How to Make a 24V Transformer Output 24V?

[russ_watters]

TL;DR Summary

HVAC unit is throwing an over/under voltage fault. 24V transformer reads 26V. How do I get it down to 24V?

My friend bought a condo last winter and now in cooling season her AC unit (water source heat pump) is faulting-off sporadically. I've done some basic troubleshooting, and the unit is showing a over/under voltage fault on its LED. It's a 208V/1ph unit. According to the manual, the controller reads the control voltage and figures if it's out of range that means the primary voltage is out of range. Control voltage is 24VAC nominal and the alarm limits say 18-31.5V. In non-coincident measurements, I found the line voltage to be 208V on the dot and control 25.6V. That doesn't seem like it's far off, and is within tolerance. But the problem seems to occur more often at night, when I'd expect the line voltage to un-sag and go up some. Old building, old city - I bet it varies quite a bit with load and I just happened to measure it when it was exactly right.

I was going to leave this to a service tech, but I'm having trouble finding one to service a WSHP. And the more I look at this, the more it seems it may be a simple fix. In reading-up on little transformers, I see it is common for them to have a low-load voltage significantly above nominal and then just let them sag. So I'm thinking about three potential options:

1. Replace the control transformer.
2. Add a resistor in series to drop the control voltage.
3. Switch the control transformer tap from 208 to 240V to drop the control voltage to about 22.2V.

For #1, the transformer is cheap and looks easy to replace. But if it's common to have a real voltage above the nominal I may just end up with the same problem. For #2, the resistance depends on the load, right? It's a 50 VA transformer, but is likely not using anywhere near that much. I calculate that if it's using 20 VA then a 2.6 ohm resistor would drop about 2V. Is it really that simple or am I missing something? Risks?

It's a ClimateMaster CXM controller. Several PDFs on the issue at the top of the google search HERE. There's a bit of conflicting info on the fault; it says 10% tolerance, but the actual voltage range given is more like 30%. The existing transformer is a pretty standard multi-tap HVAC control transformer, except that it's 50 VA and most I see when googling are 40. It does appear to be wired properly for its voltage.

If need be I can install a voltage logger on her incoming power service to find out what the voltage is really doing over time. For example, since yesterday my household voltage has been varying between 120V in the day and 125V at night.

Other thoughts?

 

[Bystander]

Summary: HVAC unit is throwing an over/under voltage fault. 24V transformer reads 26V. How do I get it down to 24V?

Other thoughts?

How old?

 

[russ_watters]

How old?

2006/16yrs

 

[Bystander]

2006

East coast, high humidity, 16 yrs. old? Any signs of arc over?

 

[russ_watters]

East coast, high humidity, 16 yrs. old? Any signs of arc over?

On what and what would the signs be?

Also, I'm not sure the installation location qualifies as "high humidity" -- it's a water source heat pump so there's no outdoor part. Unless the prior owners always had the windows open throughout the summer.

 

[Bystander]

On what and what would the signs be?

The transformer; operating on mental image of doorbell in my house, could NOT keep it working, just cheap U.S Home crap; 120 stepped down to 24, and mounted to "engineered forest products" in the basement, highlow to medium-resistance leakage paths all over, burning everything, and humming, loose laminations in the core.

Just looked over the sheet you referenced/furnished; looks like "digital hubris" and the age is about right, older is over-engineered low-tech, and newer is "humble-pie lessons-learned" tech.

Try swapping the transformer, and if that doesn't work... You might think about junking the controller; YMMV.

 

[Baluncore]

1. Replace the control transformer.
2. Add a resistor in series to drop the control voltage.
3. Switch the control transformer tap from 208 to 240V to drop the control voltage to about 22.2V.

1. Other transformers will have the same turns ratio.
2. Avoid a resistor. In place of a resistor, get a rectifier bridge that will handle the current, short the DC + and - terminals, and insert it in series with the LV AC line. It will drop about 1.5 volts from the secondary AC due to the two diode drops in the bridge.
3. Tap switching is a possibility, but make sure you flag the change and explain why. When the voltage is corrected, it will be under voltage.

