EMC issues with VFDs 6

[Marke]

I have recently been involved in a number of issues where stray voltages and currents from VFDs have been causing major issues on site.
The issues have been shocks to cows in a dairy shed resulting in severely reduced production, elevated cell count and major mastitis in the herd, major desensitization of ear tag readers in a cow shed, PLC lockups, damaged electromagnetic flow meters plus other problems.
The major cause of these problems is incorrectly installed VFDs. In my research for facts, I have been unable to find actual numbers that show the importance of the correct installation, so I have done my own tests and documented them at

http://www.vfd-emc.co.nz/vfd-emc-tests.php

These tests are ongoing and I will add to the results as they become available and time permits.
In each of the problem sites, I have easily fixed the problems by implementing the installation practices described which generally appear in almost every installation manual.
I have test curves available on request.

Best regards,
Mark

Mark Empson
Advanced Motor Control Ltd

 

[Skogsgurra]

Very good, Mark! Interesting reading. That problem has been addressed by DeLaval. The following may be interesting to a broad audience:

2010-10-25
NFO Drives signs long-term agreement for NFO Sinus

NFO Drives has signed a long-term collaboration agreement with the
global DeLaval company for annual deliveries of NFO Sinus frequency
inverters. The agreement runs until 2016. The inverter will be included as a part of DeLaval’s existing product range. Series production will start at New Year 2010/11, and the order value for the first project is estimated to exceed 10 MSEK per year at full volume.

DeLaval is a full service supplier to milk farmers. DeLaval develops, manufactures and distributes equipment and complete systems for milk production and animal husbandry. Service, sale of a wide range of accessories, knowledge exchange and advice are also important aspects of the company’s activities.

DeLaval, a company in the Tetra Laval group, is active in over 100 markets and supplies to customers with livestock herds that vary in size from 1 to 50,000 animals.

“DeLaval has a clear policy on protecting the environment and grain health as well as good milk quality,” says Tord Ringenhall, product manager at DeLaval. “Both grain and technical equipment are sensitive to electrical emissions, for example leakage currents and electromagnetic disturbances. At DeLaval we therefore want
our products to produce the cleanest electric environment possible in animal stables”.

“We have now tested the NFO Sinus frequency inverter over a long time both under laboratory and stable conditions,” says Tord Ringenhall. “We have ascertained that NFO Sinus has unique properties, which make it entirely interference free and therefore appropriate in DeLaval’s product strategy. We are now envisaging different applications in which we can benefit from this interference free technology”.

I think that you have had some contact with them already. But not sure if you knew about the latest development.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[dpc]

Mark,

Thanks for posting that link and your data. The effective of the shielded cables is impressive. As you are probably aware, the issue of stray voltages in dairies has been around for many years. In the US, the REA (now RUS)did a lot of work on this and there is a lot of reference material available - but probably not directly related to VFDs.

A good compilation of general links is here:

http://strayvoltage.epri.com/resources.asp

Regards,

dpc


David Castor

http://www.cvoes.com

 

[itsmoked]

Thanks for the info Mark E!



Off topic:

Gunnar; I have a hard time getting past the 'nasal passages' drive name.

Google 'sinus'. The first forty million hits won't have anything to do with sine waves.. They will all be about, discomfort, suffering, mucus, and drugs.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[dpc]

For my machines class I had a German professor. He had a pretty heavy accent, and still used "sinus" and "cosinus". After the first couple of lectures, we quit thinking about it. He was a brilliant guy.

David Castor

http://www.cvoes.com

 

[wolf39]

Keith:

We in Germany (and possibly in Scandinavia, too) say

sinus 60 degrees = 0.866

How do the Americans say? Also, if you talk about sinus shaped voltage curves, how do you express yourself in this respect?

Thanks for the information.

Wolf

 

[Skogsgurra]

They use the not so correct word 'sine'. I have even seen 'sign(30)'.

Little do they know about the Latin origin where 'sinus' means 'bay' and was used to describe the 'bay' formed by the sides of a triangle.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[jraef]

We know it, but we choose not to use terms that can be thought of as anatomical references.

Of wait, we use the term peckerhead too... there goes that theory!

"Dear future generations: Please accept our apologies. We were rolling drunk on petroleum."
— Kilgore Trout (via Kurt Vonnegut)
For the best use of Eng-Tips, please click here -> faq731-376

 

[electricpete]

fwiw, I am more used to seeing sin than sine.

Great writeup by Mark. I am a little bit curious about the measurement setup since it seems like a small challenge to avoid interference on the leads in a high frequency measurement of locations far apart (motor and drive frame). Of course that interference would still be related to the thing we want to understand...


=====================================
(2B)+(2B)' ?

