Wiring the fuel guage!

The diagram is here http://www.panteraplace.com/El...E%20AND%20SENDER.pdf with color codes.

Mike

Thanks Mike, I knew somebody would be awake.

Looks like the second wire inside the tank has been removed at some time. Just a broken connection inside the sender and an empty screw down the bottom. I thought it may be for a low fuel light but didn't know if the car had one of them. From the diagram I can see it does.

I've lengthened the float and that has altered the guage a bit, but seems I need to lengthen it a lot to get anywhere near accurate.

Not sure about any GT5 differences, but in earlier Panteras the yellow wire would be for the low fuel LIGHT which your gauge may not have, so it's often an unused wire. Obviously your wiring is generally correct. Assuming your gauge is a stock Veglia gauge, you would need to bend the float arm UP to get it to leave "Full" quicker and eventually arrive at "Empty." Be careful if you do so to ensure no binding on the arm throughout the full range of motion, given the shape of the tank. If it's instead an aftermarket gauge, you may need to add some resistors in the sending wire. You can also use a temporary potentiometer in line to help determine the amount of resistors needed regardless, if you wish to avoid the bending. Stock sender should be a range of about 0-200 Ohms. Late-model aftermarket gauges often have a range of more like 240-40 Ohms but would also read backwards, not your particular issue.

And if anything, I believe you need to SHORTEN the arm slightly, as from what you described your range is now more like 24 gallons (from about 29 on the upside including about 8 extra that don't really exist, down to about 5 on the downside). But tank capacity is about 21 gallons.

Lee

Thanks Lee,

There are two small lights on the dash to the left of the steering wheel near to the wiper/wash controls. Is one of these a low fuel level light?

If I put a potentiometer in the white/black line to the guage, how do I measure the output so I can then go and get the right resistor?

BTW: The guages are Autometer from the carbon fibre ultralite range.

http://www.summitracing.com/parts/ATM-4714/

Looks like empty is O and full is 90.

Is this Ohms, and if so how/where do I read this and convert the car to show properly on that guage. Seems a long way removed from the 0-200 of the stock guage.

No, I suspect the fuel level light would always have been integral to the (stock Veglia) fuel GAUGE itself (an orange light on the low end of the range, though neither of mine have such so I could be wrong, but not your main concern here).

The Autometer #4916 would be the closest match (that I'm aware of, in Autometer ultralite series, unless they have a newer product) to the ohm range but would read BACKWARD. You could add a 40 ohm resistor and would then be very close, just reversed. I don't see any good solution with the Autometer gauge you have, as you note it's a very different range. There's presumably a way to internally rewire the gauge or the sender, but that's beyond my ability (and I have yet to come across any instructions on the forum here for doing so, despite the wealth of knowledge here that far exceed mine).

I should note, Hall & Precision ProFormance sell a proprietary VDO gauge that reverses the direction on the 240/33 Ohm range to 33/240 (AND when I installed theirs, also corrected for the 40 Ohm difference on my car), and I believe I have read that another vendor(s) does the same for Autometer, but I can't recall which vendor(s). Perhaps someone will chime in.

Lee

Well that was an interesting and very smelly day. Don't light a match around me tonight!!

First thing I did was drain the tank fully. I was suprised at the amount of crap that came out of it.

From an empty tank I filled 20 litres. In an 80 litre tank this should be 1/4 full. My current guage read just under half, and putting an 82ohm resister in series bought the guage back to just under 1/4.

Then another 20 litres went in and the standard setup read just under 3/4! 82ohm resistor again bought this to a little under 1/2 which was more accurate. All good I'm thinking.

Then.........

The third lot of fuel went in bringing total to 60 litres. Standard configuration went over full. 82ohm resistor did bugger all so i then tried 164 ohms in series and this bought the guage back to a perfect 3/4.

So that's just not working. There is no linear curve there.

So along the way I checked the voltage at the white/black wire that exits the sender.

0 Litres - 2.25v
20 Litres - 2.638v
40 Litres - 4.00v
60 Litres - 4.20v

So that sucks also. Again, no linear curve to base a fix on.

So those that are still keeping up, here was my next great plan.

Using the fuel pump this time, I drained the tank, 20 litres at a time.

I took out 53 litres before the fuel pump ran dry. So there is always 7 litres in the tank that the pump can't scavenge. That's poor design right there. That's like 40 kilometers on the road. Bloody long way to walk!

So tomorrow's job.

Should I base the fuel tank on the usable 73 litres? I need to continue down the resistor road until I find a suitable value to give me the closest accuracy at 1/4 tank and reasonable indication on 1/2 and 3/4.

To do this, do I go back to a 4 x 20 litre solution, or a 4 x 18.25 litre solution, knowing that the pump can't get the last 7 litres. Therefore is 1/4 tank really 20 litres, or 4/78 (18.25) litres.

One thing I forgot to mention. Has anybody ever filled their tank from total drained empty?

The manual says it takes 80 litres but is this accurate?

If I calibrate my guage based on unusable 7 litres, and then 4 increments of 18.25 litres, I'll be buggered if the tank actually holds more than 80 litres and all my calculations will be out.

From what I know, you will never get a linear measurement due to the shape of the tank.

I think Chuck Melton and others have done this, but can't find it on Chuck's site since there's not search function that I can find.

FWIW, Thomas Tornblom in Sweden makes a fuel gauge corrector which maps the ohm readings from the tank at various states of fill to the readings which will cause the gauge to display correctly, despite the irregular shape of the tank. Contact him directly at: Thomas.Tornblom@hax.se