Mark Ve RTD issue - 13 and 29 Reading Falsely Low


I am having an issue that I hope someone can point me in direction to getting it figured out. We have a 7FA with a Mark V that was upgraded with the digital digital front end, would you call it a Mark Ve?

Anyway, I have two RTD's that are reading -165 Deg F when they should be reading around 95 Deg F. They are DTGGC13(cold gas coupling end) and DTGGH29 (hot gas center). When I lift them and measure the resistance that are around 116 Ohms and seem to be okay but when they are both landed they read the -165 value. If I lift the grounded lead in the field for the DTGGH29 the other RTD (13) returns to its expected temp are 95 Deg F.

All wiring appears to be fine in the collector end in the Mark panel, any thoughts?

Jesse, I'm not familiar with a 7VA or Mark V. I was experienced with RTDs years ago when I worked for a company that manufactured them. I designed analog electronics (signal conditioners) to work with them. So, just some basic thoughts.

A shorted RTD input will make the temperature read very low. It sounds like you have a mis-wiring somewhere that is shorting out the input, especially since ungrounding one causes the other to read properly. With the RTDs both connected, remove power from the Mark V and measure the resistances across the inputs. If they are not correct, try ungrounding the lead and see if the resistance measurements improve. Since ungrounding one causes the other to read correctly, I'd guess you have some kind of ground loop or extra unwanted ground connection. Check all your wiring.

Do the RTDs each have 3 wires? If so, two wires are connected to one side of the resistance element and one to the other side. The reason for two connected to the same side is for the circuit to compensate out the extra ohms of connection wires. If they are incorrectly wired to the instrument, the two connected to the same side will short the measurement out. Hopefully the Mark V has detailed info on which wire goes to which input terminal. If not, maybe you can just experiment and see if you can make it work.

If the RTDs each have 4 wires, my thoughts are similar except that two wires go to one side of the resistance element, the other two to the other side.

Good luck!

An RTD circuit works because the AI card pumps a constant current through the floating, ungrounded resistance of the RTD and then measures the voltage drop (Ohms Law) across the RTD resistance. No wire in a 2-wire, 3-wire or 4-wire RTD sensing circuit is grounded. Ever. The shield (or screen), if there is one, is typically grounded at one end, usually the analog input end, but the shield (or screen) is not a direct part of the sensing circuit.

I don't know what a Mark Ve DTGGH29 is, but there is no RTD sensing circuit in the world that uses a grounded sensing wire, because the keeping the AI's current source a constant value is critical to the measurement. A ground in the sensing circuit can and will lead to ground current loop that will alter the fixed constant current source value and produce an error in the measurement.

One other possibility is a broken element, that appears to the AI as an open circuit. Many AI's are designed to detect and throw an error code and/or to go to failsafe mode. Many times RTD configurations allow for downscale or upscale indication for a fail-safe condition, like a broken, open circuit element. Downscale failsafe might be -165 Deg for your system. But if you measured the RTD's resistance and got a reasonable value, then it is not an broken, open element. The RTD resistance measurement also rules out a shorted RTD element.

My experience mirrors Harry T's experience, which is that there is a wiring issue with your bad-reading RTD circuits.

There's something not right about a "grounded wire" in an RTD circuit.

On our 7FA's if you are in the collector end enclosure all the generator RTD's land on a terminal strip in JB005. They are 3 wire RTD's that that connect to the field wiring that runs back to the Mark V migration over a twisted shielded pair. For the RTD's coming from the generator it has one blue wire, two white wires and a shield. The shield connects to the field wiring shield along with the blue and one white wire to the twisted pair. The other white wire is grounded in the junction box. This is how all the RTD's are wired and am under the impression that this was originally done by GE to save money by running only a few compensation loops vs one for each RTD(they are jumpered in the mark to several channels).

To me it seems like their might be a ground loop but I was reaching out to see if any other members have come across this issue in the past which could save me a little time stumbling my way through it.
It's nearly impossible to try to help troubleshoot this problem without understanding one (little) thing:


How long has it been a problem?

Did it start with a single RTD input, or both at the same time?

Did it start after a maintenance outage? If so, what was done during the maintenance outage?

Problems like this rarely, if ever, just magically start on their own. Excessively hot temperatures can cause problems with wiring insulation, and in the case of generator RTDs--many of which are embedded in generator windings (though not usually the heat exchanger gas temperature RTDs)--many of them simply fail over time, and while they can be replaced it's not a simple, easy, inexpensive matter (usually).

No; this usually happens with gas temp RTDs after some kind of maintenance outage when care wasn't used when disassembling or reassembling.

Be that as it may, GE has done an unusual thing over decades with generator RTDs. They use three-wire RTDs, and they ground one of the RTD common leads at the main generator junction box (the one that's used for lead length compensation), and they use TSPs (Twisted, Shielded Pairs) to get the other two RTD leads connected to the temperature sensor (Mark* or temp meter). They do this with EVERY generator RTD--use a TSP to get two of the RTD leads connected back to the Mark* or sensor, and then they use one or two conductors of another TSP cable and connect them both to ground in the main generator junction box, and then use the other end of those TSP conductors as the lead length (compensation) conductors for the Mark* or sensor. They jumper the lead length (compensation) screws together at the Mark* or sensor and connect the grounded TSP conductors to the jumpered string(s). A lot of people are saying, "WTF?!?!??!!! That's insane--and it will NEVER work!!!" But it does, and has for decades. And, it's saved hundreds of dollars over those same decades by using TSP cabling instead of triad cabling (three conductor, twisted, shielded cables). Yes, this was done to save a few pennies per generator. Electrically--it is perfectly fine. But, it drives people INSANE!!!

So, to sum up--as has been said already, there is a wiring problem somewhere. If the original poster would think back to when the problem started he will probably find that something wasn't disassembled or reassembled properly, or someone didn't re-connect one of the RTD common leads to ground in the main generator JB (why? because he thought it wasn't the proper thing to do, even though it had been working for 20+ years, the person KNEW BETTER--and that's when the problem started, maybe). Or a wire got crushed when re-inserting one of the gas heat exchangers--something like that--during a maintenance outage. It's entirely possible that someone was mucking with the RTD wiring in the Mark* V panel and got those jumpers mixed up on the I/O terminal board (on <C>, I think).

Go find the friggin' wiring problem. DON'T FREAK OUT when you discover the way the wiring was done 20+ years ago and has been working fine for most of that time. Find and fix the wiring problem, or recognize the RTD itself may be failed and will have to be replaced the next time the generator has a major.

[By the way, I think the topic of RTD wiring to the Mark* has been covered several times before, and there may even be some drawings somewhere which might make things a little clearer....]
Well, I learned two things today - GE grounds the lead wire compensation sense conductor on a 3-wire RTD and the element sense cables are twisted pair, not triads. Gee Golly Whiz, it is a big wide world out there. Obviously it works when wired properly and has for decades; it just isn't 'conventional' with rest of the world.

WTF carries the ball. Go for it.

Oh, and depending on where in the circuit the element resistance was checked, there could always be an open circuit, a broken conductor, from the check point to the end destination of the cabling.