Overvoltage trip on Siemens VFD

[pittengineer]

I have a siemens MicroMaster 420 VFD. It is a 5hp 480V VFD. Ocasionally the VFD will trip on overvoltage. When looking at the line voltages on the drive we read right around 493V L-L. Should this L-L voltage be enough to drive the DC bus into an overvolatge? I would think this would still be within tolerence.

We think that this drive is only in the overvolage condition when the drive is stopping. We are going to try increasing Decel time as well as look into a brake resistor.

Someone here mentioned however that installing line reactors could help with the 493V and bring it closer to the 480V. Will installing line reactors on this drive help to reduce the voltage by much on the line side of the drive? I always thought these line reactors were primarily for harmonics and surges, but I guess they could present an inpendance and thus drop some of the voltage across them? Are these reactors a possible solution?

 

[Skogsgurra]

If it happens during deceleration, then it is most likely a question of increasing retardation time (ramp down) or adding a braking resistor. Just as you seem to be going to do. Reactors will only help if there are bad transients on the grid.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[pittengineer]

Thanks for the response.

I understand that reactors help for transients on the system.

Will the reactors presnent any sort of voltage drop during steady state conditions, or is the impedance of these reactors too low during steady state? I know the impedance of these reactors increases during transients due to XL=2*pi*f.

I would think that anything under 500V would be tolerable?

 

[LionelHutz]

Yes, the reactors will drop some voltage during normal operation and very little if any voltage when stopped.

http://www.eng-tips.com/viewthread.cfm?qid=244719

http://www.eng-tips.com/viewthread.cfm?qid=253429&page=1

 

[pittengineer]

I have seen before that these reactors are typically a 5% of overall circuit impedance. So to do this do you take the impedance of the drive, and then size the reactor to 5% of this value?

For instance to estimate the drive impedance do you divided its V/I to come up with an impedance and then use this value to size the reactor. Once you determinde the impedance of the reactor you can then specify its inductance value.

Does a 5% impedance mean that it will drop 5% of the circuit voltage under normal conditions?

I know you can give the manufacturer the model # of the drive and they will specify the correct reactor, but I want to learn the theory as well.

 

[itsmoked]

You understand a reactor won't help your problem regardless of how they're selected?

And that you could still have a high supply tripping the VFD when the motor isn't running because the reactor does nothing when the motor is stopped but the DC bus is still being subjected to line transients?

And that specifically because the reactor represents a voltage drop you will be paying more to run the motor?

And that same power dissipation will increase the heating of the enclosure possibly causing the VFD to stroke out? And, the space cooling bill to increase?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[DickDV]

pitt, to try to make this a bit more understandable, a line reactor is simply an impedance in the line. A 5% reactor will drop 5% of its supply side voltage when conducting its nameplate rated current.

Since the reactor acts similar to a resistor in a DC circuit, you can see that it will only drop voltage when current is passing thru it. Since your drive is only drawing current when the motor is motoring (driving the load), that's the only time the reactor will drop voltage. And that's not the time when you are having the faults. At idle or when the motor is decelerating (the load possibly driving the motor) the drive is not drawing any current so no voltage drop.

It is true that the reactor will tend to swamp fast AC line transients but it would be a strange co-incidence that a transient came along every time you tried to stop the motor.

What's likely happening is that the load is driving the motor when decelerating which turns the motor into a generator. The generated energy flows backward to the drive and charges up the DC bus capacitors as there is no place else to get rid of the energy. The bus capacitors rise in voltage as they charge until the drive faults on high bus voltage.

The suggestions above that you invest in a snubber resistor package is right on. This package consists of two parts: a voltage-sensing circuit to divert the excess energy off the DC bus, and a resistor bank to receive the energy and waste it as heat. This same package will help swamp any high bus voltage that is a result of AC supply transients.

Hope that helps clear things up a little.

 

[davidbeach]

A 5% reactor will have little affect on a unity power factor current. Under the conditions for which the 5% applies, it will provide a 5% voltage drop for a 0 power factor lagging current. Fault current will be much closer to the latter and load current much closer to the former. A reactor will have little impact on load voltage.

