invensys insomnia 9700i stuck on

archibald tuttle

New source of insomnia discovered. I moved a Robertshaw/Invensys 9700i yesterday to a temporary location in a cold room mounted on foil faced insulation board. It is using the remote sensor feature and had previously been mounted in the basement not far from the boiler in unconditioned space somewhat warmer from castoff boiler heat, but not as warm as the conditioned space. I did not move the sensor which was in an apartment and the setup had been working like clockwork.



These thermostats have split system with a circuit board in the base and in the removable front section with the control interface. You pull off this front interface to access the wiring connections. Insofar as I recall the thermostat does not call if this control module is removed, although I think what qualifies as relays are in the base, there is nothing to pull them in when you have the front section removed.



When I put it back together it was registering the remote sensor at 63.8 degs. (know that had to be the remote because the location of the stat was in 40s). Thus I expected it to be calling for heat we had moved out of the daytime setback into the evening call which is set for 68 deg.. So I thought nothing of it calling the boiler, other than another job well done and headed home, only to get a call at 4 in the morning that the boiler was running constantly and it was 80 in there.



I figured maybe I had made a mistake extending the wires to the new location and actually wired the TT loop closed, but checked everything as correctly wired. The stat registered temperature of 79.2 deg. and was set for 62. I pulled the front section off and the boiler continued to run, I took a wire off one terminal and the boiler shut down.



So it appears that the relevant relay in the base is stuck on. I tried the obligatory bigger hammer (well not so big for this light electronics stuff), tapping what I assumed to be the relays.



It is a plastic base like virtually all the programable stats and couldn't see any possibilty of contact through screw with the foil faced insulation, but I pulled it off the wall to double check. No change.



Out of ideas other than round up another stat as a double check. Anyone ever experience this or have any other ideas.



Exasperated in Exeter

Tim McElwain

What size wire did you

use for the extension? If anything less than 18 guage it could be your problem.

archibald tuttle

Tim [the enchanter]

thanks for thinking of me. Used the genuine item, 18/4 thermostat wire. 2 for temp sensor, 2 for TT.



NAbbed one off another application where I could replace it with a standard $30 setback stat it di and put it in yesterday. It didn't repeat the locked on status, but whereas the first one gave rock steady temp. readings from the remote sensor but had the TT locked on, now I'm getting control of the boiler but the remote sensor readings are steady for several minutes and then jump a degree or two and then maybe 10 or 15 seconds jump back down.



I checked to see if the tenant might have opened a window because of the overheat the night before plus it got warm here suddenly (in the upper 50s this morning when I got up).



I ran half the cable while I had guy in the crawl space running the other half. It's all one piece and, as I said, temp readings solid as a rock on the 1st misfunctioning stat, and now jumping on the second one.



I'm going to have him crawl under there just to make sure he didn't catch the cable with a staple or something, but the lock on was definitely not a cable problem. I pulled the RH wire and the boiler went off, no hesitation and the W to RH on that old state base is made and remained made after I took the stat down and filed it under ask Robershaw what the hell is up with this.



None of it makes any sense to me and I think it is just murphy screwing with my vacation.



The only other unusual deal here is that I mounted the stat on the foil face of a layer of isocynanate. There is no contact with any of the boards or wires or anything that I can possibly imagine, but maybe that proximity to a sheet of conductor is making this wacky, although the symptoms from the two stats I've tried are completely different and at least in the first case didn't abate when I took it off the wall.



Go figure.



Brian

Tim McElwain

Are you perhaps getting

some interference from line voltage wires you may have come in contact with while running the wire? Also how far are you running the 18 gauge wire?

archibald tuttle

tim et al, insomnia cure, cut freeze protector off circuit board

Sorry I didn't get back sooner, went back to enjoying florida after coordinating this fix via laptop.



The answer to the problem is:



The 9700 (and I presume 9701 and probably a;; the 9700s and 9800s and maybe the 9600s) have a circuit breaker style (really circuit maker style) powerless freeze and overheat protector.



Which is kind of a safety against low battery, but if you are using the remote sensor and the thermostat itself is located in an unconditioned space . . .



you get the point.



So I attached a photo of the inside of the stat base.  After a 3 hours break for consultation amongst their techs and engineers I got a call back from Invensys that the low temp protector is the compenent PT2, the large 1/4" by 5/8"  at the top left of the circuit board in the base of the unit. It makes without batteries, without the front of the thermostat installed, regardless of sensor choice when the temp at the base falls below 50 (exact temp is maybe a little variable, and I never ran tests). Just had one of my techs on scene cut the component from the board, which fortunately they are installed with accessible lengths of wire for removal. Voila, nightmare on elm street solved.  PT1 underneath it covered by a plastic sleeve is the equivalent control for overheating (I assume but I didn't ask, that PT 1 makes W to RC on rise between 80 and 90 deg. but I'm not using in cooling application and don't take that for gospel). 



Thanks again for considering the problem. It was about a 50 foot run of 18 gauge wire. I though maybe the aluminum face of the insulation was creating a time warp or something, but it all makes sense now and I thank invensys guys for running this down for me, although maybe they could have had this in bigger type. I often install these thermostats in unconditioned space although not usually in as cold a location as this one. Probably, a few of them have triggered before but they were in basements that got enough cast off heat that it wasn't perceived as overrunning.



Brian

Tim McElwain

Sometimes it

is the simplest of things. Glad you found the problem and cure.