Inductance in adjacent wires

[oldsparky52]

I'm a non-electrician involved in a project where inductance is creating control issues. I'm looking for informed thought or reaction to the situation, because it seems unexpected and not really understood by our supplier. The application is a 120 v residential sewage lift station, that involves a submersible pump; low level alarm float, a control float and a high level alarm float in a 1000 gal tank. All are 120 volt. The float and pump wires run together in conduit, underground, between a control panel with Logo plc and a junction box at the tank. Typical runs would be 15-30 feet. The low float circuit is normally energized. The middle float activates the pump when the circuit closes, shuts it off when it opens. The high alarm float is normally not energized. The PLC apparently looks for the control float circuit to drop below 40v to in turn shut the pump off. The apparent problem is that there is anywhere from 25 to 50 v induced in the pump control float wiring, presumed to be from the energized low float wire. It varies by installation, and also varies somewhat from one day to the next. There have been several situations where the PLC keeps the pump running to the point of tripping the low level alarm, apparently because it sees over the 40v threshold in the control float line. Suppliers are saying they have not had this issue at hundreds of other installed sites. Replacing the standard wires with shielded cable seems to lower the voltage readings a lot, and is the intended fix for all sites. Nobody can explain why we alone have the problem, or whether the size of the induced voltage is reasonable or not. I have no experience in this but I am really surprised that an insulated wire would show 40 volts just by running beside a 120v insulated wire. I would have a guessed a fraction of a volt would be seen.
What do more experienced people think about this? Have you seen this sort of thing before?

I think your problem is highlighted above.

I've only wired a few of the systems you are describing, and NONE of them involved a junction box at the tank, and depending on how it was done, it might be a violation on a code.

Did the floats and motor come with long cords on them?

 

[Dsg319]

How much difference shielding makes is highly dependent on how the shields are terminated. If terminated incorrectly it will not help much.

On analog signals we’ve always terminated (grounded the shield) on the supply side, and just wrapped and tapped off on the field side.

Never really put much thought into a false digital input to a PLC from capacitive coupling near power conductors. Is it possible? And would a properly installed and terminated shielded cable be helpful for digital inputs? I’ve never thought they were needed until now.

 

[gar]

210401-2018 EDT

SAF:

This thread is a good illustration of how lacking many electricians are in a basic understanding of electrical circuit theory.

I tried to adhere to the forum rules on not helping DYUs. But most responders have not paid any attention to those rules.

I will comment later if the moderators have no objection.

.

 

[ptonsparky]

This is far from a typical DIYer thread.

 

[hbiss]

I tried to adhere to the forum rules on not helping DYUs. But most responders have not paid any attention to those rules.

This thread should have been ended at post #8. Any further advice should have come from the OPs EE or preferably Siemens on how to actually remedy the problem. We went beyond the rules to say what was causing the problem.

-Hal

 

[oldsparky52]

This thread should have been ended at post #8. Any further advice should have come from the OPs EE or preferably Siemens on how to actually remedy the problem. We went beyond the rules to say what was causing the problem.

-Hal

I'm curious, was that directed at me because I said " I think your problem is highlighted above. "?

 

[SAF]

I want to thank all of you for your helpful advice, and assure you I have no intent to do it myself, and that we have plenty of electrical people involved who are the final do-ers and technical deciders.. My takeaway from all of this is that perhaps our project group didn't look closely enough at the snubber or load option as an alternative to the re-wiring fix we have planned, and I'll be asking a more senior electrical resource in our group to re-examine this option. Best case we save some time and money, worst case nothing changes.

 

[oldsparky52]

I hope you will share the outcome.

 

[hbiss]

My takeaway from all of this is that perhaps our project group didn't look closely enough at the snubber or load option as an alternative to the re-wiring fix we have planned, and I'll be asking a more senior electrical resource in our group to re-examine this option.

You can't imagine how many times we get questions from electrical professionals as to why they are getting some weird voltage reading on their meter. Gar says it best:

This thread is a good illustration of how lacking many electricians are in a basic understanding of electrical circuit theory.

I also take issue with the way Siemans designed the inputs on that PLC. They should not be susceptible to induced or capacitively coupled stray voltages when the input is open. Poor design by poor engineers.

