Wierd problem with a VFD 1

[VinnyPs]

I have a control panel with 3 VFD. Two of them are 2HP and the other one is a 5HP. The VFDs are 600V, 3PH. From the main I go through a fusebox that protects my VFDs. There is one fuse for each phase of each VFD, so there is 9 fuses in all. Nothing is plugged at the exit of the VFDs. As soon as I put the power ON, the 9 fuses blow. I have noticed that they blow only when the 5HP is plugged, so I took it out. The two other ones are ok. Because of that I thought the problem was my 5HP VFD, so I got another one. Same thing. As soon as I put the power on the 5HP, at least 2 of the 3 fuses blow instantly.
I looked everywhere for a short, and didn't find any. Because the client wanted the panel as cheap as possible, I did not put any filter before the VFDs. The only thing left was to put an L filter. After putting the filters on all the VFDs, the fuses do not blow anymore. Does anyone would have an explication? The line reactors (filters) are after the fuses and before the VFDs. I understand that they protect the drive, but because they are after the fusebox, they should have no consequence on the fuses. Why did this fix my problem? What was blowing my fuses?

Thanks for any help. All I want is to understand.

 

[Skogsgurra]

Sounds like the charging current for the DC link isn't limited. Adding an impedance (the L) reduces current so that the fuses don't blow.

Are the fuses fast or slow blow? Are they selected according to VFD installation manual? What make are the VFDs?

Gunnar Englund

http://www.gke.org

 

[Barry1961]

The capacitors on the DC buss on a VFD can draw very high current when first turned on. There is usually a current limiting resistor to keep the inrush from being excessive. The bigger the capacitors the better performance you get on some aspects of the VFD.

I do not know the exact correlation between the current limiting resistor, buss caps and starting torque but I am fairly sure you want the caps big and resistance small which leads to high inrush.

Barry1961

 

[mil3]

I had a similar problem, however my condition was one where the pre-charging circuit failed on the VFD. The pre-charge resistor bypass contactor was closed all the time and thus no soft charging of the DC Bus was happening which caused high inrush currents.
In your situation it is a new product. I am curious to know what manufacturer it is. Also like the previous posters asked....what fuses are being used and what size are they?

 

[DickDV]

A careful reading of the original post shows that all 9 fuses blow when the 5hp drive is connected. This cannot be cap charging current because the inrush would only flow in the drive with the problem.

Rather, having just worked thru a similar experience, I suspect that the power supply is one of those cursed floating delta or high resistance grounded systems. When you have more leakage current to ground than the drive fuse ratings, the virtual ground in the drives cause the fuses to blow on leakage current alone.

You can verify this by removing all grounding from the three VFD's and then turning on the power. If the fuses hold, its time to install a transformer with a wye secondary with the center of the wye properly grounded. The filters in the line probably reduced the leakage current just below the fuse clearance point but, as soon as you try to load the drives, the fuses will blow again.

 

[Farkel]

Your reactors have current lag, whereas the capacitors in the VFD's lead. So you limit the inrush with the reactors, so the fuses don't blow.

 

[VinnyPs]

First, thanks a lot for all your replies.

The manufacturer of the drives is AC Tech, The fuses are fast acting (as suggested by manufacturer) and the size of the fuses are 10A for the 2HP and 15A for the 5HP. AC Tech suggested using reactors in the following situations:

1. The supply power is subject to disturbances such as surges, spikes, dips, transients, etc.
2. The supply power is very stiff (greater than 10 times the kVA rating of the connected VFDs)
3. Harmonics are a concern

I should've used reactors because the plant is not on a power supply, but plugged directly on the energy distributor's 600V. So the line is far from clean or constant. Because of money issues, I had to cut the cost as much as possible and did not put reactors in the beginning. It ended out costing more in the end.

What I didn't understand then, was the correlation between the fuses blowing and the reactors, but what you guys told me makes a lot of sense… thanks again.