 

[DickDV]

davidbeach, my experience is that an input side drive reactor will drop voltage just as I described above. They worsen low supply voltage problems considerably in buildings with low supply voltages.

 

[pittengineer]

Thanks for all of the responses guys.

DickDV

What you are saying is that the reactors are selected to drop 5% of the supply voltage. So at 480V 5% will drop about 24V across the reactor? Obviously as you mentioned this 5% will be based on the full load amperate rating of the reactor. Is the amperate rating of the reactor selected to match the amperage rating of the drive? I would think that the only way to guarentee a 5% drop in the entire circuit would be to have the current rating of the reactor match that of the drive this way this full load current would cause this 5% drop across the reactor based on its impedance.

What is special about 5% that it is typically chosen?

It sounds like what everyone is saying is that during steady state when the drive is running the reactor will drop a voltage and help reduce the voltage on the DC bus maybe helping things a little for the drive, however when the drive is in an idle state, or in a regen state from the load, the reactor will not help at all?

I guess there is a possibility when the drive is ramping down to another setpoint while running it will still be drawing current and can be in regen and maybe this reactor will help in this case?

Will an isolation transformer serve the same purpose as a reactor? Are there differences between using the two?

 

[itsmoked]

pittengineer; I would suggest you get some local help. It kinda seems like you're not getting it from here. No offense meant, but you seem to be ignoring what we're telling you and just re-framing the same question over again.

You may feel better about some hands-on help from someone knowledgeable who can come to your place, and show you how to monitor the DC bus voltage, so you can see what we are telling you is the problem.

You should [!]NOT[/!] be contemplating a reactor until you have completely disproven the regeneration aspect.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[DickDV]

Pitt, there is nothing special about the 5% rating. I've seen 1,3,5, and 8% reactors with 3% probably being the most common.

But, as itsmoked says, the reactor or even a transformer is not the solution to your problem. You need a brake chopper and snubber resistor to deal with the deceleration faults.

Do that first. Contact your drive distributor and have them figure the parts and prices. Check your drive first; it may already have the chopper part and then only the resistor is needed. While it is also true that you do not necessarily need the same hp brake as you have for drive sizing, at 5hp the parts are so cheap that you might as well just figure 5hp. That's 3.7kw. And, from the sound of things, a 25% duty cycle on the resistor is probably enough. Very cheap!!! Go straight at it!!!

 

[Skogsgurra]

These threads tend to be unnecessarily long. Not because the advice given are bad, but because the OPs usually have an idea how to proceed and then doesn't understand (or want to understand) the advice given. I said (23 Sep 09 21:40) that a braking resistor should be used. And there still seems to be a discussion about line reactors!

It is soo uncomplicated to try a braking resistor: Get a few incandescent lamps and string them in series to get roughly the right voltage (equal to DC link voltage). Connect to the brake resistor terminals (described in the manual) and key in the resistance (hot) in parameter P075. It is set to zero if braking is not activated.

Then test and see if the trips are gone. If not, you may need to add more lamps (in parallel, of course).


Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ozmosis]

There is another factor here too. The Siemens Micromaster MM420 is rated for 380-480Vac. It does have a +10% range but this is not intended for operating, purely overvoltage tolerance.
So, you are in the 'white zone' meaning, you are very close to the over-voltage trip levels and any slight deviation whether it be caused by AC supply surges or regenerative energy coming back during deceleration, is going to trip the drive on over-voltage. The MM420 is therefore doing it's job.
You have two choices: drop your AC supply or investigate your deceleration time. If the stopping time is too quick, you need to fit a brake resistor. Before doing this, and if your application will allow, I suggest increasing your deceleration time (P1121) as long as possible and checking the DC link voltage in a read parameter r0026. The trip level of the MM420 is 810Vdc.
If this is not possible, you need to fit a brake resistor.

