Blower Overvoltage

[JankyPLC]

I've got 6 identical regenerative vacuum blowers powered by 6 identical drives in 3 identical panels all fed from the same 1 main building feed doing the same thing, robot end of arm vacuum tools in pairs, two blowers per tool.

I've also got about 20 randomly timed trips for overvoltage and ground faults between them and I'm hoping maybe someone will have ran into a similar issue and has ideas or advice.

More details:
-Powerflex525 480v 15HP 25B-D024N104
-2 Pole 3550 rpm 11kW motor direct drive to regenerative blower
-Ethernet/IP controlled from 5069-L320ERS2 PLC.
-1.00s Accel (unachievable but need blowers to start as fast as possible)
-25.00s decel but Stop mode is Coast,CF so it doesn't mean anything.
-Run static 40Hz(from PLC), start signal(from PLC) is singular so they both start and stop together
-5 minute timeout in PLC shuts them down to save energy, startup can be immediate, during coast down, or anytime after being stopped due to process variability
-Flying restart is enabled and Flying Restart Current Limit is 200%. Was having significantly more issues with Overvoltage trips prior setting these this way.
-Time between trips on any one drive is between 100 and 1000 hours which has made this a bear to debug.
-Every time I've ever seen a trip was when a pair of blowers tried to start, but I have not seen every trip to know if this is truly correct.
-Every time I've seen this happen, they are able to reset and start fine the second time.

Seems odd to me to get a bus overvoltage fault when trying to start a motor. I suspect it has something to do with the drives trying to 'catch' the spinning blower as its coasting down to spin it back up because giving it more flying restart torque helped reduce the frequency of these faults. Watching the Output Frequency when it starts regardless of success or failure is interesting, hz flashes up around 58-60 for a split second, then down to the normal 0-40hz ramp. You can also see on the attached doc that when drives fault the HZ is always in the 55-60hz range which these never spin at. Any ideas or advice is greatly appreciated.

First plan of action is to extend accel ramp a bit since process isn't running fast enough to notice it, yet, and see if that helps. Braking resistors are another thought given the fault, but what causes this on startup?

 

[joseph_e2]

I have seen weird things on a flywheel machine. What you're seeing during flying restart is the drive trying to "catch" the motor. It's figuring out where the motor is. If you leave it stopped for a long time, you'll see the frequency blip like that and then start at 0. If the motor is already spinning, it'll blip to max and then start at wherever the motor is (or close to it). I bet your max frequency is set to 60Hz which is why you're seeing it blip to there.
Our solution was a large brake resistor sized for a 50% duty cycle. Before that, we had a LOT of weird trips including overload and overvoltage during startup. I reduced the overvoltage faults by extending the ramp time.

 

[robertmee]

Is there a reason you're doing a coast stop? Why not a shortened regenerative decel, and get rid of the flying start? I've always had issues with flying starts on 525...even from a dead rotation, the drive will rotate the load backwards to determine current speed.

 

[dummy_bit]

Maybe need some dynamic braking resistors. I've had to add the braking resistor due to the regen into the drive when the load is coasting to a stop. Comes even more into play if you're trying to stop it completely in a set time. If multiple drives are experiencing this when starting, maybe look into some EMI filters to negate any voltage spikes on the line side. You can also change the PWM frequency on the drives from default 4kHz to 2kHz to avoid excessive regen spikes.

 

[kekrahulik]

Fans can be a real pain. I believe you have diagnosed it correctly that you are trying to catch a spinning motor and suffering from the regenetive overrvoltage. I would add braking resistors. Flying start makes sense. I'm not sure if accel/decel times will help you much.

I assume that one fan can pull air on another fan? That would explain why they trip when starting together. When one tries to start, it will stabilize the fan speed before it trips - that makes the fan speed more easy to start the second time.

The frequency going 58-60 then 0-40hz is probably the drive going into torque limiting in an effort to avoid tripping. Once it is there, it probably won't get enough torque to ever gain control of the fan before it trips - sort of a catch 22 situation.

 

[padees]

My guess is you have a poorly engineered system, and may not even know what these motors really are.

I agree that flying starts can be an issue, but why do you need to do that?

 

[GaryS]

If you are getting a buss over voltage on a drive the tells me that the blowers are not starting together so when the second one starts the other one is moving the other blower so the second blower starts it in in Regen mode until it matches speed
Several solutions are available
Don't setup the vfd to run in 4 quad mode Regen so they don't try to absorb the energy to slow down or sync the blowers
You could install a Buss Loader / DB resister to dump the over voltage
Why would you set the current limit to 200% it should never be more than 125% other wise you risk damaging the equipment
A few things to watch on those Flying Starts is you need to understand exactly how they worm
Some vid's start at the bottom freq and increase the speed until the catch the motor speed those seem to try slow the motor to match the speed. Others start at top speed and then lower freq until there catch the motor using a limited current then they supply full necessary current to control the motor to the commanded speed
It is more complicated than I can explain here
You really should talk to the drive engineers
Let us know how you make out

 

[JesperMP]

Why would you set the current limit to 200% it should never be more than 125% other wise you risk damaging the equipment

Why ?
Surely it is allowed to have a significantly higher motor current than nominal, when it is only short term.
The motor should be protected anyway, either by the motor model in the drive, or by a thermistor (highly recommended for anything driven by a VFD).

