Sal, I know and that is why I was asking the OP to check if he sees any wire. This mod will screw things up when using the DME with a stock AFM: it saturates the ADC input and that would explain what he sees.

I was wondering if this was an exact reading or just a round off. It looks like they go to a lot of trouble to have a low impedance source for this. Shouldn't it be 5 volts +/- .04 + the offset voltage of the ca139?

A low probability but I have to ask.

Looking at the schematic I agree: 5V with the error being the offset voltage of the OpAmp and the accuracy of the 7805. And maybe some very small losses in the harness.

Not sure but maybe he meant 4.5v was used while he bench tested the AFM? He does say he bench tested it.

By memory I recall that the source/supply voltage to the AFM is 5.0vdc not 4.5

Aprilia,

I have a spare DME and an '86 Carrera. I'd be glad to help you out. PM me.

Let me ask one more dumb question.

The injector voltage waveform should look like a longer pulse to ground for the peak, hold pulses (12 volts/ground) followed by a 47 volt spike.

When you say the pulse width is 8 mSec do you mean the initial injector voltage goes to ground for 8 mSec or is that the whole waveform.

this should help:
http://forums.pelicanparts.com/uploa...1397173776.jpg
ti is what you should be measuring.

"By memory I recall that the source/supply voltage to the AFM is 5.0vdc not 4.5"

That is correct.

"Shouldn't it be 5 volts +/- .04 + the offset voltage of the ca139?"

The offset voltage is typically 3mv (.003 volts). The op amp and an emitter
follower comprise a closed loop low impedance voltage source (5.00 +/- 3mv)
for the AFM.

Thank you all for the excellent feedback and suggestions!

I want to clarify a few points.

- The DME and AFM are stock. My inspection showed that it was the first time that the DME had been opened. There were no obvious modifications.
- The reference voltage to the AFM was measured at the harness connector to the AFM. The measured value was 4.5 V (not 5.0 V +/- 0.04 V). If the DME doesn't compare the ratio of the AFM reference voltage to the potentiometer output voltage, then I would expect the DME to drive the mixture to a lean setting since the reference voltage is 10% less than the ideal reference voltage. The alternative is that there is an impedance loss through the cable harness from the DME to the AFM.
- I will test the AFM reference voltage again at the output of the DME as well as the AFM return voltage at the input to the DME and post the results.

- The injector pulse width duration was measured independently at pins 14 and 15. The 8 ms duration corresponded to the period "ti" in the above figure. The duration included the initial electromagnetic solenoid "shock" period of approximately 200 micro-secs and the subsequent sequence of pulse width modulated signals that made up the remainder of the measured injector pulse (just like in the above figure).

- All the input sensors were measured and determined to be within specifications (as per the Benchley manual).

- The only item determined to be outside of spec was the distributor rotor resistance (measured 5 Ohms instead of 1k Ohms). I am not sure if this is related to the hard to start problem and running rich. Theoretically, the reduced circuit resistance between the coil output and the spark plug would actually increase the duration of the spark, which would aid in the combustion of a rich a/f mixture.

Finally, I received the replacement DME from ECUDoctors last night. I hope to replace the stock DME with the new one this weekend. I will also try to run a comparison (between the outputs of the old and new DMEs) of the injector pulse width as well as the AFM reference voltage and potentiometer output voltage. Hopefully, the DME is the cause of the engine running rich and flooding at startup. If not, then I may have to start considering a complete rewire of the DME harness.

Quick update:

I replaced the DME with the new one from ECUDoctors and guess what? NO CHANGE!
The engine still runs rich.

I measured the reference voltage to the AFM with ignition on but engine off and it was 4.5V.
I measured it again with the engine running at idle and it was 5.1V.

The AFM output voltage measured with ignition on but engine off was 0.5V.
The AFM output voltage measured with the engine at idle (800 rpm) was around 1.5V.

Does anyone have a graph or chart of the expected AFM output voltage vs rpm?

I am beginning to think that the AFM flapper spring is loose (as previously mentioned). I read in another thread that someone else measured their AFM output voltage at idle and the value was 0.7V. I also believe that Sal had a value of 1.1V but that his AFM was not stock.

When I bench tested my AFM, the output voltage was linear across the flapper angle deflection when using a 9V battery source as the reference voltage. There were no drop outs or variations. I did not touch the spring tension clock wheel, which I assumed was at the factory setting. If the 1.5V at idle is indeed to high, then that may be the culprit but I would like confirmation before I pop the lid on the AFM and tighten the spring a couple of notches.

Thanks again to everyone for their assistance to this 911 noob.

at idle it should run around 1.0v. Send me a private message with your email and I'll send you a document I put together on AFM theory. But 1.5v at 800RPM seems a tad high, try unpluging the afm and starting the engine. The injectors will simply fire at the min setting of about 1.8ms it's actually a good next test to unplug the AFM and see what injector pulse width is with it unplugged.

That's correct.

This was posted on 4/5/14:

"An injector pulse of 8ms is a little too long for a cold start. There're a number of possibilities:

1. The CHT is open/intermittent giving a bad value, i.e. use a paper clip to jumper
the CHT connector. Note: The CHT when intermittent will test OK and then fail later,
e.g. at different engine temps.
2. The AFM is providing the incorrect voltage, e.g. wiper arm is stuck
at a large opening or the wiper spring is too loose.
3. The DME injector driver IC has been damaged causing too great of a injector pulse.
4. The AFM input of the DME computer's A/D converter is damaged.
5. The AFM's reference 5.0 voltage is too high.
6. The CHT input of the DME computer's A/D converter is damaged.
7. There's a problem in the injector final stage circuitry."

Number 2 defined your problem.

Hopefully, you can return the ordered DME ECM for credit.

Just thinking here but this makes no sense.
I knew there was a reason you shouldn't use a comparator as an op amp :)

There is 12 volts on pins 18 and 35 of the DME right?

There's no design problem with that in the DME ECM.

The output impedance of the AFM supply voltage (pin 9) is about 50 ohms.
So if the AFM reference voltage drops .50 volts when connected, that implies that the
AFM is an effective load of about 500 ohms. The AFM reference voltage is a low impedance
supply, but not a zero impedance supply. Then under load of the AFM, one would
expect to see about 4.5 volts when the AFM is connected versus the actual open
circuit voltage of 5.0 volts coming from the DME ECM. In any case, a lower AFM
voltage would reduce the injection time and not increase it.

I thought the smiley face was enough but I fixed it

Did you ever measure the resistance of the whole potentiometer in the AFM (pin 6 to pin 9)?

I think they will all be different due to the laser trimming but the circuit in the DME will only porvide 25 mAmps before it will drop out of regulation. See if it is greater than 200 ohms.