VFD Output Filtering

[jcdammeyer]

It takes an entire book to explain the details. The electrical noise may come frome one driver but is magnetically coupled and electrically induced into the other lines.
Winding the inductors on the same core reduces the interference.
Search common mode interference

https://www.tdk.com/en/tech-mag/noise/08#:~:text=The%20role%20of%20common%20mode,1394%2C%20HDMI%2C%20and%20DisplayPort.

 

[Dabbler]

You can get stalled on tying to go for the perfect solution here, but I'm not that guy.

How about getting some torroid material and winding some 12ga wire for 4- 6 loops, and then checking the difference with a scope? sort of a 10$ test, and anything is better than nothing.

Really it doesn't matter if the core saturates, as it is the leading edge of the spike that you are modifying. (yes it is better to do it right, but you have to start somewhere)

edit: I'd try 3 of these to start:

https://www.digikey.ca/en/products/detail/fair-rite-products-corp/5980001801/9843117

 

[Susquatch]

It takes an entire book to explain the details. The electrical noise may come frome one driver but is magnetically coupled and electrically induced into the other lines.
Winding the inductors on the same core reduces the interference.
Search common mode interference

https://www.tdk.com/en/tech-mag/noise/08#:~:text=The%20role%20of%20common%20mode,1394%2C%20HDMI%2C%20and%20DisplayPort.

Ah, I see. I apologize for not mentioning that I am using special VFD cable. The wires in this cable are individually shielded from each other with individual grounds to reduce common mode interference. Think a bit like twisted pair but ignore the differential floating voltage. Part of me suspects that cable is the problem LOL. Maybe I didn't terminate it properly. I might try some regular 4 wire cable at some point just for shi*s and giggles.

The scope I used is a 400MHz 4 channel Tektronics 2465B. I did not see any common mode noise at all on any of the three outputs and I was looking for it.

I also have a 2 channel Digital 150MHz 2430A that I could try, but I have very high confidence in what the analog 2465 is telling me.

@Dabbler - you are always pragmatic - almost to a fault. I think I will do just that. It cannot hurt. Besides, I will know immediately if the VFD Output is compromised because I can watch it on the scope.

I'll have to order a few of those cores from Digikey first. No rush on my end.

You are also right about the silliness of chasing perfection. If I didn't have a nice DRO on my mill I prolly wouldn't worry about it. But I've seen what high amplitude noise can do to sensitive equipment. I don't like the thought of damaging the DRO and then in hindsight wishing I had looked after the spikes on the VFD. In this case, I think I'm better off safe than sorry.

 

[YotaBota]

I was looking into this when I installed my VFD and came across these for those who would prefer a not cheap engineered solution. I didn't buy one yet but still thinking about it.

EMC Output Filter for VFD, 6A/20A/80A/150A/300A to 1000A

Good quality EMC output filter for sale, designed for VFD load side to reduce the common-mode noise, protect the load and extend service life, 3-phase VFD output filter with rated voltage 250/440VAC 50/60Hz, current rating from 6 Amps to 1000 Amps, threaded stud or quick connect, high...

www.ato.com

If this was a real issue mechanically I would think TECO would have done something about it.

spikes were typical of short cable runs between the VFD and the motor

Did the tech define "short" cable runs? 1 foot, 2 foot, 6 foot?

 

[jcdammeyer]

Why reduce the noise spikes on the output?

Ask how you might solve the problem of moving the carriage on your lathe to a specific position via the DRO and 1 out of 19 times it's wrong. Not predictable. But it screws up the cut.

The electrical noise not captured by the individual shielded wires that escapes can play havoc with other electronics. Maybe that cut is screwed up when the tool hits the work during an interrupted cut and the load temporarily increases.

My book on electrical interference stresses the importance of removing or reducing the noise at the source rather than trying to stop it where it's picked up somewhere else.

 

[Susquatch]

Did the tech define "short" cable runs? 1 foot, 2 foot, 6 foot?

Yes, he did. Short is anything less than 50ft! LOL!

But he also said that shorter is better not worse.

To your point, he didn't think my spikes were anything to worry about. He took the view that he had never had anyone put a lab grade scope on the wires......

As per what I told @Dabbler, I think he is probably right. I have not had any problems, but I don't like the idea of wishing I had done something about it AFTER I fry my DRO.....

The only reason I even looked at it and noticed it was curiosity.

 

[YotaBota]

@jcdammeyer - I know the harm EMI can do with electronics but @Darren mentions possible damage to the motor and that's more what I was thinking about.

Short is anything less than 50ft!

That's what she said! LOL
The only electronics I'm running is the vfd itself and the tach so EMI isn't a big issue in my use but it's still interesting watch and learn.

 

[thestelster]

These are power my set -ups for the mill and surface grinder. I also have for the lathe. Don't ask me too much cause I haven't a clue how they work. I told the sales rep of the supplier of my VFD's what I had and he had his tech people tell me what I needed. Main power into the safety switch, then into the line reactor, then into the VFD, then into the load reactor, then into the motor.

 

[Susquatch]

These are power my set -ups for the mill and surface grinder. I also have for the lathe. Don't ask me too much cause I haven't a clue how they work. I told the sales rep of the supplier of my VFD's what I had and he had his tech people tell me what I needed. Main power into the safety switch, then into the line reactor, then into the VFD, then into the load reactor, then into the motor.

Don't know why I didn't notice all that when I was at your place. But very pro looking.

 

[thestelster]

Don't know why I didn't notice all that when I was at your place. But very pro looking.

Ha....I must of scrapped 50' of conduit trying to learn how to bend that stuff!

 

[Susquatch]

Ha....I must of scrapped 50' of conduit trying to learn how to bend that stuff!