 

[mohamedone]

Hello pitt
As we all know,a VFD is a one way device.
since the input voltage ia within tolerence,the trouble is in the output.It happened with me ,my assemblers switched the input and the output,one vfd burned out no protection,another was saved have one.
Therefore,you have a reverse voltage being applied at the output(due to continuous motor rotation) triggering the protection.Add a diverter and dissippater in parallel at the outport.I hope that helps.

mohamed

 

[pittengineer]

I want to first start by apoligizing if it seems as if I was ignoring other recommendations regarding the use of a line reactor. I understood from the first responses that this reactor was not a solution to my problem, however I wanted to understand in depth the reasons why this was not the solution. In part of this understanding, I wanted to know exactly how these reactors did work on VFD's so that I could speak intellegently here to others who are stating that these reactors will be the solution. I want to thank everyone who has responded and again apoligize if it seems like i was ignoring advice and heading in the oppisite direction.

I have gained some new information from the folks that were here over the weekend regarding this issue. I was told that this drive now appears to be tripping while it is sitting idle. This leads me to believe that there may be some sort of transient on our system that is causing this issue. The strange part is there is a Powerflex70 VFD located right next next to this Micromaster on the same power supply that does not seem to be effected by these transients. (Powerflex has not line reactor)

Seeing that these trips are coming while the drive is sitting idle does this lend itself to the possible use of these reactors since someone mentioned above that they were used to subdue transients. Unfortunately this drive does not have an internal chopper, so to use a brake resistor I would need to purchase a chopper/brake external assembley.

Siemens said that the DC bus trip level for this drive was 840V DC. When this dirve is powered up the DC bus level is at about 680V DC which is a bit high (due to higher line voltage) but still within spec. There then must be some sort of transient that is occuring. Siemens recommended the use of a TVSS to subdue any transients but from looking at other threads, it seems as if other reccomend using a reactor over these TVSS's?

 

[DickDV]

This new information would tend to indicate a reactor or dv/dt filter. Additional TVSS is good but limited in their effectiveness if the transient has a very low impedance source.

If you want serious protection, go with a reactor and then a TVSS between the reactor and the input of the drive. That gives you the best of both for transient suppression.

 

[Mobius44]

It sounds like you're being given 2nd-hand info so if possible visit the VFD and make your own measurements and observations. The correct solution depends on when the VFD is tripping.

Siemens is correct, a TVSS will stop voltage transients from over-charging the DC Bus when VFD isn't running.

I believe the above posts make it clear brake chopper/resistor is the solution if you find the VFD trips on overvoltage during decel.

As others have pointed out a line reactor works best while the drive is running, but doesn't do much otherwise.

You might consider what other equipment could be generating the transients (like other VFD's), in which case line reactors or TVSS on the other equipment may solve the issue. It's also preferable to fix transient issues at the source, otherwise you might add other sensitive electronic equipment in the future and have problems too.

Best of Luck!
Sean

 

[pittengineer]

I finally got a chance to go down to the VFD and take some meeasurements. Looking at the bus voltage when the drive was idle I saw about 687V DC. I had the operator then ramp the drive up to 60Hz and then give the drive a stop command. During this Decel period I did not see the drive DC voltage move very much and hung right around 687V. I also had the operator ramp the drive up and down to different setpoints and again watched the DC voltage only to find that it held steady right around 687V. Not once did I see this voltage go high leading me to belive that there regen is not the primary cause of the overvoltage issue.

Seeing that the drive again tripped last night while Idle, I am thinking that there is a transient on the line and that maybe a line reactor will help after all.

I still dont understand why the other 5hp Powerflex 70 VFD that is right next to this drive would not be effected by the same transients. Is it possible that the DC bus on this drive is more robust?

 

[Skogsgurra]

Good to have some real-world observation! After this, it is not very likely that deceleration is the problem. So, the line reactors seem to be en vougue again!

The reason that other drives are not affected by the (probable) transients is very likely that they have a less "nervous" protection level. Some inverters do not have it at all.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...