A simple and effective VFD EMI filter

[brino]

Petros,

Thanks for the additional information.
I appreciate your logical, systematic approach to problem solving.

I really enjoyed this thread!

Thanks again,
-brino

 

[metric_taper]

I suspect that the better final results were related to the capacitors the filtered socket has: 1 capacitor 0.1 uF and 2 caps 2200pF
View attachment 349247
What puzzles me is the fact that all filters I had tested initialy were giving very poor results. To be honest I was expecting to solve the problem once for good using the filtre ftom the UPS unit!

It seems that the combination of the 3 ferrite toroids was just right for the high frequencies produced by the VFD.

These current spike noises from the VFD look like broad band noise, as the edge is fast, so the Fourier series has lots of high frequency harmonics, and in your case in the wi-fi range. I didn't see if you have a DSL, or Cable modem as your internet source. The problem with those Pi filters used in commercial computers, they require a very good ground connection. You don't have that if a ground wire is your return path to earth. In a home, it is a long path, that looks like an antenna. These EMI filters are there to pass the FCC emissions part 15 tests. Your VFD upset may be from noise on the ground, injected via the VFD power connection, as the VFD will have an EMI filter needed to pass emissions testing requirements from CE or FCC compliance standards. Your photo of the EMI (apple Macintosh box) shows you having the ground with a ferrite filter. Did you do just a test on the ground path before this final packaging of the filter?

 

[petcnc]

Your photo of the EMI (apple Macintosh box) shows you having the ground with a ferrite filter. Did you do just a test on the ground path before this final packaging of the filter?

Nope!
I just followed common sense as most of the computer PSUs I have opened to test or fix have this ferrite to the ground path I thought that it must clear some noise!
Thanks for the thorough explanation/analysis!

 

[petcnc]

Petros,

Thanks for the additional information.
I appreciate your logical, systematic approach to problem solving.

I really enjoyed this thread!

Thanks again,
-brino

Thank you for the kind words brino! I' m glad you liked it.
The whole idea of publishing our projects here is exactly that! To be useful and interesting for the amateur machinists.
If they enjoy it this is an extra!
Thank you again

 

[markba633csi]

Great detective work Petros! Nice little scope you have there too- I always felt the local Rent-A-Tool shop should offer them for loan
Most folks don't have enough need to justify purchase, but they are getting more affordable with LCD displays instead of CRTs
-Mark

 

[machPete99]

They recommend using shielded cable going from the VFD to the motor, and the shield should be grounded, for obvious reasons.
There is some specially rated cable for this purpose:

McMaster-Carr

McMaster-Carr is the complete source for your plant with over 595,000 products. 98% of products ordered ship from stock and deliver same or next day.

www.mcmaster.com

 

[petcnc]

They recommend using shielded cable going from the VFD to the motor, and the shield should be grounded, for obvious reasons.

The problem proved to be not to the output of the VFD, where usually the wires act as aerials and propagate the noise, but the noise signal was travelling "backwards" through the power line to the rest of the house causing interference problems to other appliances.

 

[ericc]

Does the Internet at your home work now? It's funny: those commercial filters look like they contain toroids. Maybe they were too wimpy?

 

[petcnc]

Does the Internet at your home work now? It's funny: those commercial filters look like they contain toroids. Maybe they were too wimpy?

Yes it works like a charm! No problems when I use the VFD!

 

[DeltaWye1732]

Yes it works like a charm! No problems when I use the VFD!

Thanks for your effort of documenting the solution that worked here. I've always been curious if a isolation transformer on the input side of the VFD would work to alleviate these kind of interference problems? The isolation transformer would prevent a current path between the secondary and primary, and any current that wanted to get back up the supply wires from the VFD would be faced with the huge inductive reactance of the transformer. Would this work? I have seen reference to this being used in industrial settings too. I would post them but can't right now as a new member.

Would be interesting to hear people's thoughts and experiences regarding isolation transformers on the input side of VFDs. It wouldn't take a huge transformer for most home shop settings, given most sinlge phase input VFD's are in the 3 HP and below range. So a 3 kVA transformer for example would do nicely.
Sorry if this is an old thread, but I couldn't find anything else and think it would benefit future readers to see the transformer option mentioned here.