Variable Frequency Drive 3

[design2003]

Can any of you make comments on the following experience that I had with <5.0 HP Variable Frequency Drive?

Supply breaker to Toshiba G7 Inverter in a crane application tripped and blew 20 amp supply fuses (all 3 -phases). The breaker is upstreams of the fuses. At the time, the drive was energized but idle (i.e input, but no output). This happened twice in one week period. I ended up replacing the G7s with an S9 drive.

Thanks.

 

[wheelerTi]

It might be harmonic creating such problem and suggest you conduct load profile logging to monitor the problem.

 

[jOmega]

WheelerTi:

Since the drive was in a standby state, no power was being sourced to the load. Ergo, the input power draw was miniscule (milliamperes) at best.... and as a consequence, the harmonics, if any, were completely insignificant.

Because the unit wasn't in RUN mode, the output IGBTs were not likely to be turned on. So I think it safe to conclude that the inverter section wasn't at fault here...even a ground fault could be ruled out.

Problem most likely in the dc bus structure; bridge rectifier, bus capacitors, etc.

or...

you had an upper and lower IGBT in the Inverter section fail shorted at the same time.

It would seem that the problem resulted from a dead short across the dc bus or the failure of one or more diodes in the bridge rectifier. Shorted dc bus cap(s) could also be a cause.

Additional information would be helpful for analysis:[ul][li]did problem occur right after power was applied to VFD or had the VFD been sitting there with power applied for some time when the problem occurred ?[/li][li]magnitude of ac supply voltage[/li][li]was this a bridge drive ?[/li][li]was this a trolley drive and if so, was the bridge in motion at the time ?[/li][li]what components failed/were damaged in the drive ?[/li][/ul]

In many crane applications, ac power is taken from the rails by collectors. The collectors can bounce and ride over breaks in the rails or uneven spots which result in a power transient seen by the input of the drives. Peak transient voltage can cause the rectifiers in the converter section to fail due to excessive peak voltage.

Sometimes, it helps to mount a gapped, 5% line reactor in the ac supply to the VFD and to also install a TVSS package if required.


HTH
jO

 

[tmahan]

The comment &quot;It happened twice&quot; sound as though the drive was still functional. All of the problems above (shorted IGBT's, blown rectifiers ect would have been fatal to the drive. The G7 is smart enough to detect and announce that there was an IGBT failure.

If the supposition regarding the same drive and two faults is correct then it sounds to me almost like the drive was &quot;an innocent bystander.&quot; Additionally I would check the coordination on the fuses and breaker since the fuses should have blown instead of the breaker tripping since they are closest to the presummed fault.

 

[jbartos]

Suggestion to design2003 (Electrical) Oct 21, 2003 marked ///\\Can any of you make comments on the following experience that I had with <5.0 HP Variable Frequency Drive?
///At which voltage?\\Supply breaker to Toshiba G7 Inverter in a crane application tripped and blew 20 amp supply fuses (all 3 -phases). The breaker is upstreams of the fuses.
///What about fuses inside the VFD?\\ At the time, the drive was energized but idle (i.e input, but no output). This happened twice in one week period. I ended up replacing the G7s with an S9 drive.

Thanks.

 

[jOmega]

JB: &quot;///What about fuses inside the VFD?\\\&quot;

The 460vac 'G7' has NO AC INPUT FUSES (FU3) up thru the 200 HP rating. AC fuses only included in the 250, 300 & 350 HP unit, according to their literature. All units up thru 200 HP have a DC Bus Fuse (FU2).... No DC Bus Fuse (FU2) in the 250, 300 & 350 HP units


F Y I JB, most VFD's don't have AC input (mains) fuses in them anymore; particularly in the 75 HP & below class. Several years ago, they decided to leave them out to reduce unit cost. Also most don't have a place inside for accommodating fuses being installed by the user. So, Mains (line) FUSES are now located outboard of the VFD unit.

Ref: [ignore]ftp://ftp.tic.toshiba.com/pub/Drives/ManDrv51546_007_09-22-03.pdf[/ignore]

 

[GGOSS]

Hello Design2003,

If the fuse clearing problem dissapeared following installation of the S9 drives, I would suspect the problem associated with the G7's may be the result of aging DC bus capacitors ie drying up/failing short with temperature.

How long have the G7's been in service?

