intermittant air con

**loopyloo** · 15-02-18, 03:20 PM

Having a little trouble with my air, con it's an LG LS-K2463HL split system.
Problem is intermittent response from the remote.
It was installed around 18 years ago and has been a beauty.
About 12 years ago, it wouldn't turn on. After cleaning the ir receiver it worked normally for another say 6 years and the same process was repeated with success.
All that time I always wondered if it really was the sensor at fault.

A few days ago it happened again but this time no guaranteed fix.
Cleaned the sensor and it worked til the next day.
Turned off the mains cos I intended to remove the board but didn't get around to it. Several hours later I turned on the mains again and it worked via the remote all night.
Next day no good again so turned off the mains for a few minutes then back on thinking reset but no good. Got up an wriggled the sensor board's plug then it worked again.
A little later it couldn't be turned off and wriggling the plug doesn't work any more, so since then have been using the over-ride switch to run it. Been pretty bloody hot here.
Last night I removed the sensor board's plug pins and squeezed them in to ensure contact to the pins but to no avail.
What's the go with these types of sensors (pic from google) ? hard to clean with that cage on them.

**loopyloo** · 15-02-18, 04:35 PM

This morning it worked with the remote again and left it on for 3 hours.
Later (after a couple of hours break) it would not switch on with the remote. I started thinking maybe moisture somewhere is not helping and hit it with a hair dryer. That didn't work.
So I switched it with over-ride and run it for a couple of hours. By then I figured I'd try something else. Turned off the over-ride switch and waited 5 mins, if it doesn't turn on with the remote I'll hit the main PCB with freezer spray but yep you guessed it, the remote worked again.
At least I know what to try next time the remote is ignored.

**Reschs** · 15-02-18, 10:10 PM

The real problem is corrosion caused by condensation.
Usually the IR Sensor is mounted on it's own PCB as a module with a metal shield.
Condensation corrodes the connections and PCB tracks.
I have repaired many of these.

**Reschs** · 15-02-18, 11:02 PM

The other obvious choice is the remote it's self.
Double check for bad battery contacts from leaking batteries.

Another problem can be IR radiation from other sources overloading the receiver.
Usually from so called Energy Saver Light bulbs. Unlikely as this occurs during daylight hours.

**loopyloo** · 16-02-18, 08:18 AM

Originally Posted by Reschs

The other obvious choice is the remote it's self.
Double check for bad battery contacts from leaking batteries.

Another problem can be IR radiation from other sources overloading the receiver.
Usually from so called Energy Saver Light bulbs. Unlikely as this occurs during daylight hours.

Ah yes, you touched on 2 things I forgot to mention before.
I checked the remote with a camera as well as change batteries (twice) and also downloaded a remote app to my phone and used it each time of fail. When the remote didn't work the phone also didn't work and of course when all was normal both remotes worked.

Failure to respond to the remote happened day or night.

Yesterday afternoon and all night the air con responded normally. At the moment I'm just waiting for the next fail to try the freezer spray.

I'm still trying to find a good circuit diagram for it. I found a service manual but the circuit page must be a photo of the circuit cos when I zoom in it just turns to blur and is totally unreadable.

**gordon_s1942** · 16-02-18, 01:03 PM

I use the AV room senders to test any of my 1000 remotes (seems that many).
The Sender blinks when it receives a signal and so far has responded to all of mine.

**loopyloo** · 16-02-18, 03:47 PM

Looking at it a while ago, the air con wasn't responding but the receiver has the appropriate 5v and ground, and I measured signal spikes at 2.5v from the receiver to the main board (at the point where it plugs in), when the remote button is pressed.
From there it goes directly to pin 24 of a micom chip. Have to pull the main board to get a look at it.

Is 2.5v enough to trigger a micom input ?

**Reschs** · 16-02-18, 08:28 PM

Originally Posted by loopyloo

Looking at it a while ago, the air con wasn't responding but the receiver has the appropriate 5v and ground, and I measured signal spikes at 2.5v from the receiver to the main board (at the point where it plugs in), when the remote button is pressed.
From there it goes directly to pin 24 of a micom chip. Have to pull the main board to get a look at it.

Is 2.5v enough to trigger a micom input ?