-Hal

 

[kwired]

This is far from a typical DIYer thread.

Agreed. And the OP seems more qualified than those "professionals" he has that can't seem to figure this out.

 

[hbiss]

Nobody, at least I, never said this was a DIY thread. However, the OP by his own admission is not related to the electrical profession. I'm not exactly sure what he does and for whom, "consultant" has a very nebulous meaning. So about all we are allowed to do is to listen to his story and advise but not give instructions on how to do what needs to be done.

I think we have done that.


-Hal

 

[gar]

210401-1546 EDT

A lot of 2 wire cables have a capacitance of 20 to 100 pfd per foot. It is very difficult to find cables less than 10 pfd/ft.

Some quick measurements ---
#14 Romex --- 13 pfd/ft black to white --- 21 pfd/ft black to EGC
#18 rip cord --- 21 pfd/ft
#16 twisted together --- 14 pfd/ft

Wire size, spacing, and insulating material will determine the capacitance.

You can calculate the capacitance for any length, and in turn the capacitive reactance at 60 Hz, and knowing meter impedance estimate capacitively coupled voltage.

Measurements by the OP are likely.

.

 

[ptonsparky]

Change it all to DC.

 

[hbiss]

Change it all to DC.

And I'll bet the version with DC inputs will have problems also. Doesn't look like this was well thought out by Siemans.

-Hal

 

[gar]

210403-1035 EDT

ptonsparky:

Start a new thread here based on a PLC that you normally use that has a logic input requiring 120 V 60 Hz as an input, This should be a PLC where you have the manufacture's specifications on the characteristics of the AC logic input.

Then I will try to help you create the problem, and develop some solutions.

.

 

[ptonsparky]

210403-1035 EDT

ptonsparky:

Start a new thread here based on a PLC that you normally use that has a logic input requiring 120 V 60 Hz as an input, This should be a PLC where you have the manufacture's specifications on the characteristics of the AC logic input.

Then I will try to help you create the problem, and develop some solutions.

.

We don't have these problems anymore. Various reasons. Long distance control is more frequently done by radio,cell or DC.

 

[gar]

210406-2342 EST

SAF:

I not going to answer your question because of the forum rules. However, I will make some additional comments.

First, most electricians do not have a good understanding of basic electrical circuit theory, and therefore, don't think of many different solutions. Thus, they may not be real useful in solving your problem.

I ran an experiment with a 30 ft cable, applied 120 V to a lead, and measured the capacitively coupled voltage to another wire in the cable. With just a Fluke 27 as the load ( 10 meg input resistance ) I read 0.29 V ( 29 millivolts ). Shunt loading with 15 k ohms across the meter input the reading dropped to 0.003 V ( 3 millivolts ). 15 k is in the ballpark of a typical 120 V solid-state input.

Your problem can be solved with the correct help.

.

 

[tortuga]

Most septic system like this do not send 120V to the floats, rather 32-36 VAC or VDC.
I question weather there is really 120 to the floats?

 

[SAF]

Tortuga, there is definitely 120v AC to the floats.
Some of you asked that I Iet you know how this turned out. The electrical contractor obtained a few 10w 5000 ohm resistors and installed them this morning on a couple of places to see if it worked. The un-energized float circuits were reading 55v before and 0 after so this will solve the problem. Again, thanks for helping me get things on to this track.
I was with the electricians when they were putting them in, and they were scratching their heads, just couldn't understand the problem or the solution, which has me scratching my head. I think there are a lot of tradesmen that understand what they are doing but maybe not the theory that underlies it.

 

[ptonsparky]

Tortuga, there is definitely 120v AC to the floats.
Some of you asked that I Iet you know how this turned out. The electrical contractor obtained a few 10w 5000 ohm resistors and installed them this morning on a couple of places to see if it worked. The un-energized float circuits were reading 55v before and 0 after so this will solve the problem. Again, thanks for helping me get things on to this track.
I was with the electricians when they were putting them in, and they were scratching their heads, just couldn't understand the problem or the solution, which has me scratching my head. I think there are a lot of tradesmen that understand what they are doing but maybe not the theory that underlies it.

And any mix in between. I will admit to it.