(Disclaimer, I have no experience with flying start).

 

[lfe]

Starting large diameter centrifugal fans is not easy, the turbine has a lot of inertia, accumulates a lot of energy, and requires torque, time and high current to accelerate to the desired revolutions.

That acceleration ramp of 1 second seems impossible, I have mounted fans that require about 20 seconds to start, although they were more powerful than yours, about 50kW.

The solution is always the same: lengthen the ramps.

And if it has belt transmission it's even worse, because if you try to accelerate faster then the belts slip, generating a horrible noise.

 

[JankyPLC]

Wow, thank you for all the responses! I'll try to bullet as many answers as I can to the questions.
-Why the need for flying restart? *Variably timed process. Say product doesn't come for 5 minutes, blowers time out and shut down. While they're stopping, WMS sends a request for robot pick which means blowers must spin back up immediately.
-1 second ramp seems impossible. *It is, I timed a few starts and all were around 4.5s. I changed accel ramp to 6s on all VFDs so they can achieve their desired ramp. This morning I had a pair fault on me so that doesn't solve it.
-Why coast to stop? *Initially they were ramp to stop. I'd have to test again(this was changed months ago) but either drives would reliably fault when given run command while still ramping down, or it was just a test to see if it would stop these intermittent trips that we're seeing. I will try setting a pair back to ramp decel to see if what the results are.
-Drives are not starting together theory. *It's possible, but both get run command at same time and both drives typically fault together at the same time on bus overvoltage.
-Set drives not in 4 quad mode to stop absorbing energy. *I am not sure how to do this on Powerflex525?
-Current limit of 200% is only for flying restart torque, actual OL current limit is 125%.
-I suspect these may be drives that start at high frequency and drop till they catch the load then spin up (when I said 0-40% I was referring to a successful start and that part is it's normal ramp)
-I engineered the system so poorly, is an option, ha! All of the drives were auto-tuned to their respective motors for what that's worth.
-I will look into changing the carrier frequency and maybe try that
-I know the braking resistor is probably the answer like everyone is suggesting, but those cost money and everyone knows labors free so have to try all the program-type options to make current setup work before additional hardware will get approved.
-Talking to a Distributor/Rockwell drive engineer is next on my list. Figured I'd ask all you helpful folks before I spend the time going through the first 3 levels of drive support where they just ask me to make sure my motor HP is set correctly and such before getting to the more technical specialists.

Maybe they will simply tell me to install braking resistor, BUT flying restart feels like the drive should be ramping to match the speed of the fan, then starting the ramp from that Hz to full speed which shouldn't require so much torque in my mind. Probably more to flying restart methodology than I could ever google up. Not sure how one even finds that level of technical info for a product. Any ideas?

 

[joseph_e2]

One comment on ramp to stop. If you try to ramp down too fast without a brake resistor (or with too small a brake resistor), you'll get bus OV faults.
I too jumped through a LOT of hoops before ordering larger brake resistors. We had 2x of the standard Powerflex brake resistors wired in parallel and the drive configured accordingly and got a lot of OV and OL faults. One complication was that it was a "weird" motor (high slip NEMA design D) but that *should* have been corrected for by the autotune in SVC mode. That helped some with the OL faults, but didn't do anything for the OV. For us, the final best solution was to enable flying restart to prevent the OL fault, up-sizing the brake resistor to handle 50% duty cycle, and replacing the motor with a NEMA Design B. With that setup, we left it on V/Hz and it's been running fine for months now.
By the way, this application was another where the flywheel has a long coast down (like a couple of minutes) and the operator will be restarting the motor a lot before it's had a chance to coast to stop. We could also have it ramp to stop, but it's not necessary in our application.

One fo the hoops I jumped through was to mess with the accel time. If I made it a very long time, it would *usually* start as long as the speed ref was below 40Hz or so. I ended up configuring a drive output to turn on when the motor was above a certain speed (25Hz or so) and use that to choose a longer accel. That let me start with a faster accel and then switch to a longer one to "sneak up on" the set speed. Even that was only about 90% effective in getting up to speed without faulting and only if the set speed was around 40Hz or less. I spent a LOT of time trending in CCW trying to work around the issues but ended up with a larger brake resistor and standard motor. The standard motor was not enough by itself.

