Fixing the instrument cluster in the '85 diesel RV conversion has been a bit like playing "Whack a Mole." The issue is with the fuel gauge.

Both the fuel gauge and the temperature gauge supposedly run off of 7 volts. That is the marking on the back of the cluster assembly. The way the gauges work is interesting. The indicator needles in each of them is attached to a bi-metal strip that has a heating wire wrapped around it. The wire is fed by the supposed 7 volts and then goes to the respective sensor which is a variable resistance. The fuel sensor varies from about 3 ohms for a full tank to about 110 ohms on an empty tank. I did not measure the temp sensor but it goes from a high resistance for low temp to a low resistance for a high temp.

The way the unit reduces the voltge to 7 volts is even more interesting. A second bi-metal strip in the fuel gauge is wired around a contact point. When the contact point is closed it connects both gauges to the ignition voltage (13.8 volts approximately). As this bi-metal strip heats up it opens the contact reducing the voltage to the gauges to zero volts. As it opens it also cuts itself off so the bi-metal strip cools and the contact closes again. This process is repreated several times a second thus creating a duty cycle of approximately 50% and thus an average voltage of 7 volts..

Once the fuel gauge gets out of adjestment the duty cycle can be anything and it causes the gauges to read erroneously. If the contact stays open the gauges don't work. If it stays closed both gauges peg high.

So I decided to replace the voltage reducing part of the fuel gauge with a zener dioode. Because of the electro/thermal dynamics of the system the diode could not be 7 volts. That would be too easy. By experimentation the correct voltage turned out to be 9 volts. It takes a 10 watt 9 volt zener diode and a 6 ohm ten watt dropping resistor to make the circuit and the fuel gauge has to be modified to cut out the contact part.

The result is a gas gauge that reads pretty close. At a full tank the gauge reads just above the full mark. At totally empty it reads well below the empty mark. Takes a bit of getting used to but it works.

The temp gauge reads a bit high but is also very easy to get used to.

One change I would make if I were to do it over. I would make the 6 ohm resistor 20 watts. It heats up a bit at 10 watts even though the setup calcs suggest that the resistor should draw less than 10 watts.

 

nice job. An '80s "switching regulator" replaced with 80s electronic (zener) technology.

not that you're going to want to redo any of this, but you could consider using a 21st century switching regulator, which would dissipate hardly any power (certainly less than 1/2 watt) compared to 10 watts. Switching regulators are completely integrated, but do require a half-dozen external components to set the voltage. I did locate this outfit http://www.dimensionengineering.com/products/de-swadj that sticks all the components into their own little package for $15. More costly than your zener setup, but much "cooler."

 

Yeah that is another way to do it. But the zener method only takes 2 components, the zener itself and the dropping reistor. 4 bucks max depending on what you have in your junk box. The delay inherent in the meters themselves makes the set up pretty easy except for figuring out the right voltage. One could do it with math if you knew enough about the heating coils and the bi-metal strips but the "by-guess-and-by-god" method is simpler. Thanks for your thoughtful note.

 

I have a fuel gauge that doesn't give the right readings except for about the first 1/3 of a full tank. I've got a new sending unit, so that's not the problem.
I checked the resistance between the 3 points like the FSM says to do to test the fuel gauge and the readings are way off the spec. The Temp gauge is also out of spec for those tests.
How do I know if this is the same problem you had and If I could fix it by replacing the voltage regulator?