VFD failure because of power cycle.

JeffKiper

Lifetime Supporting Member + Moderator
Join Date
Jun 2006
Location
Indiana
Posts
2,462
Does anyone have any proof that cycling power to a VFD shortens the life of a drive? I have some concerns about cycling power all the time during a lock out tag out event. The customer is requiring the employees to lock the machine out whenever they open the door to clear a jam.
Under the ANSI Z244.1 -2003 they are allowed to use alternative means when the task is routine , repetitive, and integral to the machine operations. I am looking for supporting documentation that I can show them to help educate them on safety drives as an alternative.
 
VFD's are usually pretty well protected against power cycles, with current limiting circuitry for charging the caps.

But why not just put a local lockable disconnect at each motor? We do that as a matter of course for all AC motors.
 
Older drives used to simply use a current limiting resistor that was switched out of the circuit after startup to limit cap charging current. Not sure if the new ones still do this.

How often is this restart? If it's only several times an hour it's probably not a problem.

I'm not a safety expert but I thought most drives these days came with a Safe Torque Off input. They could use this and even couple with a safety rated load side contactor.
 
One of the machines at our facility jams on occasion and we need to access it after rendering it safe. We use a combination of Safe Torque Off(STO) contacts on the drives tied with a captive key system to access the machine.

As for power cycling: I would think smaller drives with a limited number power cycles wouldn't cause problems. I would check with the drive manufacturer to see if they have any data on that.


Pi
 
I can tell you from first hand experience that power cycling a drive too often will blow up the capacitors (this is not just removing the start signal and re-applying, this is dropping the mains and reapplying). It was loud and startling. Luckily the drive was a NEMA 12 enclosure type which contained the event. However, the metal of the enclosure was oilcaned out.

This was a while ago but still, most drives have something in their manual that says you must wait x minutes before restart. We didn't read that part but after our learning the hard way, we always put a restart delay timer in the logic to the contractor feeding VFDs. We used to joke that it was the longest 3 minutes of your life when you were waiting for that thing to tick down.

The real way to do it now is with STO (Safe Torque Off) inputs as has already been mentioned in this thread.

If you really want some excitement, power cycle a cheap soft starter more often than recommended and you just might possibly see flames in your control cabinet. I'm not going to talk about that one ;-)

For existing drive, your best tact is to read the manual and if there is a minimum wait time before re-applying power, program a timer to prevent pre-mature powering of the drive.
 
Last edited:
Like Norman Dziedzic Jr., I have let the smoke out of a drive doing this. There was bug in the control PLC code, and the "Reset" brought in the contactor and never latched it. Kept pressing button until drive smoked. I always incorporate a timer in my code that prevents reapplying drive power until x seconds after the last power up(not power down). Minimum of 30 seconds, and that is way less than most drive manufacturers recommend. Almost all new drives have STO and this should be used. I see many controls guys still using only main line contactors for E-Stops, even if STO is available.
 
Yes power cycling a drive does decrease the life. Another thing you see many do is to break the output power to the motor with and contactor and this is done when the estop is hit and the Drive output is still conducting and this is more damaging than power cycling which is one reason safe torque off was developed.

There is no need to do a physical lockout on the machine to remove a jam because in some operations you would never make the production numbers.

The machine does need to be in a safe state so what i would do is use safe torque off on the drives and you could add a safety contactor if you wanted but there is no real need other than making people feel safer.

For employee safety use a trapped key interlock system so the machine can not be restarted while the jam is being cleared.

For easier maintenance and safety a motor disconnect within sight of the motor would be a god idea.

Investing in determining why the machine jams so much would be a good thing to do also IMHO.

At the end of the day all you can do is present solutions to the customer. He may want them to lock it out because they know it's safe and it's cheaper than making a safety system as I have laid out. The will end up paying more by damaging drive after drive but some customers can't see the Forrest for the tree's.

I would give them a rough draft of a safety system that is safe and does not require lockout to clear jams and protects drives and other equipment but if they blow you off let them carry on and make you some money replacing drives for them.
 
Think the system through thoroughly before changing your procedures. Keep things in perspective, too - it is better to smoke a VFD than tear off somebody's arm.
 
In certain industries, on certain machines, I can understand the need to enforce LOTO. But shoving it down everyone throat cuz some old tart smelling up her office decides it is now the mandatory company policy, that's a motivation killer.
But that's another rant for another time.

Yes, frequent power cycles will burn up drives.
The larger the drive, the faster they will smoke.
On a retrofit I was forced to change wiring so mains was removed from all drives the moment an enclosure (with safety switch) was opened.
Smoked a 5hp drive within the first 5 days.

