Actron AM24-1 Controller Possible Thermistor Replacement.

**tready** · 30-01-21, 02:46 PM

Have an Actron AirCon unit that had decided it thinks the temp is 39C all the time, I have tried resetting the power etc to the unit nothing helps, even did a reset on the controller its self to factory settings.

I suspect the thermistor may have gone on the controller I took it off the wall today and there was a lot of dust around the edge of the board and a pantry moth dead inside, I can see some light corrosion around the thermistor and the legs were covered in dust so I suspect that moisture might have played apart in playing up as its been fine, But the last 2 days its been raining nonstop and really humid so the dust and moisture could of shorted it out.

New controllers are a good $200+ and they no longer manufacture this version anymore they have a new version which should work, but unsure if it also requires a cpu board to be changed too,

Is there a way to safely test component without damaging the board? Is it as simple as shorting out the contacts and see if it affects the temp registered or can I test the resistance using a multimeter. I have limited knowledge when it comes to component level stuff. But willing to give it a go it not so much a problem now in summer as it will cool but in winter I wont be able to use heat as it will be reading above any temp I can set it too. And if it turns out to be thermistor how will I know which one to get.

Photo here:

Thanks for any valuable advice you can help with.

**hinekadon** · 30-01-21, 03:38 PM

there should be numbers on it

**wotnot** · 30-01-21, 05:02 PM

Sanity check first -- NTC thermistor, ergo resistance decreases with increasing temperature ; I would imagine 39C would be the unit's max real life scenario ; thermistor would be a near short --- check; unit off, multimeter on thermistor, read resistance, cool thermistor (cold spray, ice-cube), remeasure resistance -- does it still behave like a thermistor? (range will be around a few hundred ohms (hot) to a few KOhms when cold)

**tready** · 30-01-21, 08:24 PM

Originally Posted by wotnot

Sanity check first -- NTC thermistor, ergo resistance decreases with increasing temperature ; I would imagine 39C would be the unit's max real life scenario ; thermistor would be a near short --- check; unit off, multimeter on thermistor, read resistance, cool thermistor (cold spray, ice-cube), remeasure resistance -- does it still behave like a thermistor? (range will be around a few hundred ohms (hot) to a few KOhms when cold)

Thanks I will check that tomorrow now I know what type of numbers I'm looking for doesn't sound good though, I checked for continuity earlier and there was continuity with a reading of 006 on cheap DMM that I have but will still check actual ohm readings.

Another thing I noticed when someone else asked for photos of the wiring connections on another forum was that the yellow sensor wire is a little crimped and wondering if there is a break in it causing problems, As I don't know if the controller reads the temp directly or it sends the sensor back to the CPU board in the indoor unit and the controller is reading the data from there. As you can have external sensors wired in place of the controllers one so if that wire is damaged and not sending data back to the cpu board the controller is going to read bad data from it. I uploaded more pictures to the onedrive link showing the damage.

**RFI-EMI-GUY** · 31-01-21, 04:42 AM

A thermistor will have a resistance spec that is published. There are two numbers, a polynomial to draw a graph. If you have the center point, usually room temperature, you can grab a spare resistor near that value and solder it in temporarily replacing the old part. If the temperature reads nearly correct. Buy the replacement thermistor and solder it in.

Sent from my SM-T350 using Tapatalk

**wotnot** · 31-01-21, 12:57 PM

Originally Posted by RFI-EMI-GUY

A thermistor will have a resistance spec that is published. There are two numbers, a polynomial to draw a graph. If you have the center point, usually room temperature, you can grab a spare resistor near that value and solder it in temporarily replacing the old part. If the temperature reads nearly correct. Buy the replacement thermistor and solder it in.

This is true - if you can get a datapoint (roomtemp @25C might be a good one), you can figure out the rest ~ plenty of graphs out there (and most parts have plots in the datasheet)

A lot depends on how they've used it (in circuit) ..it's not unusual to see thermistors as part of a resistor divider network, and in that case if the pullup power has failed, you would get abnormally low signal volts (as in, none =)

Can't really see if the external sensor is separate to the onboard thermistor ; can see the 2K7 resistor is in series with thermistor and then to ground ; if both thermistors are in series, that 2K7 resistor might be there to create a voltage differential between the 2, so the MCU can conclude which is internal/external sensor,

the yellow sensor wire is a little crimped and wondering if there is a break in it causing problems

If in doubt, meter to check continuity (or strip the wire at that point to visual conductors) -- have a closer look at the PCB (or use multimeter), to try divine if the thermistors are part of the same circuit, or individual sensor inputs.

**tready** · 31-01-21, 06:36 PM

Thanks for the reply's guys looks like the thermistor is dead I am getting a reading of 15.7ohms across it with no fluctuation in heating or cooling it, When testing continuity I am getting a beep and reading of 015. Now the question is how do I know what type of thermistor to replace it with there is no marking on the current one other then its blue and very small.

**loopyloo** · 31-01-21, 07:30 PM

I would unsolder one of the thermistor legs and measure again.

**tristen** · 31-01-21, 07:36 PM

Have you tried contacting the manufacturer-distributor actronair.com.au?

Customer Service Enquiries 1300 522 722

Also, do a little of your own research via a SEARCH engine, such as

**tristen** · 31-01-21, 07:38 PM

Originally Posted by loopyloo

I would unsolder one of the thermistor legs and measure again.

Good advice.

You CANNOT check it accurately while it is still in circuit.

**wotnot** · 31-01-21, 07:51 PM

Ugh....okay, made in china unit, thermistor with no markings, probably means they got it made custom, and saved 0.5cents per part ..like this ;

If it's made in china, that's what happens =) They *might* have used the same thermistor on the external sensor -- measure it's resistance.

