Spacepak loses 24 volts after brief power glitch, returns if 240 volt breaker cycled off/on

I haven't read the entire post, but what is causing the brief/power glitch? Can that problem be solved first?

Is this glitch happening when a generator kicks on? Is there a delay in the generator's switchgear that shouldn't be there?

Why the power glitch?

Looking at the control board there appears to be three self-resetting fuses (Red arrow) and an MOV (Yellow arrow). If your AC power anomalies that upset the board are actually mild over-voltage events the circuitry on the control board may clamp the rising Voltage causing the self-resetting fuses to trip. Typically to reset a self-resetting fuse you have to remove the power source or the load, so the device can cool down.

I would also make sure the 208 / 230 to 24 VAC power transformer is using the correct primary side tap. If it is erroneously tapped at 208 the voltage to the control board maybe prematurely high to start with.

If one of the self-resetting fuses trips causing the control portion of the system and the thermostat to appear dead I would suspect you would find 24 VAC where the power enters the control board an a very low but maybe not zero voltage between R and C.

@Jon_23

With no disrespect.

Its always better to start with the simple things first. It will cost nothing but a little time to check with the neighbors about lights dimming and resetting stuff etc.

Is changing the control board or futzing around with the controls going to fix these problems?? Like the oven clock or the UPS??

 ""But the events exist. Sometimes the oven clocks need to be reset, often I can hear my 120 volt UPS's cycle on/of, lights dim or blink… Not all of these crash my air handler, just often enough to drive me crazy (or leave me with a very hot house if I"m returning from work).""

You said lots of wires running through the trees.

"

The apparent absence of 24 VAC (R to C on the control board) should be verified, since maybe the power anomaly is actually just messing with the thermostat only. Dead thermostat, nothing works. Not sure if you ruled this out yet, I did not re-read the whole thread.

I hate to add another Band-Aid type resolution to the situation, but I will. Since you stated the thermostat goes dead when the issue happens, if the 24 VAC (R to C on the control board) is truly absent, that absence of 24 VAC (R to C on the control board) could trigger a timer that would interrupt the power to the ICM493's contractor (T3 or T4) for a specific delay time, effectively resetting the Spacpak unit. So far I have not found a suitable inexpensive timer. The timer would be powered by the 24 VAC at T3, T4.

Instead of cycling the circuit breaker, just smack the unit. Its a little more humane than shock therapy with likely the same results.

If the Float Switch simply interrupted the 24 VAC to the entire board, two things, there would be no Fault Flash error code, since the board would be dead, and why would you need to cycle the 240 VAC to reset that fault.

I believe once the Float Switch is tripped (and even if then corrected), that trip event is captured by the control board's logic and locks out operation, so the unit now requires attention to restore operation, the only way to restore operation is to cycle the 240 VAC, which clears the logical fault state if it no longer exists.

And as I said before;

You would have to experiment. I found no mention in the documentation (and I may have
missed it) that claimed the thermostat would go dead if the Float
Switch was tripped.

The documentation could be wrong or simply left out some interesting details.

@Jon_23, I believe I have found a suitable inexpensive timer. The Omron H5CX-AD which was here under my nose all along (Omron H5CX-AD-G). It checks all the boxes for me.

24 VAC power.

Trigger signal input via dry contacts (from a relay).

Isolated relay output (to briefly interrupt 24 VAC power to the control Board).

An appropriate timer mode.

Inexpensive when strategically procured.

Understanding the 10 timer modes and their options is always interesting. It appears Mode A-1 with the output time interval set for a One-Shot time period would do what is needed to automatically reset the Control Board after a poor power quality event.

The way this would work, upon a poor power quality event and the control board cuts off power to the thermostat terminals shutting the AC system down. With a 24 VAC relay coil placed across the R and C terminals of the thermostat wiring, this constantly energizes the relay so its Normally Closed contacts are Open.

With a poor power quality event the relay's coil wired across the R and C of the thermostat wiring would loose power and the relay's Normally Closed contacts would then Close. This would trigger the timer, the timer would then start an On Delay, say 30 Seconds. When the 30 second delay expires the relay inside the timer changes state interrupting the power to the control board (directly or via the ICM493's contactor control). Also the timer's One-Shot delay is started. When the One-Shot delay expires the relay inside the timer changes state again restoring the power to the control board. Restoring the power to the control board should restore the power to the thermostat circuit and energize the trigger relay, disabling the timer's trigger signal. So the timer system sits and waits for another poor power quality event.

In practice I would probably have both delays at least about 60 Seconds each to let the refrigerant pressures equalize a bit.

If there is an actual Float Switch fault, the timer system would keep cycling trying to restore normal operation. However since the Float Switch is open the A/C will never come back on until the Float Switch situation is dealt with.

The 24 VAC power that powers the timer could be from the Spacepak unit, the ICM493 or its own dedicated transformer. The trigger relay should be as small as conveniently possible, I would not use a Fan Center relay which is huge for this application. There are some reasonably sized and priced relays with sockets on eBay.

@Jon_23, As I described your situation may be the perfect storm. So others may not experience what you do as often as you do, so it is not really on their radar. For example I am not far from my substation, my power never glitches, I can't remember the last time I had to reset a clock other than for a DTS time change. If your situation happened to me once every 3 years would I be complaining about it ?

Utility re-closure devices shut the power off as a result of an over-current fault. And it is off for a very noticeable amount of time, not just a glitch. When they re-close you should get a normal equipment restart. I believe you have a loose connection somewhere. The loose connection is agitated by the storm, probably wind related. Just replacing the aerial drop to your home without discovering a real defect is just lame in my opinion. It makes it look like they are being proactive but they really are not. Have they done an Infrared inspection of the entire primary back to the substation ? Also like monitoring the voltage during good weather, lame, monitor the voltage when a known bad storm is coming.