Was allowed to change it to using the STO contacts after 3rd drive burnt and 2 full days of production loss.
 
I'm plagiarizing myself from another forum, but I'm lazy. The question there had to do with the power utility requiring that they install load shedding contactors ahead of everything, including VFDs. Same concept applies. The statement at the end had to do with his wondering if he should install a circuit to turn the VFD output off before allowing it to be powered down, not germane to this thread.

The risk is from when the VFD uses what's called a "pre-charge resistor" to deal with the capacitor charging current inrush. Capacitors charge virtually instantly, and do so by pulling the available fault current. So to avoid having that damage the VFD rectifier and the caps themselves, when power is first applied, it passes through some sore of "pre-charge circuit". ALL drives must have this, but there are several ways to do it. Small drives (up to maybe 15A) use what's called an NTC thermistor, a resistor that DEcreases its resistance drastically as it heats up. So when the circuit is dead, the resistor is cold and the resistance is high, so it acts as a current limiter when power is applied, then the resistance drops to near zero. But it's never actually zero, so it always has some heat hence the size limit. Next is to use a resistor, but have a contact that shunts power around it a second later. That means a relay and controls for the relay and a resistor and its heat. So above about 100A, that too becomes problematic. So after that, many drive designs use SCRs for the rectifier instead of diodes and when power is applied they ramp the voltage into the caps, just like a soft starter. That means having a firing control circuit, but that's now boiled down into a chip set. Then above another size again, maybe 400A or so, many mfrs go back to a pre-charge resistor because of economics.

So why do I mention all this? IF the drive in question uses an NTC thermistor or a pre-charge resistor circuit, energizing and de-energizing the drive every time it is not needed will stress that resistor and it will fail early. Not right away, but sooner than it was designed to. Adding a line reactor can help a little (and is a good idea anyway), but in general most VFD mfrs will tell you not to cycle power if it can be avoided, mostly because it increases THEIR risk if having to replace the drive under warranty (indicating how risky it can be). But if your drive is large enough to use an SCR front-end, but not so large as to go back to a pre-charge resistor, then the concept of removing power every time you are not using it becomes basically a non-issue.

How can you tell? You will have to find someone in the vendor's organization that knows, but sometimes you can figure it out by looking at the spare parts list. If it shows a DC bus resistor, bingo. As to whether or not it's worth it to trench a power-down circuit, no. Telling the drive to power down is necessary for opening a switch on the OUTPUT side, but has zero effect on the INPUT side. So if you have no control over the PoCo issue, I would just inform the owner of the increased likelihood of premature failure so that he knows what to expect and doesn't blame you, use a 5% line reactor and say your prayers.
 
Last edited:
I'm plagiarizing myself from another forum, but I'm lazy. The question there had to do with the power utility requiring that they install load shedding contactors ahead of everything, including VFDs. Same concept applies. The statement at the end had to do with his wondering if he should install a circuit to turn the VFD output off before allowing it to be powered down, not germane to this thread.

The risk is from when the VFD uses what's called a "pre-charge resistor" to deal with the capacitor charging current inrush. Capacitors charge virtually instantly, and do so by pulling the available fault current. So to avoid having that damage the VFD rectifier and the caps themselves, when power is first applied, it passes through some sore of "pre-charge circuit". ALL drives must have this, but there are several ways to do it. Small drives (up to maybe 15A) use what's called an NTC thermistor, a resistor that DEcreases its resistance drastically as it heats up. So when the circuit is dead, the resistor is cold and the resistance is high, so it acts as a current limiter when power is applied, then the resistance drops to near zero. But it's never actually zero, so it always has some heat hence the size limit. Next is to use a resistor, but have a contact that shunts power around it a second later. That means a relay and controls for the relay and a resistor and its heat. So above about 100A, that too becomes problematic. So after that, many drive designs use SCRs for the rectifier instead of diodes and when power is applied they ramp the voltage into the caps, just like a soft starter. That means having a firing control circuit, but that's now boiled down into a chip set. Then above another size again, maybe 400A or so, many mfrs go back to a pre-charge resistor because of economics.

So why do I mention all this? IF the drive in question uses an NTC thermistor or a pre-charge resistor circuit, energizing and de-energizing the drive every time it is not needed will stress that resistor and it will fail early. Not right away, but sooner than it was designed to. Adding a line reactor can help a little (and is a good idea anyway), but in general most VFD mfrs will tell you not to cycle power if it can be avoided, mostly because it increases THEIR risk if having to replace the drive under warranty (indicating how risky it can be). But if your drive is large enough to use an SCR front-end, but not so large as to go back to a pre-charge resistor, then the concept of removing power every time you are not using it becomes basically a non-issue.