If you are using a multimeter, that reading is fine. If using a CRO not good.

When faulty check you get the same reading at the input pin as the output of the IR Receiver. Corroded tracks etc.

**loopyloo** · 16-02-18, 10:12 PM

Originally Posted by Reschs

If you are using a multimeter, that reading is fine. If using a CRO not good.

When faulty check you get the same reading at the input pin as the output of the IR Receiver. Corroded tracks etc.

Yes with a hand held cro.
I took the main board out. Everything in really nice condition and I must say really well made and the tracks, pads and joints are perfect, BUT I found badly corroded pins on a 4 pin plug for 2 thermocouples. One pin broke when disconnected. I replaced the plug and socket but as I suspected, it had no effect on the remote problem. At least I found something that needed repair.
I forgot to look at the chip number while I had it out. I can tell you it's 64 pin.

I've got this receiver from a TV that's the same in appearance to the air con's . Once I check the pin configuration is ok I'll swap them over and see how it goes.

**Reschs** · 16-02-18, 10:27 PM

Originally Posted by loopyloo

Yes with a hand held cro.
I took the main board out. Everything in really nice condition and I must say really well made and the tracks, pads and joints are perfect, BUT I found badly corroded pins on a 4 pin plug for 2 thermocouples. One pin broke when disconnected. I replaced the plug and socket but as I suspected, it had no effect on the remote problem. At least I found something that needed repair.
I forgot to look at the chip number while I had it out. I can tell you it's 64 pin.

I've got this receiver from a TV that's the same in appearance to the air con's . Once I check the pin configuration is ok I'll swap them over and see how it goes.

Stupid question. Have you checked the 5 Volt supply with multimeter and the CRO, for correct voltage and no ripple. The output pulses should be 5V peak to peak.

**loopyloo** · 16-02-18, 11:52 PM

Originally Posted by Reschs

Stupid question. Have you checked the 5 Volt supply with multimeter and the CRO, for correct voltage and no ripple. The output pulses should be 5V peak to peak.

Yep 5v and reasonably clean. And you're right about the pulses being 5v P-P. I was pretty sure 2.5 wasn't right.
I just finished and came back to report. I believe it's now fixed.
I swapped the receiver and achieved hard switching, full 5v.
Air con turned instantaneously. I say that because now I've seen it, I think there used to be an 'ever so slight' delay. Just a smidgen, but I sensed something there.
I would say the receiver was prob faulty when fitted at the factory.
Anyway thanks for the help.

**loopyloo** · 17-02-18, 11:07 PM

For anyone interested.
This is the internal board showing the receiver pins and the corroded plug for the thermocouples.
Also the receiver pinouts.

**fandtm666** · 18-02-18, 12:19 AM

thanks for that i have the same ac unit so now i know what to look
for if mine plays up.

**TVguy** · 18-02-18, 08:12 AM

FYI: Almost, but not exactly, the IDU circuit diagram
BTW, the pipe and room sensors are NTC thermistors, not thermocouples. IR receiver datasheet .

**loopyloo** · 18-02-18, 12:09 PM

Originally Posted by TVguy

FYI: Almost, but not exactly, the IDU circuit diagram
BTW, the pipe and room sensors are NTC thermistors, not thermocouples. IR receiver datasheet .

Thanks for that.
Interesting....I thought at least the room sensor must have been a thermocouple because one of the leads has stiff wire like nicrome or something and was impossible to solder. I had to crimp it.

Air con still working well too.

**gordon_s1942** · 18-02-18, 03:09 PM

Most likely that wire can only be joined by induction or RF type of process. Certainly something not generally found in a DIY tool bag.

**TVguy** · 18-02-18, 03:20 PM

Originally Posted by loopyloo

....I thought at least the room sensor must have been a thermocouple because one of the leads has stiff wire like nicrome or something and was impossible to solder. I had to crimp it...

Even the most expensive VRF systems use NTCs as a cost down measure. They interface easily with 5 Volt microcontrollers and are very linear.
An embedded microcontroller eight bit ADC, via an electronic expansion valve, can control superheat to within 2 K, something a TXV or piston orifice could never achieve.
Even older splits like yours looploo are still very efficient; and current products have coefficients of performance approaching 3.5, which is remarkable.