You know what happens when we ASS-U-ME ........

A simple assumption could be hiding the problem.

Every detail matters.
.
.
.

 

If twisted shielded pair which is carrying 25mv and is in going in parallel to ethernet cable. Can ethernet cable induce noise in tsp.

 

It is certainly POSSIBLE, depending on the physical conditions. If everything was done "perfectly" then it would be unlikely. But if there is a net "non-zero" current between ends of the ethernet cable then it could cause an induced voltage.

 

Reminds me of the grounding practice when CNC (5v based) machine control first came on the scene, noise sometime played hasvoc and the general advice was to ensure one end only of any sheild was connected to earh or GND, in order to prevent ground loops.
One problem is, is that the shield can also act as an antenna , recieve or transmit!
Then a few years ago, Siemens came up with the recomendation for Equi-potential earth bonding, at the same time Ground each end of any shielded cable..!!
This recommended bonding all metallic parts of the machine in order to ensure no voltage or signal appears at different GND points.
See Pub.

 

"Grounding both ends" works best when it does not result in noise currents flowing in the shield circuit.
AND, ONCE AGAIN, if this is the same circuit that has been working in identical applications, then the problem is probably caused by some portion that is not quite the same as what has been working well. One extra ground to the frame point, or one missing shield connection, or something like that.
So I am not suggesting a design change, but rather an investigation to find what is different.

 

Not sure about store bought packaged enet cable as it was always buy and use out of the package the majority of the time, but using bulk shielded twisted pair instrument cable you ALWAYS ground ONLY one end of it for that very reason. I suspect your cable is the culprit. It was also the same running bulk shielded Cat 2&3 back in the day, only one end is grounded to prevent any picked up common current transmission between ends. Common practice with any shielded cable. You might also check your terminals for shorts during termination.

 

"Grounding both ends" works best when it does not result in noise currents flowing in the shield circuit.

Hence the Siemens recomendation for equi-potential bonding i.e. ensure same potential either end by earth bonding conductor.
Now the current recommended practice.

 

Siemens recomendation for equi-potential bonding

That's a new one to me since I've been out of the practice for many years now since retirement. Can you post a link to it, I'd like to read it. Sounds interesting. Thanks, Sam

 

That's a new one to me since I've been out of the practice for many years now since retirement. Can you post a link to it, I'd like to read it. Sounds interesting. Thanks, Sam

Post#24 has the link

 

Post#24 has the link

Actually, my point is that if it is noise then why it is not there on low rpm setting. The moment i cross idle setting (minimum throttle power), egt starts fluctuating. I think that may be it is because my voltages are tapped thru tsp behind the cockpit guage which is analog. And analogue guage is getting signals from thermocouples. Is it possible that due to vibration, my loose connection in thermocouple is causing this as analogue guage is less sensitive. One more probable reason is that these 3 stp are passing close to ethernet cable of recorder. High data rate can also cause emi. But in 03 tsp, the other parameter is not affected at all. Moreover, i removed the input signal wires from recorder and i was surprised to see that there were fluctuations at idle and very low fluctuation above idle which is totally opposite. Though i expected that with no input there should be no output. I have also freshen up all ground to eradicate any potential weak grounding. So what can it be.

 

What does the acronym PE stand for here:

I scanned the entire document and could not find it but it found every pe word in the document.

Also, what I'm seeing here is shielding for motor drive cabinets and cables from it to the driven motors. I don't think that is the case in this question. I can see the need for extreme diligence to grounding when dealing with drive cabinets and its cableing but that is not the case here. It is an interesting document, thanks for sharing it. FWIW

 

PE normally stands for Protective Earth!

 

Just a thought - have you tried a known good LRU in this installation. That would immediately indicate whether the problem exists in the LRU or the installation.

 

If the problem appears mostly at higher RPMs, then a mechanically insecure connection would be suspect. So now there is another suspect.