneous voltages is actually exceeding 800V for a time. 517V as an _RMS_ value doesn't tell you what the peak looks like.

-Jon

A transient spike on the supply isn't likely to do much to pump up that big bucket on the DC link.