Thread: TD5 Airconditioner efficiency.

Originally Posted by shack

If what doesn't spin fast enough?

I think he means the VC fan at idle.

Originally Posted by RANDLOVER

I think he means the VC fan at idle.

Thanks

So this is a pic of the "rarely heard of" and "never seen" Defender AC unit. It's called "Unit-Air conditioning".

Mine is faulty, I don't know why yet, I'm going to attempt to diagnose and repair it.

My first thought is it's one of the diodes, as the compressor will actually run even without the AC knob turned on.

If anyone with electronic experience has any suggestions, feel free to help out.

Cheers
James

Without the benefit of a wiring diagram or pretty much any context, I'd suggest if it drives the compressor then either the quad op-amp is faulty or one of the switching transistors (the one that drives the compressor) is shorted.

I will say being pretty basic it shouldn't be hard to track down the cause.

Originally Posted by BradC

Without the benefit of a wiring diagram or pretty much any context, I'd suggest if it drives the compressor then either the quad op-amp is faulty or one of the switching transistors (the one that drives the compressor) is shorted.

I will say being pretty basic it shouldn't be hard to track down the cause.

Sometimes it works ok, sometimes it turns the compressor on and off with only half a second between cycles, sometimes it won't engage the compressor at all.

I can see the commands diagnostically and the engine ECU is doing what it's asked to.

I guess a faulty trinary switch could cause the no run situation.

I have all the wiring diagrams and will post when I have time.

The fact that it will SOMETIMES fire the compressor (via the engine ECU) even when the switch is off is the one that gets me... So a signal is getting through when it shouldn't.

But not all the time, it did it today and it did it once last month as well.

The switching transistors can be tested?

Yes it is pretty basic, so I should be able to sort it, plus a new genuine one is about $500.. so that's a bit more incentive!


The condenser fan always works as expected, so maybe it is just the other circuit.

I've tested the switching transistors.

Both are "BD139" but different brands.

NPN

E C B pinout

Red lead on base.

First transistor

B and C .705 volts

B and E .706 volts



Second transistor

B and C .666 volts

B and E .668 volts

Did you measure them removed from the pcb?

It does look extremely simple. I would not even call it a ECU since there is no microcontroller on it It looks more like a logic level converter into anything 12v that the higher power electrics in the car can work with. Should be easy to test signals/voltages over those chunky in/out pins.

But first, I'd look for dry solder joints before you hook it up and measure voltages across the pins (the lm324n datasheets are all over). The BD139 can not power the AC clutch on it's own, it is rated at 12watts max and 1 amp current so those transistors are only used to power relays. (these use typically around 200mA max) so that's why they run without heat sinks.

I reckon two of the more beefy diodes are there to protect the transistors from flyback of the relais coils. Easy to test those too.

Back to captain rightfoot The VC fan would likely not run fast enough at idle to cool the condenser when it's really hot. That's why the thermo should kick in eventually. Well, if you've done some testing we'll debug it further.

Cheers,
-P