Coolant temp gauge replacement (late petrol)

[AngeloEvs]

Measure the voltage across the centre leg of the regulator to earth or measure the resistance across the centre leg to a known earth/chassis. You should be measuring a very low voltage or a resistance of around 22 ohms. The centre leg must have a good connection to earth (via the 22 ohm). Earlier vans have the centre leg direct to earth and no 22 ohm. If you suspect a problem with the centre leg to earth then make a temporary connection of centre leg to a known chassis earth point in case there is a problem with the membrane pcb and re check the regulator output.

Incorrectly fitting the regulator with input and output reversed will output the full voltage rather than a regulated 10v, but if brickwerks state it should be connected up side down then that is how it should be fitted.

I would test the regulator output, fitted but with the dash on a bench and connecting a 12v d.c. supply as well as carrying out the measurements described above. The multi connector conx are 3 for earth and 8 to 12v. I use a VW connector for tests but in the absence of one you could remove the plastic edge connector surround and carefully connect your psu using crocodile clips directly the copper connections or the positive directly to the regulator input leg and the other to the centre nut of the temp gauge.

[AngeloEvs]

Your reading of 25 ohms is acceptable.

[Cobra88]

That explains why i didnt get the brickwerx voltage regulator to work.

This info would be handy when you buy the chip

[itchyfeet]

Thanks job for tomorrow

what is tye 22 ohm for if early vans didn't have it?

[itchyfeet]

That explains why i didnt get the brickwerx voltage regulator to work.

This info would be handy when you buy the chip

it is you get a detail description and pictures

https://www.brickwerks.co.uk/voltage-re ... panel.html" onclick="window.open(this.href);return false;

[Cobra88]

I just added to basket without entering description

[AngeloEvs]

With fixed output regulators the output voltage is referenced to the voltage on the centre leg, usually 0v as it is usually connected to earth or chassis. Inserting resistance into the ground connection raises the output voltage depending on the value as there will be a voltdrop across the resistor. A 22 ohm isn't going to add much, most often diodes are placed in the centre leg as these raise the output by a fixed amount i.e 0.6v making the output 10.6v since the centre leg is sitting at 0.6v not 0v.

Why? Good question, the DOPs system was introduced and that was the main difference on later vans but does not fully explain the need to raise the regulator output slightly. From my own tests I do know that if the regulator output falls below 9.7v or thereabouts the temp gauge LED will start flashing. May be they added the 22 ohm to give a bit more headroom if the regulator got fully loaded but not exactly sure why.

[itchyfeet]

better than in vehicle I'm leaning towards a bad connection of the 22R , voltage is not stable between 10.15 ish to 10.85 ish

what can I do with the 22R crimped connections?, any body had luck soldering of does the PCB melt?

if the centre leg was open circuit or high resistance would that give higher voltage readings like I saw? maybe I should test it to prove, I first saw high resistance and it got better as I poked the resistor contacts with my meter.

P1100574 by Paul_Barr, on Flickr

much more stable like this centre leg to earth ,

P1100576 by Paul_Barr, on Flickr

[AngeloEvs]

I would not try to solder the crimp connection at this stage.

Yes, bad regulator connection to earth and the regulator output will be high.

There is about 10mm either side of the 22 ohm that you could clean up and gently solder a light gauge wire to or you could solder directly to the regulator centre leg but you have more or less proved the regulator works fine when directly connected to earth. The other end of the wire can be connected to the centre nut earth connetion on the temp gauge.

Using the centre nut as earth, solder a light gauge wire directly to the centre leg of the regulator and test for 10v out. Then solder to the 22 ohm lead that connects to the regulator socket, should be same - 10v or therabouts. If higher then the socket the first crimp conx is suspect.


Lastly, solder to the other side of the 22 ohm that has a crimp to earth and check for 10.2 v or thereabouts. If not then that crimp to earth is suspect.

Once you have confirmed that one or both crimp conx are suspect you have several options but best do the tests.

[itchyfeet]

I was thinking of taking the regulator out and bending up the centre leg then reinsert it so centre leg is connected to nothing.

Then solder a 22 ohm ( actually two 51 ohms in parallel as thats all I have) to the leg and a wire from the other side of the resistor to a ring terminal under the nut and if it works leave it like that. Heatshrinked of course.

this method is completely temovable and does not risk damage to the pcb

I assume there is no other connection from the track on the pcb between centre leg and 22 ohm?

[itchyfeet]

Bit more time on it this evening, I think it may have been a bad connecton on the regulator holder because after a gentle squeese on the crimpss it seems to be holding a steady voltage so I put the dash back for another test.

Much better and the same reading with a 100R both with engine running ( higher batt voltage from alternator) and not running

As you can see my fuel tank gauge was also affected, I have been running 3/4 tank for 18 months thinking it was full

P1100580 by Paul_Barr, on Flickr

I then connected the temp sensor and started it
as the needle was climbing I pulled the level sensor plug off and the light started flashing, good I thought until....

P1100582 by Paul_Barr, on Flickr

Aggghhhhhh

So the level sensor relay was pulled and the same thing full gauge and flashing light.

I measured the sensor resistance ( it was quite warm by now it's a hot day)

P1100583 by Paul_Barr, on Flickr

then I put a 50R fixed resistor in ( less than the 80R of the sensor) 3/4 and no flashing

P1100584 by Paul_Barr, on Flickr

next I cut the wire in the black box and connected the sensor direct to he dash and same thing full gauge and flashing.

So it looks like the resistance of the sensor is being corrupted somehow perhaps with electrical noise from the engine becuse a fixed resistor earthed at the dash end does not do the same I then thought maybe my RMS meter was not showing the true regulated 10V picture so I grounded the centre leg and got the same full gauge and red flashing light

and so I turned my attention to the capacitor mentioned by AngelEvs

Also, there is a an electrolytic capacitor across the output of the gauge on later vans, never come across one that is faulty but worth bearing in mind that these do fail sometimes.

my meter alternates between the two values does this mean this cap is shot?
I have a better capacitance meter at work but I'd expect this one to give a constant reading.

temp sensor cap by Paul_Barr, on Flickr

[AngeloEvs]

If the gauge is fine on a bench test then it is not the capacitor, if it was the cap then the fault would show during the bench test. Where did you take the sender measurement of 79 ohms?

[itchyfeet]

If the gauge is fine on a bench test then it is not the capacitor, if it was the cap then the fault would show during the bench test. Where did you take the sender measurement of 79 ohms?

the red yellow sensor wire is cut at dash because a 52mm gauge was fitted by the previous owner so i'm measuring there to earth using ignition barrel as earth

Surely a cap is there to filter noise? what other use does it have on a DC circuit?


I'm very grateful ror your help, thanks.

[itchyfeet]

this is the sensor, looks standard.

P1100537 by Paul_Barr, on Flickr

[AngeloEvs]

The fact that the gauge goes full scale when the the red/yellow wire is re-connected means that the gauge sees a low resistance. The level sender part of the circuit can be eliminated as you have removed the module 43, this leaves the temp sender, asssociated wiring or connecions and coolant temperature as the only possible causes given everything is fine on the bench. Supply noise can trigger the led but would not cause the gauge to go full scale, only a low resistance can do that.

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

A better place to accurately measure the sender resistance when engine hot is in the engine compartment black box. You will see the red/yellow wire in the barrel connector so easy to separate and measure the resistance. With it separated you can measure the wiring to the sender and the wiring to the dash. You need an accurate reading as that 79 ohm measurement is at odds with your 50 ohm test.