VSS confusion

[toc]

Prior to connecting directly to the injector and VSS input at the ECU, I was using an ALDL setup to try and get the MPGuino working that way (I thought it'd be a cleaner setup).

I started sampling the VSS sensor at the ALDL connector, and after some testing, concluded the value was 4750 pulses per KM.

I pondered more during my trial run but concluded that any request / response setup would be inherently flawed by the timing of the requests, interrupts ftw.

Finding the ECU, one time digging around, I thought we'd tap those connections. So I looked up the 4G63 ECU pin out online, and found the pinout - VSS and Injector one are taped into with jumper wire to my temporary setup.

I tested the value 4760 in the MPGuino, and ran a tank through it, the MPGuino thought we had done 500KM when we did 484.0KM.

From the start of the tank to today, I did random calculations to see how my value was travelling - it seemed wrong as the random tests gave numbers between 4950 and 4969. Most commonly, 4955.

I corrected the value to 4955 today. Drove ~6.58KM according to the cars ODO - the MPGuino logged 8.85KM. This disappointed me, the fuel economy was logging 8.2L/HK. Damned.
The calculation now suggests a value of 6664 - way too high.

If it helps, it's a 5V signal on that, as I tested it with a multimeter.
I did use a 47K resistor, and a Zener diode.

Any ideas on why the VSS value seems more dynamic than expected?
Any testing / troubleshooting suggestions? I had hoped to have an accurate setup so I could start testing the suggestions given earlier...

[dcb]

I assume this is a "homebuilt"?

the 47k resistor makes a just usable pulse well on the zener knee with 12v input.

I would suggest something more like 10k if you are sure you have a 5v signal.

Also of course I am not in a position to rule out software variables if it is customized (code branch, different fuses, extra code injected by arduino??).

[toc]

Yeh Homebuilt (I did the math, after all the screwing around, I should have just bought from you - but I'd have the same issue!).

Back to it, the 47K resistor was chosen cause it was best suited to 50K suggested in the circuit diagram.

I might clamp a multimeter to that jumper and see how it without the resistor (watch for peaks above 5V at all).

Is the zener necessary? I understood it to clamp down the voltage from 12 to 5 - I thought it might not be needed, but wanted it to work so followed the circuit to the letter.

Would the resistor really explain the pulse differences?

Oh - the code is the Metric version with smooth numbers that was published on the forum - I'm sure the code is correct for others use it, and it's not too indifferent from your work.

[dcb]

without a scope, I would leave in the zener. Just lower the resistor so you get more signal to the pin on the cpu.

I don't know if it will fix it, just something that would likely be marginal at 5v and easy to try.

[toc]

My initial testing shows this might have fixed it.
Still early testing, but it's now only out by 0.1KM (and that easily could be the wrong value used for VSS).

If the tyres are inflated higher, or deflated - would that change the VSS value at all?

[dcb]

Originally Posted by toc

...
If the tyres are inflated higher, or deflated - would that change the VSS value at all?

Possibly, dunno.

[Nerys]

change your VSS? NO change your actual distance traveled for the same VSS YES. remember your odo does NOT measure distance. it measures "rotations" and then converts that into distance based on assumptions of gearing and tire diameter.

change diameter and those assumptions are wrong.

[mcmancuso]

A question that pertains to this: DCB I added a cruise control, and now if I have both it and the MPGuino running, the speed reading on the guino gets really flaky... could I reduce the resistor on the board to give it more signal voltage?