Regards,
GGOSS

 

[jOmega]

GGOSS:
Recalling Design2003's statement: &quot;at the time, the drive was energized but in idle&quot; ..
I would not expect the bus caps to be at fault at that time as there wasn't any stress on them...At idle, no load current is being pulled out of them...and they should be virtually fully charged at that time, having gone through the precharge cycle.

So, at the time of failure... there would not be any stress on the bus caps; in fact, a lack of stress. Bus caps usually fail from internal temperature buildup. There is quite a bit of overvoltage forgiveness in them. High ripple currents will cause elevated thermal stress leading to failure.

One thing about the G7 is that it does not have a DC Link inductor supplied with the unit as standard. There is a provision for adding one outboard (remove jumper from provided terminals and connect the inductor).

The reason I mention this, is that the bus caps are much more succeptible to transient voltage incidents on the mains (ac line)...

With no dc link reactor, there is nothing to slow down the consequent surge current resulting from a fast rising voltage transient on the ac mains. (As occurs when the collector shoes bounce on the rails or ride over a discontinuity in the rails)

Depending upon where the DC Bus fuse is located in the circuit (before or after the bus caps), and on how closely it is coordinated ... it is possible that the line fuses and breaker could trip before the DC Bus fuse opens; particularly if the fuse is between the bus caps and the inverter bridge.

If the DC Bus fuse is located before the bus caps, and the precharge circuit is minimal or non-existant, then the fuse coordination woulld be such to allow initial inrush charging current to flow without opening. In which case, a fast rising voltage transient on the mains could cause the line fuses and circuit breaker to go before the DC Bus fuse opens. Breaker would have to have magnetic trip as opposed to thermal... inorder to go in such a short time.

Now, since Design2003 states that it happened twice in the same week, I would further assume that there wasn't a bus cap failure problem as it seems reasonable to conclude that they did operate the drive at some time in that week without the fuses and circuit breaker clearing.

 

[tmahan]

JOmega,
Typically Toshiba (the G2 and G3) has used a softcharge resistor on the DC bus. The resistor is switched out of the circuit once the DC bus caps were fully charged. The idea being that the Front end rectifier could short while the Caps were charging.

Has this been removed from the G7??

Additionally I would poit out that the orginal post refers to supply fuses (3-phases) so I think these are external to the drive.

 

[jOmega]

Tmahan,

Thanks for info. Don't know if precharge is eliminated from G7. On first run thru manual, didn't find anything that jumped of the page on Precharge.

Could be that it does have it... but malfunctioned..for whatever reason.

In yesterday's post to JBartos ... I stated that the G7 doesn't have factory installed ac mains fuses in units below 250 HP ...

Interesting to note: the S9 doesn't have ac mains fuses, and also does not come equipped as standard with a DC Bus reactor (inductor) .. Doesn't mean someone hasn't installed one with it.... info not provided by Design2003.
Can only guess.... Maybe it does.. and that would explain why the S9 doesn't appear to have the same failing as the G7 did.... Maybe Design2003 will enlighten us all.

Kind regards,
jO

 

[jbartos]

Suggestion: Visit

http://www.inverter.co.jp/product/inv/catapdf/clg-6491/clg-6491.pdf

for VFS9 protection. There is no internal fuse indicated.

 

[jbartos]

Suggestion: Visit
ftp://ftp.tic.toshiba.com/pub/Drives/ManDrv51546_007_09-22-03.pdf
for G7 manual. There are no fuses, however, there is a reactor circuit on the AC-DC conversion section shown.
Also, Toshiba uses fuses close to the inverter inputs, i.e. downstream of bypass contactor contacts on the inverter input, if a bypass contactor is used.

 

[design2003]

Additional inforamtion on the G7 failures:

The drives were less than one year old, one of the failed one was brand new.
The failed drives were for bridge operation.
Between the two failures, the drives were used to run the crane.
The fuses are external to the drive, for the whole crane control and power circuit.
The G7 we had might have a power supply problem that might have caused loss of -15 v bias to the IGBT casuing it to fire spikes. Can't substantiate.
The failures were discovered on a start of a day, while crane was used the day before.

Thanks for all the responses.

 

[jOmega]

JB:
Sure wish your last post made sense.

The reactor shown in the G7 manual is pictorial only. It's there to show where it would be connected. The drives ship as standard with a jumper installed across the P0-PA terminals. Instructions state to remove jumper when reactor is installed/connected.

What do you mean by: &quot;close to the inverter inputs&quot; ? ? ?

