will pay for your help.
I have a Honda S2000 fuel tank sending unit. Values are when Full: 11-13 ohms, half 52 to 58 ohms, low 114 to 120 ohms and Empty 130-132 ohms.
I want the output adjusted for the GM computer . THe GM computer needs to read 40 ohms for Full and 250 ohms when empty.
Hi dbrinus,
So, you're looking for a translator and programmable resistor to put between your S2000 and the GM computer?
[edit] automotive quality parts?[edit]
[edit] what is the GM voltage range? [edit]
Essentially, yes.
The ohms range is above. For voltage it may be 0-5v. Let me double check.
Apparently my earlier post was lost.
Basically you need to halve the current that the GM computer would see from the Honda sensor. There is an offset, but you may be able to get away without it. Otherwise a series resistor may do the trick.
This should be doable with a two-transistor current mirror and a grand total of about 50 cents in parts. I'll take a closer look when i have time. It would be far simpler than an Arduino solution.
Hi CLI.
That's an interesting approach. I was scratching my head trying to come up with a passive resistor network of some sort that would translate between the two impedances and didn't have any luck.
I eventually decided on an ATtiny + voltage divider + programmable potentiometer. Not optimum, especially if we've got to find clean 5V power on a vehicle, but...
I'm curious to see what you come up with!
Pat.
I ran a quick test and the "resistor multiplier" worked as expected. I didn't do anything with the offset. My (minor) concern is that it will need a separate power source. In addtion to the input protection components, the input to the computer is most likely biased with a resistor to form a voltage divider with the sensor. There's no way I can think of to simply insert the current mirror, so the Honda sensor will need a separate source of 5V and the output of the current mirror can be connected to the input to the computer.
When I'm back at my other PC I can draw a diagram. This is interesting: it's been a while since I used a current source for anything other than powering an LED
Something like the enclosed won't be far off...
regards
Allan
fuelsend.pdf (13.2 KB)
Hi Allan,
Could you provide a little more detail on your circuit?
Try modelling it in pspice...
The voltage the circuit sees depends on the source voltage and impedance and the current drawn - I've presumed 5v and 100 ohms for the source.
Then the amount the transistor's turned on depends on the voltage seen on it's base and the emitter - gnd resistance.
Plot emitter resistance against current drawn and it approximates the wanted law....
For different supply volts/ohms, vary R6 for best fit.
D2 guarantees turn on at low input voltage and provides crude temperature compensation.
Ideally it should be thermally coupled to Q1.
regards
Allan
ps you'd be better off with Q1==BC441 or modern small power NPN - perhaps in TO220. - it's got a higher current/dissipation spec.
I don't think a 2N2222 will handle the dissipation - sorry - overlooked that.
Worst case dissipation about 0.6W
EDIT - dbrinas if you could measure the voltage on your GM system at the sensor, both open circuit, and with a 100 ohm load I'll adjust values as appropriate. And for payment you can buy me a beer if you're ever in Cambridge, England.
For interest I append my pspice simulations
Allan
fuelsend.zip (5.49 KB)
Pat:
It's basically a constant current circuit using the fuel sender as the feedback element, so the output current varies with resistance.
Allan: I haven't looked at the simulation, but at a glance it looks like it would work and I'm facepalming for not having thought of it 
Yep - and the ratio R6/R1 provides the current gain - in this case 1/2
edit (- so it effectively doubles the emitter resistor )
D2 removes the transistor's Vbe offset.
regards
Allan
Hi Allan,
measure the voltage on your GM system at the sensor, both open circuit, and with a 100 ohm load
I assume that you are asking the OP to measure the GM system without the sensor attached. (a) open, just measuring the voltage across the disconnected wires from the GM system, and (b) same measurement, but connecting them with a 100 Ohm load.
Pat.