Else you're left with something like removing dud thermistor, and soldering a trimpot in it's place, to find the center-point (as explained by RFI-EMI-GUY) ; you might start with a 5K trimpot (multiturn Cemets are good for this)-- 15R @ 39C would seem hinky, so my guess is the unit controller has hit max @ 39C, and 15ohms isn't representative to that temperature. (the thermistor is shorted, but in normal operation 39C may be 300ohms or something)

That would be something I'd look for....the floor value of the thermistor for the controller to have it register 39C ....ie; set trimpot to 15ohms and gradually increase resistance until displayed temperature moves from 39->38C, and measure actual resistance of trimpot --- then you can keep increasing resistance & watch displayed temp, and rough-plot the resistance vs displayed temp, so you end up with something like this (you prolly only need plot @ 5C increments)

Once you've got your little plot of resistance vs displayed temp by the controller (especially min/max temps the controller will display), you'll have a pretty good clue as to what value resistance/response curve/range of thermistor to look for. Pretty much all of them have datasheets detailing the above about device behaviour.

**tready** · 31-01-21, 10:09 PM

Originally Posted by tristen

Have you tried contacting the manufacturer-distributor actronair.com.au?

Customer Service Enquiries 1300 522 722

Also, do a little of your own research via a SEARCH engine, such as

Only discovered over the weekend haven't had a chance to ring them yet, But I'm sure they will be happy to sell me a new controller is what the end result will be. And yes I have searched but haven't been able to find detailed information about this one in particular other then its based of Lesam Controls model which I then discovered became Actron Controls and the bracket is labled OEM Electronics PTY LTD Sydney Australia. Only found the wiring diagram and how to enter some calibration modes, and activating zones and choosing sensor that's in use.

**tready** · 31-01-21, 10:16 PM

Originally Posted by wotnot

Ugh....okay, made in china unit, thermistor with no markings, probably means they got it made custom, and saved 0.5cents per part ..like this ;

If it's made in china, that's what happens =) They *might* have used the same thermistor on the external sensor -- measure it's resistance.

Thanks for that information think its going to end up in the bin at the end of the day I am tempted to take it out of the circuit but don't want to destroy it completely so we have no cool or heat at all and then be stuck.

**RFI-EMI-GUY** · 01-02-21, 03:29 AM

Cant see your pictures in one drive, so I am just throwing this out.

Sometimes a common silicon diode is used as a "thermistor". I had an odd temperature gauge missing its remote sensor. I grabbed a diode, stuck it in the terminals and it has worked well.

Sent from my SM-T350 using Tapatalk

**wotnot** · 01-02-21, 09:56 AM

Originally Posted by RFI-EMI-GUY

Cant see your pictures in one drive, so I am just throwing this out.

Sometimes a common silicon diode is used as a "thermistor". I had an odd temperature gauge missing its remote sensor. I grabbed a diode, stuck it in the terminals and it has worked well.

Just so you know (although seems the OP isn't going to have at it =)

**hinekadon** · 01-02-21, 01:13 PM

I would have a educated guess at a 40 degree ntc thermister and they are real cheap so jaycarit

**wotnot** · 01-02-21, 04:40 PM

I would unsolder one of the thermistor legs and measure again.

Pointless exercise here (unless PCB is shorted at that location, which doesn't appear to be the case going by pics), because you've got a 2K7 resistor in series (and I've yet to see one of these go short circuit), so path to MCU will always be that..

I would have a educated guess at a 40 degree ntc thermister and they are real cheap so jaycarit

No, that's not how it works -- as RFI-EMI-GUY & I keep advising, you need to know the value @ 25C to figure out the replacement -- this is ostensibly how thermistors are rated ..ie;

Go have a look at the specs of those 1,260 thermistors -- as a selection guide, you first filter temp coefficient type (NTC in this case), and the next most meaningful criteria is resistance @ 25C -- this is why thermistors are often quoted as being 1K, 2K, 5K, 10K (or any other value) types, it refers to the device resistance @ 25C. The next choice criteria, is thermal coefficient (% of resistance change per degree C)...which is why I like to know the range of the controller displayed temperature versus thermistor resistance.

**tready** · 01-02-21, 07:13 PM

Spoke to Actron support today as soon I said what it was doing he straight away said the sensor has failed in the controller and needs to be replaced. Got a quote from the local Actron dealer since they wont supply it direct they want $480 to come and do service and replacement of the controller. The controller its self seems to be around the $300 on ebay so the quote is well within reason.

Tracked down a similar one that's used on ebay was listed for $99 offered $80 and they accepted it, Its an AM7-1 which is the same but has 7 day timer programming on it I asked the guy on the phone would it work without issues and he said yes.

Worst case I loose $80 and still have to pay the dealers price best case it works and is an upgrade plus I can get some type of resistance reference maybe and confirm my original diagnosis, then look into potently fixing the other one.

Once again thanks for the input guys much apprenticed.

**loopyloo** · 02-02-21, 06:54 PM

Sounds like you've had a win.

**tready** · 10-02-21, 12:26 PM

Okay so I got the new controller and the system is functioning again and displaying room temps again but is not as accurate as the old one was seems to be about 3 degree out around mid 20's comparing to a digital and analogue thermometer. Previous controller was pretty much bang on within half a degree. So just have to keep that in mind when setting the set point.

Anyway to the more important information I measured the resistance on the replacement I got and its 8 kΩ at 23.C ambient temp based of my thermometers which the controller would see as about 20.C. The resistance value decreases when temp is increased is this enough info to work out a rough replacement value thermistor?