Instrument Signal/Solenoid or Switch Cable Separation 5

[Scotsinst]

Customer specification calls for 12" separation for power and instrument signal cables running on cable tray.
Our designer did not consider 24VDC solenoid/switch power as a "power" cable (typically power assumed to be 220VAC for lighting etc) so we have analog and digital cable in same tray with a GRP barrier in the middle. Customer insists on 316sst barrier to prevent emf causing issue with signal cables.
Since 24VDC is low voltage and generates a small static magnetic field, will it be a problem for the 24VDC 4-20mA instrument signal?
The GRP barrier dielectric strength is 170 volts/mil. minimum and it's 3mm thick.
Any input appreciated.

 

[VEBill]

The worst EM interference would probably occur when the voltage is cut to release the solenoid. Ensure that there is a reverse diode on the solenoid coil to catch the back EMF at the source.

But even if there isn't a real EM problem, you might still have to meet the specs anyway.

 

[dpc]

Solenoid valve coils have a surprisingly high inductance and can generate interesting transients when de-energized. Proper twisting and shielding of the wiring to the solenoids can reduce the impact on other adjacent circuits. 24V dc power circuits to instruments, etc isn't normally an issue.

I'm not sure a metal tray barrier (the SST requirement probably is not relevant to EMI issue) is going to really solve the problem, if there is one.

If this doesn't involve miles and miles of circuits, you could propose using metal-clad cable for the 24V dc Solenoid circuits.

 

[Skogsgurra]

A free-wheeling diode will solve any DC coil issues. The magnetic field is neglible. No shielding no barriers needed. 4-10 mA is quite robust. If it were microphone (microvolt) signals, I may have been concerned. But not when it is 4-20 mA or +/-10 V. Not even encoder signals will be affected by the 24 V DC coil wires. Been there, done that - many many times.

But, as V1Bll says: You may have to follow the specs. Even if there is no technical reason for it. It all depends on what your relation to the customer (or inspector) is.

Gunnar Englund

http://www.gke.org

 

[dpc]

I agree that the diode will take care of the transients. It's just that in a large facility, this doesn't always get done on every solenoid valve. Especially since the large solenoid valves tend to fail with some regularity.

 

[Skogsgurra]

I always make the panel-builders put diodes at the driving end, not across the coils. That way, it never gets disconnected or forgotten.

Gunnar Englund

http://www.gke.org

 

[Scotsinst]

We have diodes on the panel end to protect the PLC, but to be effective for emf reduction it should be as close to the inductive source as possible.
We also used some which fit neatly onto terminal strips in the JB but some customers insist on being as close to the coil as possible.

 

[itsmoked]

You can put them on both ends..

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Scotsinst]

I don't mind putting them on both ends especially for solenoids which operate several times a day.
Most of our solenoids are on SDV's which operate every two years depending on the testing frequency.
I just hope the customer will accept this solution.

 

[itsmoked]

SDVs? Soggy Diaper Victims?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

Scotsinst,

I was waiting for that response. That myth (place diodes close to coil) has no physical or practical underpinning at all. I hear it very often. But it is wrong.

The switch-off transient (which is the worst interference source) occurs when the coil current is interrupted by a relay contact or switch (if the coil is switched via an SSR, the free-wheeling is already taken care of). Now, if you connect a diode at the driving end, the current doesnt get ripped off and there is no transient. What happens is that current decays slowly to zero. No high kick-back, no transients, no HF noise.



Gunnar Englund

http://www.gke.org

 

[itsmoked]

I don't believe it.. Whatever current you have now has to flow down the long cable to the snubber. This has got to 'transmit' more magnetic energy into the environment, then if this 'loop' remains at the coil.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

No Keith. That is not the problem.

The magnetic field is very weak and causes no disturbance in a 4-20 mA circuit. It is the repeated refiring across the opening contacts that is the problem. It produces lots of HF interferece. Often into MHz. I shall try and find a recording I once made. That will convince you.

Gunnar Englund

http://www.gke.org

 

[Skogsgurra]

I coudn't find it. So I made a new one for you. This is a 12 V DC coil being switched off. You notice that the first voltage spikes are of quite low amplitude (the contacts haven't separated very far and the restriking voltage is low). As the contacts separate more and more, the voltage gets higher and the restrikes more and more powerful. This is where the most intense HF is being emitted.

Finaly, the contacts have separated so far - and the energy in the magnetic field has also been used up - that the final so-called Power Pulse is seen. It is the rounded wave-shape at the end of the transient.

The time scale is 20 microseconds/division and the vertical scale is 200 V/div.

Adding a diode parallel to the coil (or at the driving end) will give you a very boring 12 V DC level. No transients. No Power Pulse. No nothing. As I said before: Done that, been there. This is the way it works.



Gunnar Englund

http://www.gke.org

 

[itsmoked]

Huh.. Well thanks for that new knowledge Gunnar.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Scotsinst]

Me too; thanks Gunnar. I was told about the solenoid diode location about 25 years ago and never questioned it.

 

[Skogsgurra]

You are most welcome. I know that very few people are aware of this phenomenon. So snubbers and diodes are placed in very awkward places. And the myths are spreading. I think of myself as kind of a Myth Buster (r) ;-)

Gunnar Englund

http://www.gke.org

 

[VEBill]

Interesting and informative.

 

[danw2]

Skogsgurra, Great scope screen shot. Thanks. Dan