New gauge and sender inaccurate

[steveg]

After showing suspiciously high oil temperatures during normal driving on cool days, I decided to test my new temp gauge.

With the sender immersed in 90 C / 194 F water, here's the result. 40 ohms in series with the sender brings the needle back to 90:

Lots of folks online are saying their Smith's electronic gauges run 15% high and that the problem is in the senders.

According to one, Nisonger claims to sell accurate senders.

 

[dklawson]

I cannot tell from your picture. Do you have a ground wire on the case of the Smiths voltage stabilizer? The stabilizer must have a ground/earth connection for the gauge to work accurately.

The stabilizer is supposed to deliver an average of 10V. If the ground wire is omitted the stabilizer will pass whatever voltage is supplied to it. If you are powering your test with a battery charger, that is probably around 13.5V That means that without the stabilizer ground wire the gauge will read as much as 35% high. (13.5V / 10V = 1.35 = +35%)

If you didn't have a ground wire connected, hook one up and repeat your test. (Power supply ground connected to both the threads of the sending unit and the case of the voltage stabilizer. Power supply +12V (or more) to the "B" terminal on the stabilizer. The "I" terminal on the stabilizer needs a wire connected to the spade lug on the sending unit).

 

[steveg]

Doug -apologies for not being more thorough. The return wire from the sender goes to ground and the stabilizer output is 9.9 volts.
This is my setup including ground wire from body of sender to base of stabilizer. It's properly wired per the Smiths Classic Gauge wiring diagram:

 

[dklawson]

I am not familiar with the Smiths wiring diagram you mentioned but the wiring is common for all applications where the stabilizer is used. I still cannot tell if the small black wire is making electrical contact with the stabilizer case. Even though you measured 9.9 V out of the stabilizer, make sure that the case has a ground connection via that "common" wire. Otherwise, you have what looks to be a complete setup.

That said, if you have purchased or are using a traditional bimetallic stabilizer (not a new solid state product) then you have to mount the stabilizer correctly. You should find the word "TOP" engraved on the stabilizer circuit board. "TOP" really does need to be vertically up, not horizontal as you show. Orientation does not matter on the solid state stabilizers.

If the case is grounded and you are seeing the excessive gauge reading you are sort of left with only one option. You can re-calibrate the gauge. I'm sorry, I don't have a link for you but on the back of the gauge you will find two small holes. There may be tape or small cork disks over those holes. When uncovered you will find what looks like screwdriver slots in a piece of sheet metal. Adjusting one of those slots changes the zero setting, changing the other adjusts the span of the gauge.

Do a little Googling to find which slot does what and tweak the calibration accordingly. If the bottom end of the scale is pretty close to right you should only need to tweak the span.

 

[steveg]

 

[dklawson]

I have seen that drawing before. I don't think it was produced by Smiths or Lucas. I have never seen a factory diagram that refers to the voltage stabilizer as a voltage regulator. (It is... sort of. It's just not referred to that way). The ignition circuit shows a 10A fuse. I have never seen a fused ignition in factory LBC wiring. Likewise, the drawing shows a 3A fuse for the 12V power to the gauges and voltage stabilizer. I have never seen an inline fuse on any gauge circuit. The gauges and stabilizer are typically fed by a green wire which is protected by a fuse in the car's fuse box. Right or wrong, the fuse box fuse is much higher than 3A. There is nothing fundamentally wrong with what is shown but it does not represent the way factory gauges were typically fused.

Regardless, the drawing shows a line descending from the stabilizer (voltage regulator) to the earth ground. That's the connection I was mentioning that goes between the stabilizer case and earth ground.

 

[steveg]

Doug - Stabilizer case is grounded. In the pic above the wire from the sender going behind the stabilizer grounds to the base of the stabilizer. Another wire goes from there to the neg terminal on the power supply.

Testing the gauge with voltages for alignment with the three sets of dots: 2vdc. 4.8vdc and 7.6vdc - the gauge itself runs slightly high. Testing with the sender as above results in the quite high reading shown in the first pic.

Nisonger uses ohms to test: 250, 70 and 20. I've ordered resistors to do that - will report back.

I've ordered a sender from Nisonger who claims theirs are accurate. Will re-test the gauge with that sender to see if any different results.

My point in posting this is as a warning to others not to freak out about indicated high temps from these gauges without confirming the gauge's accuracy through testing.

 

[dklawson]

Thanks for that.

I haven't seen gauges read like yours except when the stabilizer is not working properly. However, you did mention getting a reading of 9.9 V which suggests the stabilizer is working correctly. That said, digital meters typically cannot accurately measure the output of the electromechanical stabilizer. I am not convinced that analog meters will accurately show what's going on either. The traditional electromechanical stabilizer works by switching on and off (12V to zero V) anywhere from several times a second to once every second or two.

The only "accurate" way I have found for confirming stabilizer operation is to operate a gauge with an inline resistor taking the place of the sending unit. Powering the test setup with a solid state regulator delivering 10V, I mark on the gauge face the steady state reading. Then I replace the solid state regulator with the traditional electromechanical stabilizer. If the electromechanical stabilizer is working correctly the gauge will display the same value as it did with the solid state regulator. If the gauge reads differently it is time to tweak the output of the electromechanical stabilizer. That is one of the reasons I like the solid state stabilizers. You can measure their output directly and the output doesn't change with time or mounting position.

You said this is a new gauge and a new sending unit. The stabilizer looks newish also. You potentially have three contributors to the reading error. Hopefully the new sender you ordered will correct the problem.