Trane Tracer Summit odball...

[GT Jets]

I have a brain teaser for you.

So I have a fairly large facility I maintain and have been at this site for a little over 12 years. They have a Trane control system that was installed in the mid '90's.

Recently the boiler plant has been flipping out anytime the emergency generator does a test run. Not a loaded test, just a maintenance run with the transfer switch in the "normal" position. The generator room is immediately adjacent to the pump room. The generator is mounted on a spring isolated inertia base as is the control box that the PCM is mounted in (separate bases 20+ feet apart).

When the generator runs, the hot water pumps, which are on a VFD, start and stop at random, the speed signal never drops off, but the start stop at the relay on the PCM drops out and comes back like it is possessed.

While looking into this today, I pounded the top of the enclosure with my fist and was able to get it to drop out, then about 15 seconds later, fires right back up... The light (led) for the binary output goes off, the relays drop out, then right back on... Watching it with my laptop logged into the controller you can actually see the output "off" then shortly back on (not real time, but I know it is something in the controller.

This was at that point encouraging, but was short lived, I was thinking I had a problem in the controller, after reprogramming another spare PCM and installing it, the new(er) one does the same thing, exactly the same.

I am at my wits end.. Anybody have anything?, I know I have not included everything I tested in this post, but I have tried numerous other things to try and hunt down the issue.

The programming is pretty straight forward, there is an averaging pressure differential system that takes two different pressure transducers, averaging them together then using the one signal to control the pump, the pumps have a bypass, one VFD, two pumps, so the outputs are pump 1 bypass, pump 1 VFD, pump two bypass, pump 2 vfd. The inputs are rotary switches mounted in the door, there is one for each pump that does VFD/off/bypass.

Thanks in advanced.

GT

HELP!!!

[GT Jets]

Proof.... Am I losing it?

[amigo]

although pcms were pretty bulletproof for their time, they did have consistent corruption issues.
from what you are describing, I am thinking the pcm is resetting itself. If you have not tried it, I would focus on the power supply and power connections on the board. May be try a UPS connected power supply and see what happens. Also check for corrosion signs on the back of pcm boards.
good luck.

[Tech Rob]

May be reverting to comm loss values for relay positions. Do you have any PCM comm loss alarms in Summit?

[GT Jets]

although pcms were pretty bulletproof for their time, they did have consistent corruption issues.
from what you are describing, I am thinking the pcm is resetting itself. If you have not tried it, I would focus on the power supply and power connections on the board. May be try a UPS connected power supply and see what happens. Also check for corrosion signs on the back of pcm boards.
good luck.

Thank you for the reply.

I have installed local ups units on every PCM, UPCM and BCU in the campus in the last couple of years.

I also replaced the 24 volt transformer on this cabinet yesterday (started throwing parts at it out of frustration).

I put an oscilloscope on the 24volt power into the PCM and didn't see much of a fluctuation, just about what I thought the relay dropping would do (23MV).

As for the corrosion, after looking at the PCM I removed I did notice a bit of oxidation/corrosion on the contacts of the double throw double pole power switch. I did use some contact cleaner on the switch of the replacement board, but the switch condition was in tip top shape compared to the old one.

[GT Jets]

May be reverting to comm loss values for relay positions. Do you have any PCM comm loss alarms in Summit?

This was my first thought, but I am thinking it switches too fast to be comm... Did you play the video?

I did change the comm loss values to reflect one pump "on" in VFD mode. It never seems to get that far. I also changed the minimum run time to 30 seconds... No change.

Thank you very much.

[amigo]

may be you are getting a spike through the AI/BIs when the gen set starts. Disconnect all the input blocks, run the test and see if that helps.
although that does not explain why the banging would cause it, anyway worth giving it a shot.

May be unrelated, but have you gone through every program in the PCM and see if it is intact? If you open up the program and see happy face symbols then you got corruption.
good luck.

[Tech Rob]

I wasn't able to play the video. It might be switching too fast to report the comm loss in summit, but maybe not too fast to make the PCM get stupid.

If you have access to an oscilloscope, then you should check the comm link with the scope while the generator is running. You've got mail, btw.

[t-reck]

To test if the problem is software or hardware, put the PCM into Manual mode Test, force that BOP on then repeat your bang test. The software cannot change the state of the point when in that mode.

[Tech Rob]

So what'd the problem end up being, GT?

[GT Jets]

So what'd the problem end up being, GT?

Well...... Let me just say this.

I have replaced the PCM, pulled new wire for the comm, hard jumped all the digital inputs, replaced the TCM used as a point multiplexer, isolated the cabinet..

Still did it.

Ultimately I used some elastomeric silicone to "glue" some screws to the back panel of the cabinet and mounted the controller on those screws to isolate the PCM from vibration. Now it only does it half the time.....


I got nothin' ...

GT

[joey791]

Any chance you have CTs on the pumps or anything hooked to the generator?

[GT Jets]

Any chance you have CTs on the pumps or anything hooked to the generator?

As a matter of fact there are some small CT's on the outlet of the starters for the pumps.

There is nothing hooked to the generator when the thing goes whack job... The generator is just doing an unloaded run test...

GT