How can you tell? You will have to find someone in the vendor's organization that knows, but sometimes you can figure it out by looking at the spare parts list. If it shows a DC bus resistor, bingo. As to whether or not it's worth it to trench a power-down circuit, no. Telling the drive to power down is necessary for opening a switch on the OUTPUT side, but has zero effect on the INPUT side. So if you have no control over the PoCo issue, I would just inform the owner of the increased likelihood of premature failure so that he knows what to expect and doesn't blame you, use a 5% line reactor and say your prayers.

Great post jraef that explains a lot. Can you help me understand why a line reactor helps though? I know it will drop the voltage comping into the drive but how does that help on the pre charge issue?

What about drives on common dc buss and don't have pre charge circuit? Does that matter on power cycling?

Not long ago we order 2 replacement AB drives Flex 755 and they were 50 HP but did not have a bult in pre charge circuit becasue we order the model without it by mistake and i could not understand why they make one without it?

It was for common DC bus or AC connection and since it did not have pre charge for the common dc buss AB told us to connect it using AC to make it work and that is what we did to make it work but I still don't understand how that worked? Seems like there is a pre charge for AC input and one for DC input on 50 AB drives? is that right?
 
An even more robust solution, at least with Allen-Bradley Powerflex products (I'm not overly familiar with others) is to use a Safe Speed Monitor on the drive. It's similar to the SafeTorque Off but monitors the motor's encoder at well, and can be configured to give an output once the speed is below a threshold (or 0) that can be used to allow a gate to be opened or a captive key removed. This module also performs the function of the STO, keeping the drive from enabling until the Safety Circuit is reset.
 
Great post jraef that explains a lot. Can you help me understand why a line reactor helps though? I know it will drop the voltage comping into the drive but how does that help on the pre charge issue?
A line reactor slows down the rise-time of current flow due to the Inductive Time Constant. So even though the capacitors WANT to pull current at the available fault current rate, the reactor will not change state that fast. Ultimately it doesn't change the actual current amount, only the rate of rise. But because caps charge so fast, that helps a little.

What about drives on common dc buss and don't have pre charge circuit? Does that matter on power cycling?
The precharge circuit would be part of the common bus supply. What you are protecting is the FEEDING of the DC bus capacitors. After the capacitors, it doesn't matter any more. So if, for example, you are feeding VFDs with a DC power supply from somewhere else, cutting line power would have no affect on the VFDs, but would have an effect on that big DC power supply (unless it used the SCR ramping pre-charge method).

Not long ago we order 2 replacement AB drives Flex 755 and they were 50 HP but did not have a bult in pre charge circuit becasue we order the model without it by mistake and i could not understand why they make one without it?
So again, if you ALREADY have a separate DC bus supply unit, the pre-chage circuit would be incorporated there, it does not reside in the INVERTER portion. But I'm not sure what you may have ordered. There is no version of a 755 without a pre-charge circuit. There is, however, one that has a pre-charge circuit ONLY for the AC input (by virtue of the SCR ramp control), but if you use the DC input, there is no pre-charge circuit. That is intended to be used in a common DC bus arrangement (i.e. uses inverter only), fed by a line supply (converter) unit sized to feed all of the inverters connected below it. THAT common supply unit is where the pre-charge circuit would have been.

It was for common DC bus or AC connection and since it did not have pre charge for the common dc buss AB told us to connect it using AC to make it work and that is what we did to make it work but I still don't understand how that worked? Seems like there is a pre charge for AC input and one for DC input on 50 AB drives? is that right?
As I said, there is a version in which the pre-charge is only on the AC side. There is another where the pre-charge is on both sides. Depending on the size of the unit, you may or may not have had that no DC bus pre-charge option available but at 50HP it can be ordered either way. If you made that mistake and ordered the version with no DC pre-charge circuit, the solution would be to feed them with AC. The DC bus sharing option becomes a little more complex however.
 
So again, if you ALREADY have a separate DC bus supply unit, the pre-chage circuit would be incorporated there, it does not reside in the INVERTER portion. But I'm not sure what you may have ordered. There is no version of a 755 without a pre-charge circuit. There is, however, one that has a pre-charge circuit ONLY for the AC input (by virtue of the SCR ramp control), but if you use the DC input, there is no pre-charge circuit. That is intended to be used in a common DC bus arrangement (i.e. uses inverter only), fed by a line supply (converter) unit sized to feed all of the inverters connected below it. THAT common supply unit is where the pre-charge circuit would have been.

