ANALOG NOISE!

[UncleT]

Anyone got a recommendation on how to isolate A/I from 480 noise...other than remove it from conduit/enclosure? (4-20ma)

 

[itsmoked]

There really is no way.

However!

If you have control over the A/D what you do is synchronize the A/D to the power line. You then read the A/D channels each at the exact same time 180 degrees apart and average the two readings. This can remove the positive and negative power line contributions by averaging them out. Any timing errors will allow the AC(or some part of it) back in.

Alternatively you can use a V/F for the A/D conversion and run it over 180 or 360 degrees. Same result.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

480 volts in the same enclosure as 4-20 ma is not much of a problem. I have not yet seen 4-20ma signals run in the same conduit as 480V. I suggest removing the 4-20 ma from the conduit. Twisted pairs help.
yours

 

[dpc]

Does this involve a Variable Frequency Drive by chance?

 

[UncleT]

Yes, it does involve a VFD (Cutler-Hammer). I have sinced moved the twisted pair and it has helped but didn't completely remove the noise. The poblem only exists when the VFD is on, I can practically track the VFD ramping up with this particular channel.

T

 

[waross]

Is this common mode noise or noise on the signal?
What type of shielding and or grounding do you have? Multiple grounds can put a lot of noise on a signal. More than one ground point on the shield can increase noise.
respectfully

 

[dpc]

It's the RFI from the drive and associated wiring more than the presence of 480 V that is causing the problem, I'm guessing.

The 4-20 mA circuit should be twisted, shielded pair - sounds like it already is.

Is this 4-20 mA cable in the VFD enclosure itself? Does it run in its own conduit from there, or is it running with the drive ouput to the motor?

 

[ozmosis]

Is your 4~20mA signal an input signal to the VFD or an output signal from the VFD to another device?
If it is an input, then you need to make sure the device feeding is at the same ground potential to the VFD. This may mean connecting a very thick ground cable between the two to ensure you create a balanced ground pot.
You can simply test this if the distance isn't too great by running a thick cable between the two devices as a test and see if the noise goes.
I have just been involved in a similar problem where the drive was mounted on a brick wall, they had the minimum earth cable running about 100m to the MCC and then a 4-20mA signal running to the drive. The earth cable provided the high frequency noise with a high impedance and therefore didn't go there, so the path of least resistance was down the AI cable and the distruption to the setpoint was terrible (up and down the speed signal went) A scope showed up all the noise.
A quick test with a (temp)cable strapped to all devices in the system to equalise the ground solved the problem.
Then I advised the installer to take all this new found knowledge and install it correctly.