Could you me to deal with noise due to a electric motor?

[wzhit]

I am using an electric motor to impose displacements on a specimen,
but when the motor is on, there is too much noise in the displacement
and force signal.

Everything seems to be nice when the motor is off. The frequencies
of the noise are less than 10Hz, too low to be filtered. In addition,
I am sure that each cable is effectively shielded with the same ground cable.

Does a motor, in my case, which is a Permanent Magnet Synchronous
Motor (AC), always introduce noise like that? How can I reduce it?

I will greatly appreciate your help.

Grazie@@

 

[Bob S]

If the motor is a permanent magnet AC motor, what is the AC frequency, and what is the motor RPM? If the RPM is over 600 RPM, how could it make noise less than 10 Hz?
Bob S

 

[wzhit]

If the motor is a permanent magnet AC motor, what is the AC frequency, and what is the motor RPM? If the RPM is over 600 RPM, how could it make noise less than 10 Hz?
Bob S

Thanks for your reply.

AC=50Hz,
RPM=3000

I know less about the mechanics of the PMSM, and can you explain more why there should be not any noise less than 10Hz?

 

[Carl Pugh]

Could the problem be vibration from the motor?

Does tapping the specimen cause noise?

Does placing an electric motor on the specimen cause noise?

 

[wzhit]

Could the problem be vibration from the motor?

Does tapping the specimen cause noise?

Does placing an electric motor on the specimen cause noise?

Good questions! I forgot to mention this.

First, it is not the vibration of the specimen.
When the load cell is put on the ground and is connected to the motor via a cable,
the noise is there; if disconnecting the cable, the noise disappears.
It clearly shows the noise is due to the EM field of the motor.

Now I use the same plug-in unit for
the DC power supply, motor drive, data acquisition device.
Of course, the other cable supplies the voltage of 400V for the motor.
I think both of them share one ground cable.
In addition, the ground cable from the plug-in unit is also used
as the shielding cable for the load cell.

Is this connection correct?

Many thanks to Carl Pugh.

 

[Carl Pugh]

It is best to have separate ground cables.
Sometimes it works best to ground things and other times the opposite is true.
Operating sensitive equipment with isolation transformers sometimes help.
Operating from batteries solves many problems. (At low power levels this is excellent solution)
Maybe use different motor and controller.
Place capacitors across motor and/or controller. (Check with motor and controller manufactures first and only use capacitors with AC rating)
Can you use a 60 hertz motor and no controller?


Why are you using 400V for motor? If this is a small motor, 400V is very high.

 

[0xDEADBEEF]

How do you measure displacement and force? Can you look at the signal with a scope?

Frequency filtering can help. For very high frequencies ferrite beads can help. You can never have enough shielding. Transzorbs could help. Just plugging the motor controller into another wall outlet could do the trick.

At this point you should try to get as much information about the noise signal as possible.

 

[Bob S]

As a last resort, I have wound 10 or 15 turns of a power cable (both conductors) thru the hole in a high-mu tape-wound toriod (1-mil thickness mu-metal alloy). This is the equivalent of a high series resistance in series with any ac ground current.

Bob S

 

[wzhit]

It is best to have separate ground cables.
Sometimes it works best to ground things and other times the opposite is true.
Operating sensitive equipment with isolation transformers sometimes help.
Operating from batteries solves many problems. (At low power levels this is excellent solution)
Maybe use different motor and controller.
Place capacitors across motor and/or controller. (Check with motor and controller manufactures first and only use capacitors with AC rating)
Can you use a 60 hertz motor and no controller?


Why are you using 400V for motor? If this is a small motor, 400V is very high.

I try to response to your reply.
1 I tried to use a Meta System power supply for all the devices except the motor. This indicates two ground cables are connected, one for the motor and the other for all other devices. However, the results didn't change better.

2 As for an isolation transformer, I have no idea about the influence. Maybe it is one choice in the future.

3 As for "different motor and controller", we have four motors and four controllers, and all of them have the same problem.

4 We are waiting for a filter from the manufacture.

5 I don't know how to operate the motor without a controller. Is it dangerous?

6 The nominal torque of the motor is 4Nm. Is it small?

Anyway, you widen my choices. Thanks, Carl.

 

[wzhit]

How do you measure displacement and force? Can you look at the signal with a scope?

Frequency filtering can help. For very high frequencies ferrite beads can help. You can never have enough shielding. Transzorbs could help. Just plugging the motor controller into another wall outlet could do the trick.