As I said, there is a version in which the pre-charge is only on the AC side. There is another where the pre-charge is on both sides. Depending on the size of the unit, you may or may not have had that no DC bus pre-charge option available but at 50HP it can be ordered either way. If you made that mistake and ordered the version with no DC pre-charge circuit, the solution would be to feed them with AC. The DC bus sharing option becomes a little more complex however.

We use this to get DC from the 480 for all the drives http://www.pwrx.com/pwrx/docs/ld41_60.pdf

WE had 1316 drive in there that was original and it died. Looking at other 1336 looks like when on common DC buss they have a precharge board in the drive with a resistor the size of a flashlight. Board in each drive part of 1336 is labeled pre charge and sits above the dc terminals.

I think we order the 755 with no precharge by accident thinking it would have precharge board like 1336 did and we did not have external precharge so it was blowing fuses as soon as you tun it on so we wire it up 480 and not use the dc common buss terminals.
 
I'm plagiarizing myself from another forum, but I'm lazy. The question there had to do with the power utility requiring that they install load shedding contactors ahead of everything, including VFDs. Same concept applies. The statement at the end had to do with his wondering if he should install a circuit to turn the VFD output off before allowing it to be powered down, not germane to this thread.

The risk is from when the VFD uses what's called a "pre-charge resistor" to deal with the capacitor charging current inrush. Capacitors charge virtually instantly, and do so by pulling the available fault current. So to avoid having that damage the VFD rectifier and the caps themselves, when power is first applied, it passes through some sore of "pre-charge circuit". ALL drives must have this, but there are several ways to do it. Small drives (up to maybe 15A) use what's called an NTC thermistor, a resistor that DEcreases its resistance drastically as it heats up. So when the circuit is dead, the resistor is cold and the resistance is high, so it acts as a current limiter when power is applied, then the resistance drops to near zero. But it's never actually zero, so it always has some heat hence the size limit. Next is to use a resistor, but have a contact that shunts power around it a second later. That means a relay and controls for the relay and a resistor and its heat. So above about 100A, that too becomes problematic. So after that, many drive designs use SCRs for the rectifier instead of diodes and when power is applied they ramp the voltage into the caps, just like a soft starter. That means having a firing control circuit, but that's now boiled down into a chip set. Then above another size again, maybe 400A or so, many mfrs go back to a pre-charge resistor because of economics.

So why do I mention all this? IF the drive in question uses an NTC thermistor or a pre-charge resistor circuit, energizing and de-energizing the drive every time it is not needed will stress that resistor and it will fail early. Not right away, but sooner than it was designed to. Adding a line reactor can help a little (and is a good idea anyway), but in general most VFD mfrs will tell you not to cycle power if it can be avoided, mostly because it increases THEIR risk if having to replace the drive under warranty (indicating how risky it can be). But if your drive is large enough to use an SCR front-end, but not so large as to go back to a pre-charge resistor, then the concept of removing power every time you are not using it becomes basically a non-issue.

How can you tell? You will have to find someone in the vendor's organization that knows, but sometimes you can figure it out by looking at the spare parts list. If it shows a DC bus resistor, bingo. As to whether or not it's worth it to trench a power-down circuit, no. Telling the drive to power down is necessary for opening a switch on the OUTPUT side, but has zero effect on the INPUT side. So if you have no control over the PoCo issue, I would just inform the owner of the increased likelihood of premature failure so that he knows what to expect and doesn't blame you, use a 5% line reactor and say your prayers.

great information!
 

Similar Topics

Hi there, I have above mentioned VFD which is communicating to my control logix(L62) plc using devicenet DNB scanner(plc side) and 20-comm-D card...
Replies
3
Views
102
Hello, first of all let me state for the record that I'm not a tech of any kind, I have limited experience with industrial automation devices, so...
Replies
3
Views
125
Sorry in advance for the long post, but this requires a little back story. I work in a facility with a couple hundred VFDs Most are in a web...
Replies
14
Views
265
Customer want to use contactor for their 5HP motor blower application. im leaning to VFD. anyone use contractor for blower application...
Replies
12
Views
352
Hi, I would like to assemble a simulator/practice booster pump system that uses PID to maintain steady water pressure under various outlet demands...
Replies
0
Views
99
Back
Top Bottom