Puzzling electrical readings taken at VFD bypass contactors 1

[pbcd]

I have a question concerning an electrical test reading taken at the VFD bypass contactors. This VFD provides power to a 15 HP motor for an air handler unit. To check resistance across my contacts on my contactors I take voltage readings under load for each phase (from L1 to T1, L2 to T2 and L3 to T3). Readings are checked at the 3 contactors (bypass, VFD input & VFD output). I also do the same check on my overload which is in parallel to the load side of both the bypass and VFD output contactors.

In the VFD bypass mode, I get zero volts from line to load for each phase of the bypass contactor. The readings for each phase at the overload is also zero volts. In the VFD mode, I get zero volts from line to load for each phase of the VFD input contactor. At this point all readings taken are normal, but when I check the VFD output contactor I am puzzled by my readings. All readings at the VFD output contactor are 2.1 volts for all phases. The readings on the overload are also 2.1 volts. All six reading are exactly the same down to a tenth of a volt. When I raise the VFD speed to 60 Hz, all readings rise to 3.2 volts.

I assume there is a logical theory for these readings and would appreciate it if someone could shed some light on this.

 

[itsmoked]

Yes. The signal put out by the VFD is not one your meter can accurately handle. It is a very, VERY, complex waveform that you are measuring with a meter that likely expects a sinewave.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[pbcd]

Thank you for your response ......... very interesting.

 

[itsmoked]

For your situation I would measure the contacts and if they all read about the same voltage you're good to go. If one was very different then you could surmise a contact problem. Another words the delta not the absolute readings. (VFD output of course)

Suggestion: Carry an invaluable temperature gun. Follow up your across-the-contact measurements with a thermal reading. Several volts at the amperages you are working with would result in rapid elevated temperatures at the contacts. 'Gun' them as close as you can to the contacts and you will see if the electrical reading is true or not.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

To elaborate Keith's discussion.

There are high-frequency components in the current and they produce an inductive voltage drop that your voltmeter tries to interprete. A few MHz and one or two microhenries of inductance can easily produce a couple of volts.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[gepman]

Fluke states that their 87V will work on VFDs (for voltage measurements) and supposedly Allen-Bradley backs up that statement. It has a switchable low pass filter. Fluke also has a good application note on VFD measurements. See

http://us.fluke.com/usen/support/ap...ukeProducts)&parent=APP_NOTES(FlukeProducts)#

(the # sign is part of the URL), Case Study: Allen Bradley and the Fluke 87V DMM AND Electrical Measurements on Adjustable Speed Drives.

I am looking for a good Power Meter which will measure VFD outputs. All the meters say they are True RMS but don't say much about the crest factor or the frequency they can handle. Eventually you may want to by an oscilliscope.

 

[pbcd]

Thanks for taking the time to reply. Learning things everyday!

 

[pbcd]

Would the VFD output voltage reading taken across the phases be accurate with a VOM?

 

[dpc]

Not with digitial multimeter. An old analog Simpson meter might give you something somewhat proportional to rms voltage because of the averaging nature of the meter movement.

 

[Skogsgurra]

The analog Simpson doesn't produce a good RMS reading either. It was calibrated in RMS using a sinewave. Any other waveform will be wrong on the Simpson (and all other d'Arsonval movement meters with a simple rectifier).

It is only moving iron meters, thermal meters, electrostatic meters and electrodynamic meters that show TRMS*. And the problem with them is that they show RMS of the motor voltage. And that is not what you want to know, usually. What you want to know is the RMS of the fundamental frequency.

You need a spectrum analyzer or a selective voltmeter to take such a measurement. Or simply read the inverter's output voltage from its display. They are accurate.

*Modern TRMS meters also show RMS. No doubt about that. But, some cannot take the EMI around an inverter and some do not have enough band-width.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[controlsdude]

An analog Simpson?! The last time I used or saw one of those was in the Navy back in the 80's and by this maintenance guy who was very sharp. Do they actually still make analog Simpson? I think your showing your age Skogsgurra.

 

[DickDV]

Skogs and I must have gone to the same school at about the same time since I still prefer to have an analog (Simpson 260 or equiv) meter in my bag.

The general rule about drive output leads is "don't try to measure anything on them". Let the drive tell you what's going on out there and use the drive outputs instead.

This is due to the same reasons given above---few field instruments can give you accurate readings. And, there is a hazard issue that some ignore, as well. 460V drives produce 660V peak pulses. Due to ringing and reflected waves, these pulses can exceed 1300V especially at the motor end. I'll bet your meter isn't rated for those voltage levels.

 

[itsmoked]

Like I say. You want to know about contacts in the drive's output? Use a temperature gun.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

I wish people really read the posts.

I said nothing about using a "Simp" other than it won't do the job.

But, OK, dude. I have been along quite a time. You are right there!

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[jraef]

And yes, Simpson analog meters are still being made.

Fluke and AEMC have several very good digital meters now that are specifically designed to read VFD outputs. The trouble is, they are also very good at draining budgets as well. Last year I was offered one as payment for a consulting job I did, I should have jumped on it but the wife and kids insisted on eating that month, you know how that is.

Santa?... I have been a very very good boy this year...

 

[ScumPunk]

Hi, all,
The Fluke 87V meter is an excellent bit of kit for measuring the VFD outputs. Not only is it accurate (i.e. it agrees with the VFD's we've tried it with, and other much more expensive analysers) - it's catIV and rated to handle the voltages mentioned by DickDV. Only snag is that it goes to manual ranging when you switch the filter in - and starts at 6V range.
Jraef - Farnell currently have them on offer for £265 - £40 off normal price.

 

[ScumPunk]

Sorry, really must read posts more carefully...
Itsmoked was complaining about DMM prices! Sorry all.

 

[startermotor]

Would it be possible to get a reading if you made a filter box for your meter and flitered the signal.
I use something like this for working on RF that consists of a box five inputs two in two out and an earth in the middle.
I connected a series of 1kv mica type capacitors 0.001pf to each terminal.

This stops my meter from going nuts and is cheap.

 

[Skogsgurra]

A very useful filter is one that has a 1.00 Hz corner frequency. It will show the V/Hz ratio when connected to a VFD output.

Use two resistors, one in each leg, and a capacitor between the resistors. Selact resistors and capacitor so that 2xRxC = 1/6.28 (i.e. w = 2 times 3.14 times f). Make sure that the resistors can take the power developed and that meter loading (1 or 10 Mohm, usually) doesn't load the filter too much. Capacitor voltage is not a problem since it will usually stay in the below 10 V region.

Such a filter will show V/Hz when the inverter is running. Example: a 460 V at 60 Hz output will show as 7.67 V, regardless of inverter frequancy. A good check to find if the inverter is doing what it is supposed to do. At least voltage-wise.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Me no complain about DMM price...

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com