They sell a really nice manual pipe/tube bender at Home Depot. I don't have one but my neighbour does. He borrows enough of my stuff that I never hesitate to borrow his!

 

[thestelster]

They sell a really nice manual pipe/tube bender at Home Depot. I don't have one but my neighbour does. He borrows enough of my stuff that I never hesitate to borrow his!

I have both 3/8" and 1/2" EMT benders, but, it's figuring out were to put the bends/off-sets and by how much. I finally found an App which helped immensely, but of course after almost all was done.

 

[jcdammeyer]

I was loaned a 1/3HP 3 phase motor to test the Baldor 3 phase drives I picked up. First one worked nicely. A bit of whine up in the higher speeds. But with the pot I can make it turn 60 RPM and it appears to have decent torque.



I'll do some testing and see what sort of electrical noise it generates and then make up some inductors to see if I can suppress it a bit.

 

[Susquatch]

I'll do some testing and see what sort of electrical noise it generates and then make up some inductors to see if I can suppress it a bit.

Based on all the stuff I see you do, I know you must have some very good instrumentation. Just wondering what kind of scope you have to do this with? Which reminds me that I have to get some better scope probes for this kind of work......

Also wondering if you have ever tried one of those computer enabled USB digital scopes that have no screen of their own. Hantek seems to be quite popular these days and are even available on Amazon at decent prices.

Products - Hantek Electronic & Your testing solution provider

 

[Susquatch]

My book on electrical interference stresses the importance of removing or reducing the noise at the source rather than trying to stop it where it's picked up somewhere else.

I don't know how I missed this post, but I did. Sorry about that.

I agree completely. It's not always possible though - especially when the source is out of your control - for example the guy arc welding next door..... LOL

It's easy to forget that though, and also easy to fall into the trap of dealing with symptoms instead of cause.

Good reminder for me and prolly others too John.

 

[jcdammeyer]

I have one of these that I bought specifically to be able to decode the alternating bit levels of RS232 and translate into the communication bytes. It allowed me to see the timing differences between specific characters because a Pan & Tilt system wasn't talking 'friendly'.

Analog Discovery 2: 100MS/s USB Oscilloscope, Logic Analyzer and Variable Power Supply

Buy Analog Discovery 2 USB oscilloscope, logic analyzer, and multi-function device that allows to measure, visualize and generate mixed-signal circuits.

digilent.com

I also have the BNC breakout board for it along with scope probes and the adapter for the digital side of things.
Essentially it's an "Oscilloscope, Waveform Generator, Power Supply, Voltmeter, Data Logger, Logic Analyzer, Pattern Generator, Static I/O, Spectrum Analyzer, Network Analyzer, Impedance Analyzer, and Protocol Analyzer". I like it but haven't used it in several years.

The reason for buying it was it was half the price of the serial decoder module for my Tektronics MSO3034 scope which is 4 channel, 300 MHz with 16 logic analyzer channels and has the 2 modules for decoding Automotive CAN bus, and also SPI, I2C. Those modules are $1200 each hence the USB type scope for a PC. This scope has been discontinued.

https://ca-en.alliedelec.com/product/tektronix/mso3034/70136946/

I also have two pieces of Agilent equipment. A 6.5 digit meter, the 34410A which is also now discontinued.

34410A Digital Multimeter, 6½ Digit [Obsolete]

The 34410A DMM will be discontinued December 1, 2016 and is replaced by next-generation 34465A Truevolt Series DMM.

www.keysight.com

And the Agilent 33220A Function and Arbitrary Waveform Generator also discontinued.

33220A Function / Arbitrary Waveform Generator, 20 MHz [Obsolete]

The 33220A function / arbitrary waveform generator will be discontinued December 1, 2016 and is replaced by next-generation 33500 Series Trueform waveform generators.

www.keysight.com

For RF work I have a R&S FSH3 portable spectrum analyzer also discontinued along with the VSWR Bridge for it.

https://www.aaatesters.com/pub/media/datasheets/rohde__schwarz_fsh3133_specifications_spec_sheet_9s48.pdf

So all my stuff other than the Digilent is old. Like me. But it still works. I see on EBAY the FSH3 sells for as low as $895. I believe I paid $15K for the kit including the carry bag.

I hesitate to dive in and figure out what it would cost to update all of the above. Likely the price of a small car.

 

[John 345]

It's related, but he is filtering the input not the output. I've already done that.

Not sure if this comment in the right place as it's my first.I started playing with vfd a few years ago ,had noise problem affecting my blue tooth sound system so I just use shielded cable on vfd output to lathe motor have not checked with any fancy gear but solved my noise problem apparently as blue tooth NC headphones work fine . Can't explain technically the reasoning though.

 

[Former Member]

@jcdammeyer I have 2 VNA, HP8753C calibrated recently to 3GHz and a AEA Bravo2500 handheld, which also has Spectrum Analyzer and FDR built in. Difference between the 2 sensitivity the HP while significantly older is still at the top end. My main Spectrum Analyzer is a HP141T (18Ghz).

 

[jcdammeyer]

@Degen There's nothing wrong with the old stuff. Although my FSH3 was originally bought to work with the WORMs I also used it for my CANRF modules. Because my FSH3 has extra code features also for short circuit detection in long cables. That came in handy for finding 'mistakes' in the cables for the YVR Olympic Rings.

 

[jcdammeyer]

Not sure if this comment in the right place as it's my first.I started playing with vfd a few years ago ,had noise problem affecting my blue tooth sound system so I just use shielded cable on vfd output to lathe motor have not checked with any fancy gear but solved my noise problem apparently as blue tooth NC headphones work fine . Can't explain technically the reasoning though.

Sure it's the right place. Often just the right type of shielding will prevent the noise from going too far if at all.