Trouble shooting 1973 240z Ammeter/Fuel gauge

[Hayden]

Trouble shooting 1973 240z Ammeter/Fuel gauge. Ammeter works fine, it is the Fuel Gauge not working at all. With the gauge out I tested the car wiring harness 4 wire plug with 3 wires with a test light. With ignition off NOTHING (as expected).  Turn on Ignition. NOTHING on black wire (obvious ground). NOTHING on Yellow wire (UNIT connection on gauge) I believe this runs to gas tank. LIGHTS UP on Yellow w/Red Strip wire (IGN connection on gauge) Ignition connection. So this is normal correct? I need to check the gauge?

(Back Ground Info: Fuel gauge never worked. RENU Gas Tank - restored my tank. I had a shop install tank with new float and check the wiring at that time.  Fuel gauge still didn't work).

[cgsheen1]

Easy way to check the gauge itself is to ground that YELLOW wire with the ignition on. It should peg the gauge to F. You don't need to let it go that far, if the needle moves the gauge is good.

The wiring and fuel level sender are easiest to check with a multimeter reading the Ohms value between the two contacts on the sender. You can do that at the sender itself, at the connector under the hatch, and at the yellow wire going to the gauge in the dash. The fuel level sender is a "variable resistor" (and built like a rheostat) - it's resistance value changes with the level of the float. As long as you can read a resistance value (ohms) between the yellow wire and ground (black wire) the fuel level sender is probably okay. (the gauge is expecting a certain range of resistance between empty and full so that could be an issue)

IF the gauge is good then you need to check the sender and the wire path between the sender and the gauge.

[Hayden]

 I grounded the yellow wire with the ignition on as cgsheen1 suggested. Sometimes it would start to move, I would stop it short of full,  Other times nothing. I cleaned all the Fuel gauge wire contacts (all 3) and connected the gauge. Without grounding the gauge it started to work. I am getting 10.72 volts at the yellow/red ignition wire  (I checked at the gauge). It would go to half full then fade then start back to just over half. (the gas tank is near full).  It's an all new sender/float. I am going to check the sender with a multimeter next to check the Ohms value.  I'll clean all the contacts at the sender.  Any ideas why the fuel gauge would go to half full then fade?

[SteveJ]

I'll try to remember to set up a test rig to demonstrate testing this weekend.

[SteveJ]

So I set up the gauge with my power supply and a variable resistor. It's interesting to me to watch the power supply ammeter while I'm doing the demonstration.

 

[Captain Obvious]

There will be some DC current draw variation due to the change in resistance of the sender unit, but the approx. once-per-second fluctuations in the current draw while you aren't changing the (simulated) sender resistance, that's not due to any divider. That's the temp/voltage regulator built into the switch. You're seeing the PWM draw from the regulator contacts opening and closing.

You won't see this when you first hook up the gauge because the regulator is cold. And a cold regulator will run full bore (100% duty cycle) until it heats up.

[SteveJ]

Thank you for the correction. What I wanted to show most in the video was that the voltage regulator should be connected while testing.

I have seen in the archives about people testing the fuel gauge out of the car, and it failed when they only connected the green and yellow wires (no wire from black to ground) for testing. The gauge I used was one I purchased about 10 years ago and never tested. Fortunately it is a good gauge.

Again, when testing, don't forget to connect the black wire to the negative of your power source.

[Captain Obvious]

Glad to help. That temperature compensator/regulator is really an ingenious little detail. I should really put together that "how the gauges really work" thread I suggested some time ago. In fact, I suspect the OP's issue with his gauge is dirty contacts in the regulator portion of his fuel gauge. (But I haven't officially suggested that yet because it's jumping the gun until you work through the other possibilities.)

So, you had previously mentioned (in another thread) the caution about damaging the gauge if you didn't connect the ground. That ground connection is the low side of the regulator heater, so if it's disconnected the PWM will run full bore 100% to the sender unit. But I'm surprised to hear it's enough power to burn out the sender heater. You've seen this happen? It's only about 240mA 130mA.

 

Edited November 1, 2021 by Captain Obvious

[SteveJ]

I actually did that once before I tore apart a fuel gauge to look at it. I only connected the green and yellow wires to a 12 volt source (2 lantern batteries in series) without a load. The gauge went to full, and I can't remember how long I had the circuit completed. However, later on the gauge would not register at all. This was 10 years ago before I started studying the wiring diagrams to death.

It was interesting to see in the video how much the current dropped after the first heating coil came up to temperature and started operating the voltage regulator. Mind you, that was on the second run of the day after I verified I had a working circuit. I just ran a test. It took about 7 seconds for the VR to come up to temperature with an ambient temperature of 64 degrees F. I don't really want to see what the power supply ammeter reads without the VR just in case I'm right that it could burn up the gauge.

[Captain Obvious]

48 minutes ago, SteveJ said:

It was interesting to see in the video how much the current dropped after the first heating coil came up to temperature

The gauge current will actually drop completely to zero when the regulator opens. You're looking at a wildly swinging PWM signal with the digital current meter on your power supply and I didn't see a zero fly by on the gauge, but it does actually cut off completely.

And that comes back to the burn out the gauge question... In normal operation, the sender resistance could be zero. So in normal operation, the gauge heater will pass current proportional to it's heater resistance which is about 50 Ohms (which I believe is about 90-100 Ohms). But it doesn't pass that current 100% of the time. It's pulse width modulated and the average power is cut down because of that*. When you operate it without the regulator ground connected, it's running 240mA 130mA at a 100% duty cycle. And it sounds like given enough time, that's enough power to eventually heat up the nichrome wire to the point that it doesn't like it.

*I didn't write anything down, but memory says the regulator runs at about 50% duty cycle once it's warmed up.

Edited November 1, 2021 by Captain Obvious

[Captain Obvious]

Well that's what I get for relying on memory... I just took a quick look at a fuel gauge, and I got the heater resistances wrong. The heaters (both regulator and needle movement) resistances are about 50 Ohms each, not the 100 Ohms I mentioned above. The SENDER unit is 90-100 when empty, but the heaters in the gauge are about half that.

If the regulator is closed, it will pull about 240mA. And if the fuel sender is at zero (full scale), the needle heater will pull another 240mA.

So if you run the gauge without the regulator heater (no ground connected), the needle heater will see a constant 240mA at 100% duty cycle (not the 130 I mentioned above).

[SteveJ]

3 minutes ago, Captain Obvious said:

Well that's what I get for relying on memory... I just took a quick look at a fuel gauge, and I got the heater resistances wrong. The heaters (both regulator and needle movement) resistances are about 50 Ohms each, not the 100 Ohms I mentioned above. The SENDER unit is 90-100 when empty, but the heaters in the gauge are about half that.

If the regulator is closed, it will pull about 240mA. And if the fuel sender is at zero (full scale), the needle heater will pull another 240mA.

So if you run the gauge without the regulator heater (no ground connected), the needle heater will see a constant 240mA at 100% duty cycle (not the 130 I mentioned above).

That lines up pretty well with what I was reading on the power supply.