Actuator Interference

[jcole]

It does sound like interference over the shared 480V line. Can you temporarily install an isolation transformer or EMI filter in the line to the valve as a test?

I assume what you mean by isolation transformer is one that has a 1/1 coil ratio. Correct?

Thanks for the advice.

I have been doing control work in a manufactoring enviroment for 8 years and never had a problem like this. This is a fairly new job for me and by far the oldest the equipment I have ever worked on. I am more familiar with industrial bus networks like devicenet. Never had a communication or interference problem of sorts.

 

[jcole]

Would it be worth getting the power quality meter out that as far as I know has never been used or is there any body there that knows how to use it?

Its a real nice one. Fluke 1534 I think. Would this pick up and show interfernece on the 480v?

Anybody ever used this model? Comments.

 

[justdavemamm]

After the blower is started the actuator would read 100% open with a 20mA signal, which is correct (demand ok), but the actuator would only be 50% open mechanically.
But as soon as you turn blower off the display reads right and actuator acts correctly.

So, when the blower is running and the signal is 100% (20ma) the display reads 100% and the actuator is only open 50%. Then the blower is shut off, the signal is kept at 100% and now the actuator opens to 100% ?

What is unclear is that the display read out both the commanded position and the actual (as it see it) position. I think you say so.

Any 'slipage' I was imagining would only reset itself if the command signal went to 0% after the blower turned off as that would be a 100% command change and the actuator only moved to a hard stop at 0% from 50%.

 

[jcole]

So, when the blower is running and the signal is 100% (20ma) the display reads 100% and the actuator is only open 50%. Then the blower is shut off, the signal is kept at 100% and now the actuator opens to 100% ?

What is unclear is that the display read out both the commanded position and the actual (as it see it) position. I think you say so.

Any 'slipage' I was imagining would only reset itself if the command signal went to 0% after the blower turned off as that would be a 100% command change and the actuator only moved to a hard stop at 0% from 50%.

Close. The reading on the actuator changes to the correct mechanical position. Not the mechanical position changes to the correct display reading.

Basically, the only thing that changes when the blower is turned off is the display reading.
Ill say again, when the blower is not running the actuator is electrically and mechanically correct. In other words if the signal is 12mA (50% demand) the actuator displays "demand ok" and mechanically it is 50% open. As soon as I turn the blower on the display will read 100% open and "demand ok". But the valve will not be 100% open it is about 50% percent open. The reason it goes to 100% percent electrically is because the header pressure is what controls the position of the valve and since the valve is not opening any further than about 50% the header pressure set point is never being met so it keeps trying to open it further until it maxes out electrically (20mA).

I spent a full day trying to find mechanical slippage but it just isnt occuring.

Thanks for all the replies.

 

[ELA]

Thus far it seems that the encoder feedback signal is reading incorrectly while the blower is on.

Since the encoder is an integral part of the value assembly it seems the only way to protect it is to filter any noise that might be present in its power feed.
(assuming it is conducted and not emitted noise).
If there are no other connections to the value when you use your fluke to drive the control signal then I do not see any potential ground loop path?

I have assumed the valve is powered by 480V? If so then you could try a 1:1 isolation transformer. Or a high voltage rated EMI filter if you have access to one.

A power quality meter reading would be a good bit of data to help diagnose potential noise issues.

 

[nawao]

I miss something

I miss something

All connections carying analog signals must be done with shielded multicore cables. The shield must be grounded at one end only, preferably at the actuator side.

All connections carying analog signals must be done with shielded pairs multicore cables. The shield must be grounded at one end only, preferably at the actuator side

 

[G & G Electric]

but I have solved situations like this by cutting off the drain wire (shield) that is grounded to minimize EMI.

This may not relate to your situation, but is your drain wire connected at both ends? In low voltage controls this will create an interference loop, and will cause computer programs to act strange. Cutting off one end of the drain wire (terminate it at only one end) has solved interference problems for me in the past. Not sure if this applies due to the incredible voltages you've mentioned.