 

[Tom.G]

Summary: HVAC unit is throwing an over/under voltage fault. 24V transformer reads 26V. How do I get it down to 24V?

If need be I can install a voltage logger on her incoming power service to find out what the voltage is really doing over time.

I would start with your first choice:
1. Replace the control transformer.

Then:
Power companies here in the USA will often install a logger if you make enough noise.

Though one item in one residence may not be enough inducement. See if you can find any neighbors with problems and give it a shot.

It could also be a bad connection anywhere between the heat pump and the distribution transformer on the power pole. For instance a loose Common connection on a 120V/240V supply can cause undervoltage on one leg (phase) and overvoltage on the other one. Which leg is high (low) depends on which one has the lower (higher) load at any particular time.

Somewhat Relevant Story
I'm within walking distance of the Pacific Ocean here. A few years ago the power company had a somewhat similar problem in a 3 block area. It turned out that some splices on the 16kV 3-phase lines (that's the primary side of the consumer distribution transformers) had corroded. It took them three tries to find all of them. they would repair one set and several hours later some splice further up the line would fail.

(Customers too were getting... shall we say... a bit annoyed.)

A conversation with one of the repair crew revealed that the original splices, which were Aluminum, are now replaced with Copper.

Cheers,
Tom

p.s. Please keep us updated on what you find.

 

[George Harlan]

As a power engineer I vote for the tap change as the least likely to cause harm. My suspicion is, however, aging caps in the power supply filter causing sags in the rectified transformer output.
I base this on the information that this is a probably a new problem (nine years is a long time to live with an issue like this), where as a site over voltage is not a likely issue to suddenly occur (although possible).
If the problem occurs more frequently after the tap change then I would definitely check out the filter caps.

 

[Tom.G]

I just looked at the manual for the controller. The cover photo of the controller shows two Molex connectors, those white ones along the top edge of the board.

Those Molex connectors are infamous for corrosion/oxidation that causes intermittent connections!

The usual 'quick-and-dirty' field repair is:

1. unplug them
2. wipe the pins with bond paper (NOT newspaper, too much Sulfur in it that causes further corrosion)
3. liberally spray both the male and female connectors with contact cleaner
4. mate & unmate the connector several to provide some scrubbing action
5. spray again to remove the loosened crude
6. wipe with bond paper again
7. blow the excess contact cleaner out of the female connector
8. spray lightly and re-assemble
9. while crossing fingers and eyes, give it a try

If failure recurs after 3 attempts of the above, either replace the connectors or look somewhere else for the problem.

The female connector contacts are crimped onto the wires. It is possible to solder the new contacts to avoid purchasing a crimp tool. The crimp flags of the contacts must be bent over the wire to allow insertion into the connector body though.

The male pins on the circuit board are difficult to remove without damaging the board... and the main trick is Do Not Apply High Force Or Scrubbing Action to the Circuit Traces While Hot! The glue that holds the traces to the board softens with temperature and you will damage/remove the trace.

Double-sided board
Removal is often done one or two pins at a time by cutting the Nylon body that holds the pins; be careful not to cut any traces that may be under the connector. This also requires a large size Solder Sucker to free the pins without board damage; you must get all the solder out of the board holes.

Single-sided board
For the Solder Sucker that is needed with a double-sided board, you can substitute SoderWik (Copper braid with soldering flux embedded in it). It soaks up the molten solder for removal.

Sometimes, you can get a few weeks/months use just by a few unplug/replug cycles of the connectors; it won't last but is worth a try.

Please let us know how it goes... and
Good Luck.

Tom

 

[Rive]

Just for the log: a failure sign of certain cheap, small transformers (mostly wall-plugs or few VA sizes) can be the rise of output voltage due turn-shorts on the primary side. Often caused and comes with increased heating.
But the size (50VA) is above the usual range of this kind of problem => unlikely.

 

[russ_watters]

Thanks everyone. I bought a new transformer. It's a much smaller commitment than buying a new controller board ($30 vs $300) for an uncertain fix. I'll let you know how it goes.