Sorry, that turned into a longer-than-expected ramble...

Edit to add:
In my experience with flying restart, while looking at trends in CCW while the drive starts, the drive quickly and briefly ramps from 0 to max Hz with a very low current. I suspect it's analyzing the voltage and current waveforms or something to figure out where the motor is, and then it starts at wherever Hz corresponds to the motor speed to do its full-power ramp up at whatever rate is configured. As I recall, if you command is 40Hz and it finds that the motor is at 20Hz and your accel time is 10s, it will take 5s to go from 20Hz to 40Hz, so it's using the accel rate rather than the accel time if that makes sense.

 

[GaryS]

First let’s look at 200% current limit that means the motor is developing at least 200% torque on the motor shaft, but please keep in mind the motor windings in the motor are always moving around in the winding slots. The higher the torque the more they will be moving around So every time you do that is one step closer to a winding failure so why increase the risk of motor failure if you can prevent it.

As was said the 1 sec ACC time is way too short for most any fan to get up to speed so every time you start the blower the actual acc time is controlled by the current limit and the load the acc time difference between 100% and 200% in minimum at best I would guess only about a second or 2. So why put the extra strain on you equipment.

Let’s look at the Flying Load some vfd manufactures also call it Catch a Spinning Load. Either way the vfd uses the counter EMF of the motor to know the motor speed and direction to know when it can start controlling the motor the way you ask it to. Now what happens if the motor is actually spinning in reverse when the vfd is commanded to start now it must first find the speed of the motor and bring it to a stop as quickly as it can. To do that the vfd must pull the spinning energy from the motor and send it into the vfd buss. The buss will very quickly trip on an over voltage unless you can dump that energy as heat in the buss loader or a better solution would be to change the vfd’s to line regeneration type that way the energy can be recycled for use by the other vfd’s.

There are so many things to consider when designing drive systems.

A few things to look at to improve the system

I would recommend that you VPI ( Vacuum Pressure Impregnated ) all the motors that will seal and stabilize the motor winding and prevent them from moving I the motor slots it also helps dissipate the heat generated in the motor so they will last much longer. I have done this on many projects with new motors and I have never regretted it.

Depending on the vfd’s you are using you could check out setting them up I a common buss arrangement that way the energy collected to slow down or stop the motor could be recycled for use in the other motor. I have done that a few times. In this day of recycling and energy saving I am surprised that it is not dome more often but it is challenging to do.

There so many things to look at that it’s hard to cover them here.

 

[lfe]

That's like wanting to accelerate in 5 seconds to 60 mph when the car can only do it in 10 seconds.

The acceleration ramp should be extended in accordance with what the blower is capable of.

 

[JankyPLC]

joseph, I'm going to see if there's a way I can catch this on a trend. I've not been able too yet, but maybe that'll reveal something I've missed so far. I do know about too fast ramp to stop being problematic. DC bus/built in resistor can only absorb so much. I have experienced what you describe trying to start the motor at various speeds when I had ramp to stop on. trying to switch from no run command (while ramping down) to run command often resulted in a fault. Switching to coast helped.

Gary, maybe you missed some of my second comment, I turned up the accel from 1s to 6 second which the drives can achieve. Flying Restart is on, the 200% torque is the not Overload limit, its Flying Restart Limit. The drives have a much higher fault rate when trying to catch the load during flying restart with lower torque maximum %. I'm not sure why as it should take very little torque to catch the load if it's measuring it's speed correctly. I have verified these are not spinning backward when this happens, both are spinning down in the correct direction. a common bus arrangement would be nice as would different motors but the hardware is in place and getting approval to swap it all out isn't going to happen preemptively. I appreciate you're suggestions though and will keep in mind.

lfe, acceleration rate has been extended, see my second post in this thread, still has issues sadly.

 

[GaryS]

Wow you post here asking for help with a problem and seeking advise.

I offered my advise you are free to do what you want.

Unlike others I don’t like to put others down I just offer my knowledge and experience.

It’s clear to me that you don’t understand how vfd motor drive system works

When you say that 200% current limit is not an overload I don’t know where you came up with that anything over 100% of the motor full load is overload. It how much over you for how long and how often you do it is what matters it all adds up. Without question at some point the motor will burn up. As I have said before I think your motor is actually driven in reverse by the air flow through it when you try to start it that why you are getting a buss overvoltage fault. Because you need to stop the motor very quickly when you first turn the vfd on. At 200% OL energy feeding into the buss is about twice as much as the vfd can handle. When it trips the control turn off the vfd output transistors and the motor coasts to a stop

I have run into these types of problems many times over the years careful selection and set up of the vfd will solve the problem but you have to first understand the problem. I am not going to chase this any more

If you want PM me and maybe I can help you one on one