 

[waross]

Hi Pete. I understand that "sin." is an abbreviation for "sine". I also understand that the period (full stop) although once mandatory following abbreviations is becoming optional in several style guides.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[Marke]

Hello Gunnar
Yes, I am certainly aware of NFO and have tried to get them into some problem areas here, but too many asian drives at super cheap prices and people anaware of the issues makes for problems.

David
yes, the stray voltage issues are known, but try to find references covering stray voltages coming from VFDs in dairies. In NZ, we use primarily rotary sheds where the contact to the cow is around the whole body at one common potential (platform) and across the rear (shed) and so the body resistance is very low and the sensitivity is far higher than all the USA based testing that I have seen.

Pete
The 470 ohm termination resister is at the scope end to try to soak up the interference.
The measurements are designed to be relative rather than absolute, to illustrate the effects of the different types of termination etc. Any interference will be a constant and add to all results, so if anything, the ratios would get worse!!

We use Sine and Cosine here in New Zealand, or we did when I went to school.
Best regards,
Mark.

Mark Empson
Advanced Motor Control Ltd

 

[Skogsgurra]

We use that, too, Jeff. But we learned from you!

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[xj25]

Thanks Marke, actual measurements are always welcomed

btw anyone knows what actually is the "sinus switch" pantented by NFO¿?
i couldn´t find any clue about it

regards

 

[Skogsgurra]

I did a writeup more than ten years ago. It is still valid.

http://gke.org/pub/files/The Switch Circuit.pdf

It is for 'educated laymen' so do not expect much math. There's none, I think.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[xj25]

Mark, could you sketch the setup for measuring the relative voltage and how the 470ohms resistor is connected?

"The voltage between the frame of the motor and the frame of the reference VFD are measured using a 100MHz oscilloscope with a 470 ohm termination resistor."

regards

 

[Marke]

Hello xj25

I will try to draw out a diagram and pperhaps a few photos when time permits, but it is really quite easy and does not require a circcuit.
1. Use an isolated 100MHz oscilloscope with a film 470ohm resister across the input connection. (BNC - banana plug adapter)
2. Connect flexible cables between the input terminals to the frame of the motor, and the frame of the VFD.
That is all there is to it.

Best regards,
Mark.


Mark Empson
Advanced Motor Control Ltd

 

[GGOSS]

Great work Mark.

 

[sibeen]

Skogsgurra, I had a read of your PDF and then went and had a look at the patent application. The patent doesn't really say that the output of the inverter is a sine wave but appears to deal more with ways of firing an inverter which avoids catastrophic misfiring.

Is it just the use of the LC filter in the output that causes a sinusoidal output, or am I missing something in my reading?

 

[Skogsgurra]

Not many people read patent applications. And fewer get anything out of that reading. You obviously did. And were surprised that nothing is being said about sine waves. The inventor never thought about sine waves. I more or less had to convince him that they are good.

The sinewave is a by-product of the Switch Circuit. The Switch Circuit was designed to avoid hard switching and also to avoid closing the flywheel diodes by driving reverse current through them.

Ragnar Jonsson (the inventor) never thought about the benefits with the sinewaves. We had a rather tough discussion where he meant that the crisp and good torque control that the NFO algorithm offers was the main thing with the drive. It was only when I told him that fans do not need servo properies but it is nice if they do not emit EMI that he reluctantly started to accept that the real value with the Swich Circuit *could* be the interference-free sine wave.

The LC components are not a classical filter where you input a PWM wave and get a sinelike waveform out. The LC components and the switching co-operate to accurately reproduce the signal that is fed to the control amplifier. So, you can produce complex waveforms (useful when you need a power generator with arbitrary waveform - lots of harmonics, for instance - for testing). That is not possible with a standard filter. A standard filter also hinders the use of vector control.

The switching frequency can be high, due to the lossless switching, and that keeps the LC components small. At low load, the carrier frequency gets close to 200 kHz. Ragnar is now working on a modular design with modules in the 10 - 15 kW range so that you can have a 'row' of modules that produce 40 - 60 kW and a 'square' (four 'rows') that produces up to 240 kW while a 'cube' with four 'squares' will produce 960 kW. More 'squares' can be added for higher output.

The beauty with this is that all power modules are optimised for cost and use very readily available semiconductors. That makes the production of large inverters an easy task and spare parts are the same for all sizes.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[sibeen]

Thanks for that explanation, Skogsgurra. After reading the patent app I was worried that my "stupidity quotient" was somehow set higher than normal. You've set my mind at rest

I've never really worked with VFDs but worked in the UPS industry for 20 odd years or so, so novel inverter systems are always fun to look at. It sounds as if the inventor is going down the UPS path with a modular system. I think Chloride (Now Emerson) go up to 9.6 MW with their stackable UPS configuration.