Hello all

I'm planning on updating the gauges in my Z to something more modern and a bit more relaible. Does anyone know the resistance range of the fuel sending unit? I know I can replace them with something new but if I could make the stock one work it would mean less work.

Thanks mates

Jan

See page FE-5 in the 1976 FSM.

Thank you my good sir,

You'd need an electrical engineer, or at least a really good tech to make the existing factory gauges more accurate.

The way you'd do it is using 741 "Op Amps" to correct either the signal coming from the sensor, or the gauge display, or both. You'll start with what's ccalled a "unity gain amplifier" circuit and go from there.

Consider a fuel gauge. Most of 'em either come off FULL waay too fast, and then take forever for the last 1/4 tank to disappear, OR the opposite. They sit on FULL forever then ZOOM to empty from 1/4 tank.

So you're going to have to design and build circuits that CORRECT that non-linearity.

That ALSO means that you're going to have to completely empty your tank, pour in 1/4 tankful, 1/2 tankful, 3/4....and measure the signal coming from the sender so you KNOW what the actual curve looks like. In theory it should be a straight line, linear, but it's probably not, which is why your gauge reads too fast on one end of the spectrum andn too slow on the other.

It may be a curve, or have a slight knee in it, until you plot it out you don't know if the inaccuracy is in the sender or the gauge or a combination of the two.

You'll ALSO need a calibrated, mechanical sender to measure oil pressure, temperature, etc. in order to correct THOSE gauges.

So you're going to be drawing six sets of curves --- fuel, oil pressure, temperature --- both signal sent, and display shown

And then you'er going to adjust the resistor values associated with the op amps to "straighten out" the curves into nice straight lines.

Got it? Got an electrical engineer friend who's bored? Specifically, an ANALOG engineer?

Your other choice is to buy some MORE accurate aftermarket gauges and either mount 'em on the dash or replace the existing ones. This would be approximately 100x faster than

designing circuits.

You can, once in a blue moon, get away with "correcting" a temperature sender, for instance, by adding a single resistor, either in-line, or pulled up to B+ or pulled down to ground. If the non-linearity in the sender's output is a simple curve that loses or gains steepness as temp goes up, sometimes this'll work.

I had, for instance a Jeep Cherokee where the owner SWORE it was overheating. It wasn't. It never spat coolant, nothing. But his GAUGE read 20 degrees high. Adding a simple 60 ohm resistor (if my memory is correct) brought the gauge closer to where it should have read, AT LEAST AT ONE SINGLE OPERATING POINT. We tried 4-5 resistor values until the gauge APPROXIMATELY met what our IR thermometer gun told us the temperatuire was at the thermostat housing.

There are also some "gauge correction" boxes out there on the Internet you can purchase, I remember seeing them for odd shaped fuel tanks to "correct" the readings. Probably want to google that.

Wade

That is a lovely explanation. Being an engineer myself - just graduated with a degree in Indsutrial engineering - I understand the need to correct readings based on ACTUAL data and not relative date. I run megasquirt on my car and thus I can read the temperature using my laptop and I've confirmed it multiple times that the temp reading on the computer (which usually sits at about 170) is accurate. (I used a group of thermometers I placed in the radiator to test temps). This past year I worked a lot with microprocessors and other electrical components and understand the need for signal smoothing and signal conditionining. I spent quite a few nights trying to make a signal equal the range I needed

As for fuel gauges and others, my fuel gauge works alright, usually when it sits at E I have about a gallon or so left in the tank so its actually a useful gauge. I'm just planning on updating the daul guages to somethig more modern that tells me a bit more. For example, the temp gauge just has to readings 120 and 250. While this is lovely, knowing where in this range it sits is a bit more useful. (The gauges I wish to buy have marks every 20 degrees roughly temp wise).

When I do get around to purchasing the gauges I will correct them to the range needed. I'm not just going to throw them in there and hope they work. For what Mr. SteveJ says and the book the range of the sending unit is 10-80 ohms, the guage is 0-90 ohms. Clearly there's a difference, small as it is, it would affect the readings the guage will show. I'm sure it would be the same for the temp sensor and oil pressure. Its a matter of getting the right measurements into the gauge to get the right readings on the gauge.

Thank for the help mates

Jan

Holy cow folks, lets not forget that the fuel level is not an accurate indication of volume remaining.

The tank is a not a big rectangle, the relation ship of cross sectional area to volume remaining at all levels is a complex function. Best you put a micro computer based circuit in there so you can properly input the necessary volumetric relationships.

Which reminds me of a story...A farmer was convinced to allow an engineer to attempt to increase his dairy farm efficiency. After 12 months and $500,000 spent the farmer asked the engineer to explain what he had done and how the new system works. The Engineer said, "well, first we assumed a spherical cow..."

I only care to know with any accuracy, when the tank is about 5 litres from empty....

Holy cow folks, lets not forget that the fuel level is not an accurate indication of volume remaining.

The tank is a not a big rectangle, the relation ship of cross sectional area to volume remaining at all levels is a complex function. Best you put a micro computer based circuit in there so you can properly input the necessary volumetric relationships.

Which reminds me of a story...A farmer was convinced to allow an engineer to attempt to increase his dairy farm efficiency. After 12 months and $500,000 spent the farmer asked the engineer to explain what he had done and how the new system works. The Engineer said, "well, first we assumed a spherical cow..."

I only care to know with any accuracy, when the tank is about 5 litres from empty....

:-) Yes engineers are kind of known for over complicating thing, I've been guilty of that. My whole point here, as from the beginning, is will I require replacing the sending unit or would it be possible to use the stock one.

Thanks

Jan

You could try to 'calibrate' your gauge, by putting the vehicle on level ground, and adding gas in measured amounts (such as 1 gallon or 5 liters at a time) to the empty tank, and noting the gauge after it settles.

I've adjusted my mind to compensate for the non-linear curve of my gas gauge.

The range is 8 ohms full and 90 ohms empty, or something very close to that, as suggested look in the FSM for actual numbers. I tried to quickly find the actual numbers, but didn't find them.

I listed the approximate range to show something here. GM fuel level senders are 0 ohms empty and 90 ohms full (after 1965 to 1997). The range is almost exactly backwards. I had an Auto Meter GM fuel level gauge installed in my 240Z for a long while, mostly because I had it from a previous car. I just lived with it being almost exactly backwards. What I had planned to do was to get the fuel low and then flip the sender 180* to show the fuel level more correctly. I just never got around to and and ended up re-installing my original volt (from a 280Z) and fuel level gauge, mostly because the fuel level, narrow band O2 and volt gauges were taking up the radio slot and wanted to re-install a radio in the car.

people have also use Ford ranged fuel level gauges and calibrated them using certain value of resistors. There is information on this at hybridz.org.