Gallons Per Hour Error?

[dennyt]

The polarity of the scope is correct. It was grounded to the cigarette lighter, and read +13V when I probed the power lead for the guino.

So, the INJ lead is reading 0 most of the time (there's an underscore next to the A in this image, that marks 0V). Then the pre-fire square wave is +14V for 1000 us (no matter the load, this part stays the same even after I key-off the engine and it decelerates to a stop). Then the first big negative spike, then PWM from 0 to +14V, and a second big negative spike.

[dcb]

ok, so we want to toggle the flip flop on a falling -20v signal (I think it is still doable with just a voltage divider), then adjust the Injector DelayuS to 0 and see what happens.

Just curious, are you sure the thing was still firing with the key off and that the scope wasn't just freezeframing on the last known signal?

[dennyt]

Yeah, it's interesting. When I rev the engine then let off the gas, the right half of the signal goes away but that square pre-fire shape is still there until the engine slows down to idle & needs fuel again. Same thing when I key it off, I think the pre-fire happens even when the injector isn't opening. Weird.

[Coyote X]

The opamp circuit in post 103 works with the gm tbi peak and hold injectors with the addition of a capacitor on the avr input pin and to ground. This keeps it from missing pulses and it seemed more or less right on mine. I can't test it any further though since it is no longer gas.

I can't remember what size cap it was but it was a small electrolytic. It was just enough to smooth out the small gap on the very short pulse width signals that the mpguino would miss the interrupt on.

I still have some of the current sensing chips I ordered and never uesd. I could send them to you if you want them dcb. Just let me know and I will send them to you to play with.

[CrazyLee]

Originally Posted by dcb

not until someone with a tbi tries it at least.

I built this circuit today. It does square up the signal quite well. My design is a 2X2 perf board with the IC in the middle of it. I robbed a power connector from an old 31/2" floppy drive for the signal input and clipped a power supply cable from a dead power supply with the 4 pin connector to plug into the MPGuino.

I used direct connections to the other components.

Now all I have to do is wait for the gas price to go down to fill the beast up. Over 60 gallons. OUCH!

Then I can run some mpg tests.

I'll post my o'scope pictures on my website below.

Lee
crazycomputers dot com/Fleetwood_Flair_26/]1993 Fleetwood Flair

[winker]

@dennyt

I have started a thread about my software soultion for peak and hold in the following thread

http://ecomodder.com/forum/showthrea...tml#post258881

You can see my peak and hold appears to be the inverse of yours, but it may be of use to you, maybe inverse the triggers?

[CrazyLee]

Originally Posted by dcb

not until someone with a tbi tries it at least.

I tried the comparitor circuit. I built it on a perf board and put the IC in the middle.
Here is the picture.

I used one 4 pin plug and one 4 pin jack to allow it to be easily removed or added in.

I found that the circuit squared off the whole wave form from the injector.

Here is a scope trace


Chevy 454, TBI, @ 2000 RPM. Top trace injector pulse, bottom, comparitor signal.

The result was I was showing 50 MPG! That's on a large motor home.
I think it was counting the narrow pulses as the injector signals, not the wide pulse.

Maybe some bias circuit would get the actual pulse width modulation we want.

I tried a variable resistor in the cold side of the reference voltage to adjust the switching point, but that didn't do the job.

I also tried a transistor inverter circuit between the injector pulse and the comparitor input but that didn't work either. I haven't had time to check out why yet. I wanted to see if the flipped signal would be accepted by the MPGuino.

I am going to try the filter circuit next. I'll build it on the perf board.

Lee

[dcb]

the comparitor is not picking up the hold portion of the injector signal.

so you can't fiddle with that potentiometer and get a longer low pulse out of the comparitor? Have you looked at the opto-isolator version?

[CrazyLee]

Originally Posted by dcb

the comparitor is not picking up the hold portion of the injector signal.

so you can't fiddle with that potentiometer and get a longer low pulse out of the comparitor? Have you looked at the opto-isolator version?

I'll check it out. The comparitor picks up all of the positive portion of the injector pulse. I need to bias the signal so it only sees the middle to the top. Then it will see only the pulse width.


I am adding a variable resistor to a B- to Injector in and a blocking capacitor to the ECU input signal to lower the signal B-. I may have more to do.

I hope to lower the signal so the bias is set lower than -5 to -15 volts so all the comparator will see the actual pulse width.

I need a - 15 volt source. I am trying a - volt pump from the actual injector pulse! I'll keep you posted.

[consaka]

this is like a good book that someone ripped the last 3 pages out of. lol did it work?