AFM reading widely fluctuating in Rovergauge

[Steve_D]

blitzracing said:

Nothing that advanced in the 5 AM im sure- especially as the output is non linear. Id think it may be worth my while reverse engineering one if its £300 to get a referb'.

Do you think the sensors themselves would be available?
There does not look to be too much in the way of 'circuit' and should be easy enough to trace if it is only a double sided board.
Have bad memories of laying out 14 layer boards in my youth.

Steve

[taylormj4]

taylormj4 said:

Measuring the dc voltage across these two with engine running gave:

0.9-1.1V at around 900rpm
up to
1.3-1.5V at around 3000rpm

The reading seems relatively stable but the meter show regular fluctuations of around 0.1V, like it is pulsing up and down around 0.1V at a frequency of around 1-2Hz.
Is that correct ?

Mark,
Would you call 0.1V fluctuations a "wild swing" as you said.
Not sure at the moment if I need to replace the AFM or start looking at other things like the ECU earth.

Anyone know where teh ECU earth is - in the footwell or under the car somewhere ?

[Steve_D]

taylormj4 said:

.....Anyone know where the ECU earth is - in the footwell or under the car somewhere ?

At the front of the timing cover.

Steve

[blitzracing]

taylormj4 said:

Measuring the dc voltage across these two with engine running gave:

0.9-1.1V at around 900rpm
up to
1.3-1.5V at around 3000rpm

The reading seems relatively stable but the meter show regular fluctuations of around 0.1V, like it is pulsing up and down around 0.1V at a frequency of around 1-2Hz.
Is that correct ?

That makes no sense at all- The idle voltage is around 1.6 volts at 800 rpm on a 3.9. Also to get the reading to jump many rows down the map it would have to jump a significant voltage- it would be a lot more than .1 of a volt. I dont know the 3000 rpm voltage on no load, but its should be way higher than 1.5 volts

This plot shows the start and idle voltage:

[taylormj4]

Thanks Mark,
Wish I had an oscilloscope !

Sounds like the AFM is bust then ?

[taylormj4]

Unless my earth is degraded and I'm dropping some of the volts across that.

A temporary connection between the AFM black/red and the battery -ve would test that wouldn't it ?

Edited by taylormj4 on Sunday 2nd April 23:40

[blitzracing]

The earthing is very specific- each sensor has its own earth path back to the ECU, and then from there there is a big single earth point for the ECU to chassis but certainly if you go from chassis ground to AFM ground, there should be no significant voltage on it. TBH, before you get too in depth id borrow an AFM from someone local if you can just to do a swap test. It is possible something the the loom, or ECU is pulling the AFM output down, so you need to eliminate the unknowns.

[blitzracing]

In terms of 'scopes this data logger well for the ECU sensor outputs. I went to the effort of building onto into an ECU case that intercepts all the sensor signals, but it was a big job. You can just plug it into a PC on USB, and you would need to pop some probes onto the inputs. Its wont measure much below 100 milliseconds, but is fine for a lazy old ECU.

https://www.rapidonline.com/velleman-pcs10-4-chann...

[taylormj4]

Borrowed an old 5AM AFM off a mate with a Discovery and thought, before I swap it over, I'll just check that Rovergauge still shows the AFM reading jumping around, and of course, the car runs fine and the AFM reading is normal !

The car has been standing for around 1 week and the weather has warmed up but other than that, nothing has changed. WTF ?

On starting, I noticed that the engine runs more smoothly until the lambdas cut in. It still runs OK after that but more lumpy.
The lambdas are showing fluctating -ve readings and they rarely agree. Is that normal ?

So took car for a run tried high revs, high loads, full throttle from low revs in a high gear etc and all OK.
Running a little bit rough but AFM reading normal.

I made a log of Rovergauge. It's just a massive text file. What's the best way to display it on PH for your comments?

Cheers.

[TV8]

The log files have to be saved as csv format and then you can look at them under the column headers

[blitzracing]

taylormj4 said:

Borrowed an old 5AM AFM off a mate with a Discovery and thought, before I swap it over, I'll just check that Rovergauge still shows the AFM reading jumping around, and of course, the car runs fine and the AFM reading is normal !

The car has been standing for around 1 week and the weather has warmed up but other than that, nothing has changed. WTF ?

On starting, I noticed that the engine runs more smoothly until the lambdas cut in. It still runs OK after that but more lumpy.
The lambdas are showing fluctating -ve readings and they rarely agree. Is that normal ?

So took car for a run tried high revs, high loads, full throttle from low revs in a high gear etc and all OK.
Running a little bit rough but AFM reading normal.

I made a log of Rovergauge. It's just a massive text file. What's the best way to display it on PH for your comments?

Cheers.

The reason for the added roughness is the cars idle better with a richer mixture than the 14.7:1 AFR the lambdas target, so you may be running nice and smooth at say 13.5:1 AFR until the lambdas kick in and lean the mixture off. The RV8 was never designed to be emission compliant in the first place, and all the systems put in place with mixture and timing lead to lower poor speed running.