Coolant temp gauge replacement (late petrol)

[itchyfeet]

Thanks

Will do some more trst tomorrow and report back.

[AngeloEvs]

Inspect the conx at the sender.

I would set up a work bench with dash, 12 v supply close to the engine bay and run a cable with spades direct to the sender. Switch on, run engine and see what happens..........

[itchyfeet]

Inspect the conx at the sender.

I would set up a work bench with dash, 12 v supply close to the engine bay and run a cable with spades direct to the sender. Switch on, run engine and see what happens..........

yeh good call easy to do as its in my back garden right by my workshop (shed)

something very odd going on

it clearly had a duff gauge and a bad regulator conection but it appears there is a third fault

no wonder it never got fixed.

[itchyfeet]

Remind me, does the gauge go full scale with engine cold and ignition on?

Tested that this morning No not without engine running which I think is the key

So I measured cold resistance at dash end as you would expect

P1100592 by Paul_Barr, on Flickr

and measured at black box

P1100593 by Paul_Barr, on Flickr



and then added separate wires direct to sensor.

P1100595 by Paul_Barr, on Flickr

same

P1100594 by Paul_Barr, on Flickr

connected up a battery so now we are totally seperate from the engine with the exception that the sensor wires are running past the loom and coil in the engine bay ( this bit I think may be important)

P1100596 by Paul_Barr, on Flickr

started it up normally by now it would be flashing...

P1100597 by Paul_Barr, on Flickr

takes longer but gets there ( it is flashing but camera missed it)

P1100598 by Paul_Barr, on Flickr

resistance at this point

P1100599 by Paul_Barr, on Flickr

In the event of low coolant or a wiring, sensor fault, etc, module 43 will output negative pulses that trigger the circuit in the temp gauge and then the LED flashes.

.

So I'm thinking noise picked up from the coil and green tacho wire causes pulses and this is affecting the gauge and thats why I suspect the capacitor to ground

Next test is to route the wires away from the coil and see what happens
I don't think I have any caps suitable so I need to order one

I also broke the high beam lights bulb

[AngeloEvs]

Its not the cap.

So the dash is isolated from the van wiring, you have a separate 12v supply and the gauge is connected directly to the sender and to nothing else but, after running the engine, the needle goes full scale and resistance is 100 ohms (which is normal engine temp) - is this correct?

[itchyfeet]

Its not the cap.

So the dash is isolated from the van wiring, you have a separate 12v supply and the gauge is connected directly to the sender and to nothing else but, after running the engine, the needle goes full scale and resistance is 100 ohms (which is normal engine temp) - is this correct?

yes correct

[937carrera]

I've been trying to get my head around this for you, a couple of suggestions

Where have you identified that the coolant temperature sender is actually working correctly. (you tested a spare sender and got 165 ohms @ 90 C I think, not the one in the van)
You are in the fortunate position of having another test vehicle. Can you hook up the gauge to your DJ and see if the gauge is behaving the same way.
I think you have ruled out the level sender, if it were me I would pull the connector off (and bridge the connections if that's the way the switch works) to eliminate that from the equation

I'm thinking along the lines of good gauge / bad sender or bad gauge / good sender. Of course bad gauge / bad sender is also a remote possibility

[AngeloEvs]

Given that the sender resistance is at around 100 ohms (which is pretty well what it should be for a normal engine warmed up) the gauge should be centre or thereabouts. Take the same measurement with engine running just to eliminate any possibility of vibrations causing intermttent problems with the senders resistance.

You tested the gauge out of the pod and everything tested ok using resistors to emulate the sender. When you used a dummy load of 100 ohms did you leave it connected for a few minutes?

I would eliminate the dash pcb if you haven't already done so. Remove the gauge fom the pod, centre nut to negative, 10v supply to gauge input and pair of wires direct to sender, one wire to the gauge output and the other to the centre nut. Use eyelet tags to make the cnx to the gauge.

Monitor both the sender resistance while you run the engine and the behaviour of the gauge.

Something has been missed and sometimes you just have to re do the testing procedure by a process of elimination. Starting with gauge, sender, then gauge and sender, then gauge in dash pod with dummy resistor loads, dashpod fitted, etc, etc.

E D I T, you have another van? If uses the later horizontal connector then do as Carrera suggests, test the dash pods in the other van.

[itchyfeet]

I tested the broken gauge out of the pod as this was broken its not being used for now

This is the replacement I bought which was not tested out of the pod.

I have another almost identical van (.see my signature below) but I'm reluctant to remove the dash pod any more than absolutely necessary as they are delicate so I will but only when I have exhausted all other possabilities.

remove gauge first In this one I think after a few more tests ( like leaving the 100r for longer) and rig it up with just a regulator and sender

Certainly got me confused and I'm an electronics engineer!

[937carrera]

To be clear, I was suggesting hooking the new gauges up in the same temporary way as you have for testing in the tin top - not disturbing the actual dash in the DJ

[itchyfeet]

Its the DJ dash I have a problem with
tin top is fine and I want to.keep it that way.

Wil do some more tests later just eating an ice cream right now

[itchyfeet]

When you used a dummy load of 100 ohms did you leave it connected for a few minutes?

.

Double checked this with 100R same setup as this morning when the sensor was connected ( sensor was not conected to vehicle electrics) and waited 10mins it's fine, so the sensor must either be duff of picking up noise, as AngeloEvs is convinced it's not noise I'm going to swap the sensor, I don't have the right one just the earlier type but it will do for the test.

P1100600 by Paul_Barr, on Flickr

P1100602 by Paul_Barr, on Flickr

[itchyfeet]

Swapped the sensor same thing,
I stopped the engine at 3/4 gauge it before the light flashed and the gauge stayed at 3/4,
without disconnecting 12V DC I swapped the fixed resistor back in and the level dropped to just below LED.

All on a remote battery no connection to vehicle electrics

It makes no sense to me engine was not running and it stayed high, the sensor is just a thermally variable resistor so why does it behave differently to a fixed resistor?

Next to strip the gauge out I think, it maybe try the old 'mended' one.

[937carrera]

What resistance were you seeing from the sender when you stopped the engine ?

[itchyfeet]

This is the old broken now mended gauge, I wired up another regulator with 0v on centre leg

checked I had 10V and checked again with the fixed resistors of 100R ( sits just below LED) and also 25R to check it goes full scale and flashes and it did,

So connected the sensor ( remember nothing connected to vehicle electrics) and started and it goes full scale, then stopped engine and it stays up there.

Disconnect batt and reconnect and it returns to full scale.

So this is behaving the same as the dash with replacement gauge.

It's leading me to think some of the other components like cap or diodes on the dash have a role to play and they are broken hence why this without those components behaves the same, I can't explian why a sensor behaves different to a fixed resistor but maybe it has some capacitance.

I'm temped to throw in the towel but I know that's just not me, I'm doomed to waste more time

Off to bang my head on a wall.

P1100614 by Paul_Barr, on Flickr