VFD & Noise

[C3PO]

Here is the situation: I have a piece of test equipment that sends a pulse of oil into an oil filter (for example 0 to 100Psi) This pressure is measured by an o scope that is hooked to a transducer. My problem is that I need to put a VFD on the motor on this piece of equipment. The VFD really messes with the o scope and causes it not to give me a valid pressure reading. I have put both line and load reactors on it and it didn't help much. Does anyone have any ideas on how to keep the scope from picking up this noise???

 

[ohmhead]

Here is the situation: I have a piece of test equipment that sends a pulse of oil into an oil filter (for example 0 to 100Psi) This pressure is measured by an o scope that is hooked to a transducer. My problem is that I need to put a VFD on the motor on this piece of equipment. The VFD really messes with the o scope and causes it not to give me a valid pressure reading. I have put both line and load reactors on it and it didn't help much. Does anyone have any ideas on how to keep the scope from picking up this noise???

Well now iam no expert on vfd s but lets look at noise to o scope when signal comes out of transducer have you tried a snubber circuit .


Your line reactors incoming and out going are they solid core type or air ?

Maybe O cable is picking up noise shield it or shorter cable to O scope ?

Have you tried blocking or i should say ferrite toriods in vfd on cables and is this maybe already a factory thing inside vfd ?

A high band pass hf filter to o scope with a isolation transformer are you at 10 volts or 5 volts signal ?

Line reactors dont limit all there are better but it cost lots of dollars there is electronic filters today .

Go to Digit Key web site punch in VFD filter tell me what you think ?

 

[Besoeker]

The VFD really messes with the o scope and causes it not to give me a valid pressure reading.

o scope?
Oscilloscope?

 

[ohmhead]

o scope?
Oscilloscope?

Well i think its a Oscilloscope he talking about Besoeker i was thinking to eliminate the trash to the scope input from what ever he is doing .

Even maybe a isolation circuit like a photo infared transmitter circuit to isolate between or maybe his O scope cable is grounded on both ends picking up trash from ground loop .

What would you do ?

 

[76nemo]

"Have you tried blocking or i should say ferrite toriods in vfd on cables and is this maybe already a factory thing inside vfd ?" ~Ohmhead



Good question. Are there any there?

 

[Besoeker]

Well i think its a Oscilloscope he talking about Besoeker i was thinking to eliminate the trash to the scope input from what ever he is doing .

Even maybe a isolation circuit like a photo infared transmitter circuit to isolate between or maybe his O scope cable is grounded on both ends picking up trash from ground loop .

What would you do ?

Well, I routinely an oscilloscope to make measurements on drives and other power electronic equipment. I have posted some on this forum. I think the only time I have seen significant interference was in the old days of cathode ray tubes and very high current rectifiers - physically placing the oscilloscope in vicinity of a bar carrying 10kA resulted in the waveform being distorted.

Maybe the problem is the transducer rather than the 'scope?

 

[ohmhead]

Well, I routinely an oscilloscope to make measurements on drives and other power electronic equipment. I have posted some on this forum. I think the only time I have seen significant interference was in the old days of cathode ray tubes and very high current rectifiers - physically placing the oscilloscope in vicinity of a bar carrying 10kA resulted in the waveform being distorted.

Maybe the problem is the transducer rather than the 'scope?

Well i guess the kit era is over but ive got a 1012u oscilloscope which i built in 1968 and it still works and yes you are correct today there much better to use but we didnt have harmonics back then to screw up the tube .

HEATHKITS were big then .

 

[Besoeker]

Well i guess the kit era is over but ive got a 1012u oscilloscope which i built in 1968 and it still works and yes you are correct today there much better to use but we didnt have harmonics back then to screw up the tube .

HEATHKITS were big then .

Yes, I'm old enough to remember them too!
Most of our guys now use a Tektronix TDS 2012 which is convenient to carry around and has a decent performance. I routinely carry one around in my box of kit in the car boot (trunk) I also have an old TDS 360. I like it because it has a built in 3.5 inch drive and you can store several hundred waveforms in *.csv format on one disk. And you don't have to connect your PC to the scope - one of lads has managed to fry three laptops doing just that
On the down side, floppy disks are not so common these days.

 

[gar]

100117-1112 EST

1. Short the scope input at the scope. Is there any noise observed?

2. What kind of pressure transducer are you using? Straingage, quartz, or what?

3. What signal conditioning is on the output of the transducer?

4. How and where is shielding provided and connected?

5. If the signal conditioning is connected to the scope and the input to this shorted is the noise observed?

6. More questions after you provide answers to the above questions.

.

 

[StephenSDH]

• Use VFD Cable (3 twisted conductors, sheilded) between the drive and the motor. Generally ground at both the motor and enclosure.
• Use a 5% choke to reduce the inrush.
• Use vfd cable from the power source to the vfd.
• Keep the wires away from each other.
• If you have an external brake, use shielded wire also.

http://www.omnicable.com/site/catalog.php?id=372

 

[Jraef]

• Use VFD Cable (3 twisted conductors, sheilded) between the drive and the motor. Generally ground at both the motor and enclosure.
  ...

http://www.omnicable.com/site/catalog.php?id=372

I think this is an important point. The most common problem I see is in situations like this is people using "portable cord" on the output of the VFD going to the motor, or they use "Sealtite Flex" conduit that is the non-metallic type, then pull regular stranded wire inside. Having no shielding on the output cables turns them into nice little FM transmitters. Filtering on the output will help, but the best thing is to use shielded cable or use metal conduit, either one with good grounding.

