HAI 'stat troubleshooting help

I've recently started graphing temperature data throughout my house, including the thermostat's current temperature value. It's been a great tool to be able to visually note what is happening with my HVAC system and general comfort. Turns out, it's also revealed an issue that I didn't notice before. On a very inconsistent basis, the thermostat is "dropping out" and I am logging a temperature of -40 degrees. The drop is very intermittent. One day it will be stable all day and the next it may drop 2-3 times. Having seen this behavior via the data logging, I have tried to catch the thermostat red handed and I have. Twice, I have been able to get to the 'stat when it drops out (which typically lasts for a few seconds to a minute). What I have noted is that the display is dead but that it's relays are still energized. Meaning that if it's calling for heat when it goes blank, the relay's wiper remains in the same position and the call for heat is maintained. If I pull the face of the 'stat off the wall, of course the circuit is broken and the call for heat is ended. When I place the 'stat back on the wall at that point (at least the two times I've caught it), it fires back up. It's the danged sporadic nature of the problem that makes it more difficult to troubleshoot. My gut is thinking that there is an issue with current draw as that has always been an issue with HAI thermostats but, why then wouldn't it always be a problem? I do have current stealing via the additional common properly wired (this isn't my first rodeo). Anyone else care to chime in and steer me elsewhere before I drop time and money into a larger transformer?

As it turns out, my water heater $hit the bed and today is the day for replacement (minor flood in the mech. room). As long as I'm in there, I'm just going to go ahead and replace the transformer. Hopefully that will get it.

I went ahead and "hot rodded" my furnace's 24VAC transformer. Went from a factory installed 40VA model to a spare 96VA model I had collecting dust. Now I'll just keep an eye on the log to see if I'm still getting sporadic drops. If my lousy math skills are up to par, I should jump from 1.66 amps available to 4 amps. Can't hurt.

No love. I reviewed the data logging this morning to find that, despite the increase in available current, the stat is still dropping out. Any ideas? I may have to just shoot an email to HAI.

Very strange problem. Is there anything else connected to that Xfrmr (another t-stat?)? Maybe it would be worth it to ring out the cable for leakage and continuity (high resistance in a line?)? Are you getting full voltage to the t-stat?

BraveSirRobbin said:

Very strange problem. Is there anything else connected to that Xfrmr (another t-stat?)? Maybe it would be worth it to ring out the cable for leakage and continuity (high resistance in a line?)? Are you getting full voltage to the t-stat?

Click to expand...

I will put the DMM into service and check these items. However, define "leakage". I'm guessing you mean voltage drop but, it's AC and a short (approx. 20') run so I doubt that's an issue.

Disconnect both ends of the cable and make sure none of the cable ends are touching anything or each other. Then check for continuity using the multimeter in ohms mode. In long runs I measure between two cables with the meter on one end and just jumper those same two cables together on the other end.

The meter should be in the lowest ohms setting (unless it's an auto scaling meter then it doesn't matter). This will insure no 'breaks' are in the cable and the resistance readings should be approximately the same.

You also want to measure for "leakage", which makes sure the insulation is intact and the cables in the bundle have infinite resistance to each other. The reason this is important is in cases where a staple was driven through a cable to secure it to a stud, but it was driven to hard and it made contact with the cables inside the bundle (or crushed them compromising the insulation). The continuity test above may not show if this problem exists (I had it happen to me with 22 gauge four conductor cable on a motion detector cable run by the builder).

On one end of the cable make sure the individual cables are not touching each other or anything else. Then on the other end measure between two cables with the meter on the highest ohms setting. The reading should be infinite on all possible cable combinations.

As always make sure your fingers are not touching the meter's tips when performing the above measurements as your skin will influence the resistance readings (especially when performing leakage tests). You can quickly test your meter by shorting the tip's ends (near zero reading for continuity) then separate the tips and the resistance should be infinite.

OK. Yeah, I am very comfortable with the meter, I had just never heard the term "leakage" before. I will test tonight and report back. Neither issue seems likely, though. If I had lost continuity, the stat wouldn't function and if I was shorting (leakage) AC, I'd have blown the fuse. Weird resistance is what I will be focused on looking for. As a side note, I haven't seen the stat drop since mid day Sunday (nearly three days now).

Quick update. I have checked all wiring and it's all solid. While I can't explain the hiccup that I reported (stat dropped out last Saturday, AFTER the new transformer install), it has run well for seven days now. No drops, though I am keeping an eye on it.