I'm also going to plug in my Kill-a-Watt to watch the house voltage (rather than take the effort of installing my whole house power monitor).

Photos of the controller/xfmr and wiring diagram, if these help.

 

[russ_watters]

As a power engineer I vote for the tap change as the least likely to cause harm. My suspicion is, however, aging caps in the power supply filter causing sags in the rectified transformer output.
I base this on the information that this is a probably a new problem (nine years is a long time to live with an issue like this), where as a site over voltage is not a likely issue to suddenly occur (although possible).
If the problem occurs more frequently after the tap change then I would definitely check out the filter caps.

Speculative, but the current thermostat is not the same one that the previous tenants used, it's a basic programmable whereas they were using a Nest (they took it with them and re-installed the old one). Perhaps a Nest or other modern one uses more power and pulled down the voltage.

 

[Baluncore]

Photos of the controller/xfmr and wiring diagram, if these help.

That circuit shows TRANS (note 3), the 24 V secondary transformer with;
Black com 0V, Orange 240V, Red 208V, and Brn 265V.
But it appears they have confused or swapped the tap positions.

 

[russ_watters]

That circuit shows TRANS (note 3), the 24 V secondary transformer with;
Black com 0V, Orange 240V, Red 208V, and Brn 265V.
But it appears they have confused or swapped the tap positions.

In what way, exactly, does it look confused? It's tough to see in that photo, but it's the red wire connected to incoming power that goes to the transformer. That's correct; 208V. The tied-off wire (covered with electrical tape) is the orange wire. I don't see a brown wire.

Or are you just saying the diagram looks wrong? I don't think that should matter.

 

[Baluncore]

Or are you just saying the diagram looks wrong?

The primary and secondary should both have the same volts per turn. But the 208 V tap is at the far end of the primary winding, while the 240 V tap is part way along. That is backwards.

I would check that the windings were correctly terminated, by measuring the voltage on the other available taps.

 

[russ_watters]

Plot twist: it looks like the problem is not electrical.

Well that didn't go at all as expected. I got the transformer and before swapping it out I did some more electrical testing: 207V primary, 26.3V secondary with unit off, very little change when unit turns on. Doesn't look like an electrical fault, and I didn't replace the transformer. Incidentally I hadn't noticed before, but there are some scorch marks on the upper right of the controller which led me to think it was going bad and needed to be replaced. But first...

I dug a bit more into the fault codes. It's a bit confusing, but the slow flash doesn't just indicate a voltage problem. It can also indicate "fault retry", which just means "I got a fault, but I'm going to try again to run in 5 minutes". Otherwise the controller doesn't actually report fault codes on the fly. There's a couple of terminals that if you short them puts it into "Test Mode", which then reads out the stored faults. I did so and got code 6: CO fault. No, not carbon monoxide: condensate. There's a high level sensor in the condensate pan. It's a real pain to get to, but I did and sure enough, it was touching the water. And man, was the pan and back of the coil nasty. I scraped and cleaned as best I could which wasn't much because I couldn't reach the drain from inside - it's behind the fan. And I disconnected the level alarm. Theory being it's probably just draining slow, not really getting set to overflow. As a backup I put a water alarm (from the water heater pan) into the unit and I'll also check from the filter side where you can see into the pan. Also, it does have a secondary pan under the unit. Tomorrow I'll spray on some coil cleaner, pour in some Draino and disconnect the condensate drain pipe to try to clean it from the outside. And re-connect the level sensor.

For now, the unit is cooling/maintaining temperature for the first time today.

 

[russ_watters]

Spoke too soon on the temp fix. Drain pan is full/overflowing. I'll have to leave it off until tomorrow when I can get tools to disconnect the pipe and clean it from the outside.

 

[Tom.G]

Spoke too soon on the temp fix. Drain pan is full/overflowing. I'll have to leave it off until tomorrow when I can get tools to disconnect the pipe and clean it from the outside.

A little late, but...
A wet towel (or any woven material) can be used as a siphon that will restart as long as it doesn't dry out. Might be handy to keep the system running until the drain is cleared.

Cheers,
Tom