VFD and Voltage Distortion

Retracted, I mis-read the post

Thanks mr.JRaef for your suggestion.

You have mentioned that we have to install 1:1 input transformer. But , We forgot to mentioned that we have already installed input transformer 380/440 ( all motors are 440V/60Hz ) , 18 KVA , delta-star.

Best regards from sunshine Croatian Adriatic sea.

Does the VFD give any fault when it doesn't start? I would look there to try to determine wether it's an incoming power issue of if it's an output noise issue. What mode is the VFD running in? I would assume V/Hz since it's driving two motors but it may be worth verifying. If you can get the fault there may be a parameter in the drive which can be tweaked to allow it to start with your DC drive running.

Shawn

Thanks for your comment

After start OSF fault is occured ( line voltage too high ). This voltage is coming from transformer 380/440V

Probably the voltage level is to low.

As JRaef told, the biggest current consumer will take the voltage down.

The VFD will see this as an under voltage fault and will not start.

• An other solution is using three small separation transformers each 200VA
• input voltage 254 volt, output voltage 50 volt.
• Connect the three transformers in star and supply them with 440 volt.
• Then connect each secondary in serie with the input voltage.
• In this case the VFD will see a voltage 254 + 50 Volt = 304 x SQRT 3 volt.

This can be enough to overcome the under voltage problem.

information given without any liability

Zmahmet,

If you have access to a scope-meter, not just a multi-meter, you could look at the incoming waveform. Some multi-meters incorrectly read the voltage on distorted waveforms; others (typically called "True-RMS") do much better. I concur with the essence of others' posts, that the drive is receiving a distorted AC input waveform, and not able to handle it.

For diagnostics, check the input voltages both before and after the line choke you have, with DC drives running and off, and with AC/VFD running and off. This would give you 8 measurements, one for each set of conditions. Even in the absence of a scope-meter or True-RMS meter, this will help to see what is happening.

Here are some possible steps you can take:

• If the transformer has primary taps, change them so the output voltage is higher (assuming that the 440V input you mention is at the lower end of the VFD's input range).
• Try a longer acceleration time for the motors you are driving, if this is possible for the process you are using. Since your post shows that they will run simultaneously with the DC loads, but not start when those loads are on, it sounds like the starting inrush current is too high for the environment you are in.
• Consider using a larger line choke. This will have the effect of smoothing out the waveform at the input of the drive, but will probably reduce the voltage a little.

Do you have separate running overload protection for each motor? This usually has to be provided when a single drive runs more than one motor.

--JMM

Mr. Jmueller

Thanks for your explanation. Now we have information that in few second after start vfd drive fault OSF is displayed. ( line voltage too high ).

Yes, we have protected each motor with overload relay.

????

Ah, the line transient jmueller found issue may be another good clue. If the DC drive is of an older design, it may have an SCR front end, or even worse, a GTO thyristor front end. If so, the thyrisor firing is likely creating sever line notching and subsequent spikes on the incoming feed. That line noise may be resonating with your AC drive's rectifier and caps, creating the condition that your VFD is interpreting as a high line voltage. As jmueller says, a scope would be very beneficial here.

If it doesn't already have it, a specially designed shielded "Drive Isolation Transformer" may be necessary in front of that DC drive. These are almost a standard practice for old thyristor DC drives here in the US, many products are available for it.

http://www.sola-hevi-duty.com/products/transformers/Drive/index.html

I have no idea what you have available to you in Croatia, but I can't imagine this issue doesn't come up everywhere. If you can find someone with a scope to look at your power quality, it might be worth the investment in having them look at your system. If you do indeed have sever transients being created by the DC drive, a shielded drive isolation transformer may be the answer.

Good luck.

Yes. Since this is reported to occur only during starting of the VFD while the DC drive is on, and not if the VFD is started first, I wonder if this is a problem that is just at the margin or threshold of the drive's design performance. If so, considering the location, treating the symptom may still be an acceptable alternative to diagnosing and fixing the source.

--JMM

Hi GW

We recognized your big experience in the field of the AC and DC drives.

No, we dont have AC line filter to suppress AC ringig on the Generators common bus.

Our drilling rig has 9 pcs. of DC Converters -1500A (SIMOREG DC MASTER).

Ahead of each DC converters commutating reactors are installed.

Can you get some "scope" pictures and post them for us?

I am almost ready to bet you have the ringing at the notches. The RC filter is simple and inexpensive.

Yes, we have recorded a few pictures

Take a look

1st. & 3rd. pic. 1DC motor cca 250A

2nd. pic. 1DC motor cca 350A

Oh boy, you've got more problems.

There is far too much noise there to be just DC drives. There should be notches present at the DC drive terminals that go all the way to zero. Everything looks like it is ringing.

The scope you are using may be aliasing. Take a close look at the notches at the peak and set your sweep to get the detail of it.

By the way, what is your scope bandwidth? You should have atleast 10megHz.

GW,

the fluke 123 has a 20MHz bandwidth

Chas

Yes, I also checked the specs for the Fluke 123 (first picture).

There is still just too much distortion in the wave form and the system needs to be carefully analyzed as to where it is coming from and and then systematically clean it up.

It may even be due to existing power factor correcting caps that non-linear loads can cause havoc with.

A proper single line power system needs to be drawn, what loads are attached, and waveforms checked at various nodes.