At this point you should try to get as much information about the noise signal as possible.

My replies:
I measure the displacement and force via displacement transducers and load cells. The signals have been shown by a scope, and the noise is there.

The noise is introduced by the running motor. It has something to do with electricity or EM field.

The same thanks to you!

 

[wzhit]

As a last resort, I have wound 10 or 15 turns of a power cable (both conductors) thru the hole in a high-mu tape-wound toriod (1-mil thickness mu-metal alloy). This is the equivalent of a high series resistance in series with any ac ground current.

Bob S

This idea is very interesting,haha. I'm to consider it.

 

[wzhit]

Now, the question is:
if all the motors, or PM AC synchronous motors, have the low frequency noise.

I mean firstly I want to know if it is possible to solve the problem for me. Maybe it is a general problem, and I can do nothing.

Thanks to everyone.

 

[lionprecision]

You do not have a grounding issue as that nearly always results in line frequency noise.
I don't know where you are seeing this noise. Is it in the output of the load cell? the displacement sensor? What technology is used in each?

Please say more about "disconnecting the cable from load cell to motor." What cable? If you disconnect the load cell, of course it will stop producing an output so that must not be what you mean.

One insidious source of noise is beat frequency. If you're running 60 Hz and 50 Hz in the same system, the difference is 10 Hz. Is your noise sinusoidal?

 

[wzhit]

You do not have a grounding issue as that nearly always results in line frequency noise.
I don't know where you are seeing this noise. Is it in the output of the load cell? the displacement sensor? What technology is used in each?

Please say more about "disconnecting the cable from load cell to motor." What cable? If you disconnect the load cell, of course it will stop producing an output so that must not be what you mean.

One insidious source of noise is beat frequency. If you're running 60 Hz and 50 Hz in the same system, the difference is 10 Hz. Is your noise sinusoidal?

www.lionprecision.com

I found the noise in the output of both the load cell and displacement sensor.


See the previous post:
When the load cell is put on the ground and is connected to the motor via a cable,
the noise is there; if disconnecting the cable, the noise disappears.
It clearly shows the noise is due to the EM field of the motor.

First I used one cable to connect the load cell, and then disconnect it. The change of the noise during this procedure is observed. So I think the noise, at least this part, is not introduced by mechanical vibration.

I'm in Italy, and I guess the line frequency of the power supply system is 50Hz. I don't have another source with the frequency of 60Hz.

The noise is not sinusoidal.

Hope this can make the question clearer. Thanks.

 

[lionprecision]

I'm sorry, but I don't understand "connecting the load cell to the motor" with a cable.
"On the ground" means there is no "load" on the load cell. Are you connecting the frame/chassis/housings of the load cell and motor with a cable?

If not sinusoidal, what does it look like? What exactly is happening at 10Hz?

 

[wzhit]

I'm sorry, but I don't understand "connecting the load cell to the motor" with a cable.
"On the ground" means there is no "load" on the load cell. Are you connecting the frame/chassis/housings of the load cell and motor with a cable?

If not sinusoidal, what does it look like? What exactly is happening at 10Hz?

As is well-known, mechanical vibrations and electrical currents can introduce noise. To avoid the mechanical vibrations, I took the load cell away from the motor(in fact, it is a thrust), say, put the load cell on the ground. For sure there is no load. To keep the electical current there, I used a cable to connect the load cell to the motor. This sounds strange, but it works well to show the type of the noise. The conclusions are: the noise comes from the current, or something like that, but not a mechanical vibration.

You can see the attachement of the force noise.

 

[wzhit]

See attachments

 

[Carl Pugh]

Motors are not my specialty, however with that in mind:

400 volt is too high a voltage to be used around sensitive circuits.

Motors with magnets must have a controller? The controller switches very fast, so there is noise coupled into everything.

Motors with brushes will have high frequency noise.

My suggestion would be to try a motor without magnets and without brushes. This should be an induction type motor similar to the motor in a 50 or 60Hz electric clock.

The idea by Bob S to use a torroid is a good idea. You may be able to obtain line filters that clamp around the power cable (or other cables). These have a snap and open and then are clamped around the cable.

 

[wzhit]

Last post

We bought a filter and connected the ground cable with a different scheme, and now the noise is reduced by about 50%.

One can't pay too much attention to ground cables.

Thank you everyone for your help.