 

[ELA]

As has already been asked , what is the output impedance of the transducer you are using? Is the output 0-5v, 0-10V , a voltage output? And if so how is it terminated? Hopefully not as this is probably a high impedance output and will be more susceptible to the VFD noise.
4-20ma would be better as it is much lower output impedance.

The O'scope is a very high input impedance and the output impedance of the transducer is in parallel with it. The higher the impedance the more susceptible to noise it will be.

You can address the source of the noise by shielding at the VFD which is good but if the monitoring transducer impedance it too high you may not suceed without reducing the impedance at that end.

An open circuited O'scope lead picks up all kinds of noise ... way too easily.

 

[C3PO]

I appreciate everyones replies.

The transducer output is 0-10v

The cable from the scope to the transducer is sheilded with the shield connected at the scope end.

The transducer is made by validyne, I suppose it could be the tranducer picking up the noise but we have some transducers on other equipment ran by drives and there isn't any noise on them. The difference on them is that they dont use an o scope it just goes to a display.

The wires to the motor are stranded THHN in nonmetallic sealtite.

The scope is a new Techtronix brand.


I will try to change the leads to twist cable and see what that does.

I am going to have limited computer access for the next few days so bear with me if it takes me a little bit to reply back. Thanks again!!

 

[gar]

100117-1459 EST

C3PO:

If the transducer output is actually 0 to 10 V, then there is signal conditioning somewhere in the transducer. This means there is power to the transducer. I will guess it is a straingage type.

This might be supplied with +/-12 and a common for reference. The 24 V could be the excitation to the bridge transducer followed by a differential amplifier referenced to the common wire. Also means a signal line. Possibly a total of 4 wires plus shield. This amplifier would have a low output impedance, there would be substantial shunt filtering capacitance on both + and - supplies to the common line. A lower supply voltage might be used, but you need a way to get the 0 to 10 V swing. It could be an unbalanced supply with the differential amplifier shifting the output to the 0 to 10 range.

Very likely the transducer bridge is balanced at about 0 pressure, but not necessary. Is there some additional wiring for bridge balancing?

I would physically disconnect the transducer from the pressure piping, but leave in the same physical position and insulated from the piping and connect the transducer housing to the shield and see if the noise is still present. This would be only after shorting the scope input at the scope to determine if noise was getting into the scope by some other path.

Provide more information on the transducer.

.

 

[gar]

100117-1523 EST

Another point. Belden 8723 probably would be a good wire to the transducer if 2 twisted pairs and the pairs individually shielded would meet the wiring requirements. A twisted pair for the signal wires will reduce magnetic noise induction.

.

 

[ohmhead]

Well just a small add to all the good post but years ago we use to disconnect the ground at the oscilloscope supply this had a effect on any ground noise from any neutral on the system or back feed it kinda would sometimes clear a problem .

And only ground the cable on one side of that transducer another ground or double ground loop can make or break your input less of a path for current to flow .

Gar good post the old 741 op amp circuit ?

 

[ELA]

The transducer output is 0-10v

The transducer is made by validyne,
The difference on them is that they dont use an o scope it just goes to a display.

So this particular transducer is not terminated by anything but the O'scope whereas the others are terminated in a display?
Read the specification on the transducer to see what its maximum output current is? Probably not very great as most 0-10V input modules are high input impedance.
Try terminating the transducer output in as small of a resistance as the specifications allow.

 

[Jraef]

...

The wires to the motor are stranded THHN in nonmetallic sealtite.
...

Ding ding ding... I think we have a winner!

Either use metallic sealtite with a good ground connection, or pull shielded 'VFD Cable" in it instead of the plain THHN.

I once worked at a project in Nevada for a concrete mixing / pouring machine for underground nuke test sites (the ones we don't have...). Because the machine was rotating around the bole hole, they used unshielded SO cord to hook up the motors to the VFDs. Test scientists triangulated on us from 15 miles away because we were screwing up their instrument readings. Put the leads in metallic sealtite and the problems went away.

 

[gar]

100117-1636 EST

ELA:

The transducer almost certainly has internal amplification and thus I expect an AC small signal output impedance is very close to zero from the negative feedback of the amplifier.

There are limitations by the available drive current. The AD524 can drive a 2 k load to +/-10 V.

The wiring from the amplifier may have a higher impedance than the amplifier itself.

If internally the output impedance was 100 ohms with no feed back, the gain was 1,000,000, and 100% negative feedback, then output impedance is about 0.000,1 ohm. A lower gain amplifier and it is about 0.001 ohm.

See following references:
http://www.electro-tech-online.com/general-electronics-chat/35701-op-amp-cct-output-impedance.html
http://zebu.uoregon.edu/~rayfrey/431/notes9.pdf

.

 

[StephenSDH]

underground nuke test sites (the ones we don't have...)

If this is your last post, I think I know why.