Without the utility doing a proper investigation to find out who is affected and then carefully inspect the wiring they may never find the issue. Bad weather may eventually rip down the bad section and when it is replaced the problem may be gone, but that is a poor way of repairing power quality issues.

And yes I understand it may take time for you to capture the next event and document the state your equipment is in when it happens.

In the mean time I'm just putting out some thoughts and ideas if the utility never repairs your issue properly.

any time I’ve seen a unit that needs a power shut down to restart that’s a safety shutdown. High or low voltage doesn’t matter.

How Linemen find Bad Connections - Thermal Camera

https://youtu.be/nhREcg_PyQU

Ahhh… I did not understand there was actually two different switches to detect condensate issues.

Either way if there is no satisfaction from the utility the timer method would still most likely work for an automatic and timely recovery. The timer's interruption point of the power to the A/C equipment may need to be different.

@109A_5 I'm having a hard time dismissing @Jon_23's observations.

But didn't he say there were no trouble codes?

Possibly, I'm not sure, I will go back and look. I thought he simply did not know, since he was just cycling the circuit breaker to restore the operation of the A/C.

Also up until recently I did not realize there is actuality two methods to detect a condensate overflow. I incorrectly thought the WetSwitch he mentioned in the past was the device connected to the Float Switch terminals on the control board.

Now I wonder if the two condensate switches are wired directly in series or if the WetSwitch actually interrupts the power to the control board or just interrupts the thermostat circuit.

EDIT:

@pecmsg I found no mention of any inspection for trouble codes. Just the access to the unit was a pain and/or the A/C interruption was inconvenient for troubleshooting, so the circuit breaker was just cycled.

Not sure what this follow up posting means ???

Are you agreeing with me or trying to point out something that I missed ???

I think I need to clarify something. I do NOT think the first place to look is the utility company. It's handy, and big, and distant, and therefore a nice target. It can't possibly my equipment!

Wrong.

We don't know that. We don't even know what it is about this particular individual unit is the problem, or whether it's a characteristic of the overall Spacepak design. We hear a lot about a float switch. But are we sure?

I will agree — quite happily — that intermittent failures are really really hard to find. I've chased a few from time to time.

So. A few more comments about the utility. What I think I'm hearing is that there is something, somewhere in the Spacepak control board or logic, which causes it to lose 24 VAC power after a brief power glitch. To be blunt, and I don't mean to dump on Spacepak, this simply should not be able to happen. But the first thing which needs to happen is to determine from the full control circuit diagram, what controls the 24 VAC output from the board; it's likely to be an integrated circuit controlling a relay, either ground side driven or power side driven. Then one needs to find out what the logic of that control is.

I can think of a number of possible scenarios where a timing conflict within the control logic could do what I THINK I'm hearing: a situation where some control does NOT reset properly after a cycle or two of power loss, but does reset properly after say 10 to 20 cycles. It would be nice to know what it was.

But in the meantime either come up with an on delay timer — which has been suggested and would probably work, but is a kludge, or stop blaming the power company or the float switch or warp drive or whatever and figure out the problem instead.

For what it's worth (nothing!) my money is on a bad solder connection on a capacitor in the board's power supply section…

I have been following this since the beginning and still cant believe the problem source has not been discovered. This reminds me of a rather similar situation form my time as a Lennox Dealer.

In the 1980s, Lennox Industries had a problem with an electronic ignition control on one of their furnace models. The furnace would fail to start after some unknown electrical event. Even though the thermostat was calling for heat and the power to the furnace was on, the burner would not operate.

Resetting the furnace by cycling the power off and then back on, solved the problem. The furnace would then operate as it should and cycle off by the thermostat, limits and everything else for days or even weeks before the glitch happened again, and you ended up with no heat until someone manually cycled the power.

Since this was a problem that many furnaces encountered, and the engineers at Lennox Industries could not easily correct it, they came up with a Band-Aid-type fix. They added, at no cost to the consumer, a duty cycle that would break the thermostat circuit every hour if the thermostat had been calling for heat that long. It was called the Watchguard control WC-1

Since most calls for heat were less than one hour, the consumer never actually knew it was happening because the duty cycle only started after one hour of continuous calling for heat. Although Lennox no longer offers that control, you can find it on sites like eBay that offer New Old Stock. Here is one you can get right now (August 2026): eBay Lennox Watchguard Control WC-1.

Since it was good enough to solve the Lennox ignition failure problem, I think you can use it to restart your AC whenever it fails. The only difference; I would add is a 5-minute time delay for the compressor contactor, since some air-conditioning units do run nonstop for several hours during extreme weather. You don’t want the short off-and-on interruption of the 24 volts to cause a compressor event as a result of this device.

@EdTheHeaterMan, as much as I considered a cyclic solution like your Lennox example, I did not like it so I came up with an inexpensive solution that only activates when actually needed.

Since the control board could be the issue or the WetSwitch and the ICM493 now exists in the basement, I would simply (unless not practical) extent the thermostat circuit's R and C wires to the basement and add the timer circuit I illustrated above to briefly interrupt ICM493's contactor control. Easy, at least for me from here.

I think @Jamie Hall missed a bit of the posts in the middle of the thread. @Jon_23 amusingly this thread is short compared to some. You seem correct, folks appear to be missing information.

In a situation like this, my impatience and aggravation may overrule the my strong curiosity's need for data.

I like evidence based troubleshooting.