 

[VinnyPs]

Hi again,

It was all good until..... it happened again. The fuses of the 5HP blow again. After 1 week of running I got called, the fuses are blown. Do you guys have any idea of what should I do next? A RLC filter on the input? Output reactors?

When I change the fuses, they hold, but for how long? I can't leave it like that, I have to find out why it happened again... Do you have an idea of what to do next?

thanks a lot,
Vince

 

[Skogsgurra]

Standard procedure is to record mains currents with a fast recorder. You need to resolve down to about 1 ms/division. Use DC clamps and find out what relly happens. Also record DC link voltage.

You say that harmonics are a concern. Does that mean that you are concerned or that the harmonics level really is high?

One thing that inverters don't like is voltage asymmetri. If one phase is down more than a few percent, the other phases take over the current. That usually means that one of the fuses don't blow (the one fed from the low phase). Check phase-phase voltages. They shall stay within about 2 percent from each other.

I hope that you are not using very fast "semiconductor" fuses. That shouldn't be necessary and is sometimes the reason for problems like yours. I would even suggest that you use slow-blow fuses. It is not common to have any overcurrent at all once the inverter has started (output is current limited). And since the fuses blow during start, I think they are to fast. Simple as that.

The next option is to talk to the manufacturer. He should be able to help you.

Gunnar Englund

http://www.gke.org

 

[DickDV]

I would suggest, in line with my earlier post, that you install new fuses and then measure the current thru the fuses with a clamp-on ammeter with the drive idling (not running the motor).

I predict that you will read surprisingly high current in at least two of the three phases.

This current is leakage current from phase leg to ground elsewhere in the facility and finds a path from ground back to another phase pair thru the drive and the drive input fusing.

This only happens on floating delta or high resistance grounded wye power networks.

The solution is to install a drive isolation transformer with a wye secondary and the center of the wye directly connected to ground.

As I mentioned in the earlier post, you can test to see if ground leakage currents are the problem by lifting all ground connections from the drive frame by insulating the mounting and removing all ground wires. (I assume the drive is inside an enclosure for safety). If the problem goes away with the drive floating, the transformer is the answer.

 

[jraef]

Just a hunch, but you mentioned that they are AC Tech drives. Have you checked the diode bridge on the input for shorts? I don't want to denegrate any manufacturer, but at 5HP and below I have problems with their 575V design because on the ones I have seen, they use exactly the same IPMs that everyone else uses for 480V max. (IPM is the Integrated Power Modules, a singular module that contains the diodes, transistors and firing circuits all in one).

I am not saying they are doing anything wrong because I do not know anything else about their design or testing, but that has always bothered me. If you have noticed, most other manufacturers only sell 575V drives at above 5 HP, or if they have <5HP they are larger dimensions and more expensive than the AC Tech. That's because they have to use stand-alone diodes and IGBTs to get that voltage rating. IPMs allow for a smaller and less expensive design, but AFAIK, nobody officially offers an IPM rated for 575V use. I may be wrong, and someone please correct me if I am, but that's how I see it.

If a diode is shorted, and the new one shorted immediately upon energizing it, it is likely that your drives are not surviving the applied voltage, especially if it's a little over spec.

"Our virtues and our failings are inseparable, like force and matter. When they separate, man is no more." Nikola Tesla

Make the best use of Eng-Tips.com
Read the Site Policies at faq731-376

 

[mil3]

Jraef is correct. Maybe we should step a few steps back and go back to basics and first confirm that we have a 100% functional drive first. Very likely could have a damaged front-end, plus what makes it more likely is the fact that you mention bad power at your plant and you never installed line reactors. For those sizes of drives a line reactor is very cost effective and very cheap "insurance". Especially at 600vac I would have installed them.
So if this gets resolved, I still suggest that you install reactors on the line side.
There are some really good technical commenst mentioned above by everyone, and I am interested to hear your feedbakc after measuring all they had to offer.
Let us know.