Gallons Per Hour Error?

[Coyote X]

I got an ALDL connector for my Astro and now I can monitor the computer with a laptop. I noticed that the MPGuino tracks along perfectly with the mph and injectors at low pulse width, but when the injectors go over around 3-4millisecond pulse width the MPGuino does not increase the gph number any further.

I have 0055314688 usec/gal in it but changing the number to a smaller or larger number doesn't seem to get it to read the larger injector opening times, it just changes the displayed number but still does not read larger injector opening times. So if there is an error I figure it has to be either somehow the injectors are going above 5 volts after 4ms or a timer just cant get above ~4ms.

I have been trying to figure out the code to see if there is something overflowing and keeping it from reading a 8ms pulse width or larger but I don't see anything. Is there any quick way to get a pulse width number displayed on the custom screen? Or does anyone have any idea of why it won't read anything larger?

[dcb]

Hmm... you can get duty cycle with:

change:
void doDisplayCustom(){displayTripCombo('M','G',instantmpg(),'S',instantmph(),'G','H',instantgph(),'C',current.mpg());}      

to:
void doDisplayCustom(){displayTripCombo('M','G',instantmpg(),'S',instantmph(),'%','D',instant.injHius*2000,'C',current.mpg());}

It uses a much less complicated arrangement than the instant fuel consumption so it might tell us something.

Might need a scope, or record the injector with the laptop soundcard?

Just curious How does one get to injector pulse width with ALDL ?

I'll try some different ranges with the signal generator in the meantime.

[Coyote X]

I'll take an oscilloscope out and get a video of everything running together so it gives some good diagnostic information. The van has an inverter built into it so plugging in stuff is real easy

I am using a standard usb ALDL cable and Tuner Pro RT. Tuner Pro RT is pretty nice. I will probably pick up the memcal adapter so I can burn my own eeprom chips to adjust some of the stuff in the computer this summer.

[dcb]

Well, it looks ok on the generator. Played with it to an on time of 1 millisecond and off (injector open) time of 200 milliseconds. It will be interesting to see the scope perhaps.

[Coyote X]

The instant.injHius value stays consistent with the gallons per hour so it has to be something in the injector signal. I will get a scope out to it and see what is going on as soon as I get time today hopefully.

[Coyote X]

http://www.youtube.com/watch?v=LvxQlx0w2WE

Quick video uploaded to youtube. I will try and get something better in a little while. I didn't do much with it other than set it up and grab a quick video to make sure youtube had good enough resolution to see the numbers on the display.

[Coyote X]

2 more

http://www.youtube.com/watch?v=HcKuUb-amYw

http://www.youtube.com/watch?v=-BNdUK3JpPY

Still doesn't make a lot of sense to me. There is like a 1V difference between the injectors being on and off on the scope. Look at the second video where I set it to a larger time per division and you can see the on time is higher than the off time but not by much. it is 2V per division on all the videos. 0V is the second to last line on the display.

[dcb]

Ok, I think we are looking at a "peak and hold" setup here.

The lowest line is the peak (0 volts), the middle line is the hold. The upper line is 12 volts, right?

I think you *might* be able to add a 20k or so potentiometer (ahead of the guino injector resistor) and set it up so that middle line is below 2.3 volts on the mpguino and the upper line is above 2.5 (or do some figuring and use resistors).

[Coyote X]

That is about what I was thinking after digging around a bit. I should go out and get a video of the signal after the zener and resistor so we can see exactly what the avr is seeing. Each line is 2V and the spike that goes off the top is 14V. The middle line is 4V and after the second spike it is around 5V. It really doesn't make sense because the 0V pulse is always the same size then it spikes to 14V then goes to 4V and the 4 volt is as long as the pulse width reported on the ALDL plus ~0.5ms. But then to close the injector I would figure it needs more than a tiny spike at 14V before sitting at 5V. I always assumed a closed injector should read 14V since it is switched on the low side.

If it is a peak and hold injector that means probably a good bit of the GM TBI cars use the same type of injectors and it might be something that needs checked on a few other engines. I will check with some neighbors and see if I can find any other TBI engines around here.

[Coyote X]

http://www.youtube.com/watch?v=73YbVHRa0fI

It is dark so the lines aren't visible but you can pretty much tell what is going on without them. I had it on 50ms at the start then changed it so a single pulse could be seen. I switched it to ground and back so you could see where 0V was at on the scale. The zener is pretty much only cutting off the 14V spikes and not doing much else it looks like. If I had a second probe I could have measured before and after the filter at the same time. I might do that next week when I can pick up a second probe from school.

Looks about the same as before the zener/resistor but there is even less of a difference between on and off. I did notice that at idle the 10x probe was loading the circuit enough to make it read 99% duty cycle. It would go back to normal reading when off idle. I am kind of surprised it even can read a signal like that and get any useful data out of it. It might only be reading the curve of the peak and not the sustained injector hold signal. That would possibly explain why it has such a narrow range of values it reads. I am just guessing about that though