Are you talking about the &quot;INVERTER&quot; section of the G7 VFD ?

and if so, what page of the manual are you pulling this detail from ? ? ?

Take a look at Appendix H, JB. You will see that the particular VFD that design2003 refers to (5 HP) does not have the following:
FU1 - Control Fuse
R, S, & T - Input Fuses
FU3 - DC Bus Fuse

It does have FU2 - DC Bus Fuse

It does NOT have a bypass contactor .

Question to JB:
How many crane applications (Bridge, Trolley, or Hoist) that have a bypass control for bypassing the VFD and going to DOL operation ? ? ? ?

 

[TWW]

As a side bar, the US Navy no longer considers microproceesor based AC drives trustworthy enough for emergency stop operations (with some exceptions). Per CSA 121 (crane saftey advisory 121), the microprocessor based controls are to be bypassed for emergency stop.

 

[jOmega]

TWW,

I believe that NFPA 79 defines, for industrial applications, that 'emergency stop' must remove the power from the motive device. Since solid state devices can fail &quot;on&quot;... it is necessary to have a disconnect device that physically removes the power from the motive device. Ergo, a contactor, Circuit Breaker, Disconnect Switch, etc. are appropriate. An E-Stop button could be used in conjunction with the shunt trip coil of a circuit breaker installed in the ac supply to the VFD and serve as the E-Stop means.

Some VFDs are not able to tolerate opening of an output contactor or switch between the VFD output and the motive device... without damaging the VFD... Ergo, the scheme described above provides the necessary E-Stop function without causing harm to the VFD output devices.

One caveat: repeated short time sequecing of power into a VFD can overstress the DC Bus Precharge Circuit elements causing them to fail.... Manufacturer's recommendations about how often input power can be sequenced should always be obeyed.

 

[Baldor]

Just read through this thread...
Someone I work with sometimes (at another location) had a simialr experience. I don't know all the details but this is what I do know:
It was a crane (not a bridge crane though)
drive 3 &quot;blew up&quot; energized but not running
drive 1 and 2 had been running for a couple of hours beforehand
there is a common DC bus supplying the 3 drives
2 fucntioning choppers, each rated to brake all 3 drives at the same time at full load

I won't mention the drive manf. since it's an OEM thing, but the manf(not just a rep, but the actual drive design engineer) was stumped as to why it happened

If there has been any light shed on the problem, I'll reply back here

 

[jbartos]

Comment on JOmega posting:
Some Toshiba VFDs have only one upstream protection, e.g.

http://www.standardcontrols.com/PDF/TOSH PDF/Toshiba E3 Master PDF.pdf

Toshiba E3 Extender Box Drawings:
Drawing No. 18082F02 shows two sets of fuses:
1. For the protection of VFD input
2. For the protection of Contactor Bypass Loop
which is not a bad idea to have the motor occasionally started without the VFD. This obviously needs different fuses size.

 

[jraef]

Just my 2 cents from experience. The thing that caught my eye here was the fact that it happened in the morning after being run the previous night. I would suspect the conditions were just right to create internal condensation within the drive, possibly right on the cap bank. I have seen this a lot where VFDs are used in exposed areas without proper attention to keeping them warm when idle, i.e. enclosure heaters. Look for a seemingly inocuous change in proceedures that had unforseen consequences. Case in point, I had a customer who noticed one night that the breakers feeding his crane drives were left on and the fans were running. Concerned that they were wasting energy, he turned them off. The next day, the drives blew! Turns out the breakers were left on to supply power to the strip heaters.

Another possibility that I have seen quite a bit is the caps actually freezing, although the net result of re-energizing them when frozen is catastrophic. The fact that it &quot;happened twice&quot; implies that the caps didn't actually blow.


Quando Omni Flunkus Moritati

 

[Marke]

From my experience, I would suggest that the most likely cause of the problem was a deflection in the supply voltage. If the voltage dipped, but not low enough to reactivate the slow charge circuit, and then rapidly returned to normal voltage or higher, there would be a high charging current into the capacitors. If the supply is very strong, and there are no current limiting inductors in the aaAC or DC side, then this can certainly cause the fuses to fail. If it was a problem on the IGBT side, there would more likely be obvious damage to the output stage. If fitted, the DC bus fuse would have failed. Generally, the energy stored in the capacitors is enough to do major damage to the IGBTs if they are not correctly controlled.
I would expect that fitting a DC bus choke or AC Line reactors would have solved this problem.
Best regards,

Mark Empson

http://www.lmphotonics.com