Gallons Per Hour Error?

[drbobwoolery]

What's wrong with the PNP level shifter in post 24? I have a 91 Geo Metro, and the GPH went off the wall one day; goes down when I push the accelerator. It had worked perfectly with occasional resets a 9V backup fixed.
I thought the program was munged, but found this thread just before shipping it to dcb with a little cash for a reflash. Seeing this thread, I suspect that hash on the 12 v is being counted.
BTW, could someone please put a pointer to whole default setting info in the Metro wiki?? I must have changed every setting twice. Now it appears there is a false triggering issue.
I'll try the PNP setup in a 4 wire extender I need anyway. If that fixes the injector count, with inversion that IIRC is a menu option in v0.82 that I have, I should be back on the road with MPG info.

[racprops]

I posted this after having a lot of problems with resetting settings:

VERY IMPORTANT:

I still had a small touch of resetting, so with the advice of a electronic expert, I added a 100 MFD Cap across the power leads and I added a diode on the power in line, you cut the line and put the diode in line.

This stops all the EMF and backwards voltage that can happen in a car.

It has now held it tank records for 3/4 a tank, 269 miles...and mant stops and starts.

Rich

[dcb]

Originally Posted by drbobwoolery

What's wrong with the PNP level shifter in post 24?

Worked on the bench, and not in the car.

Originally Posted by dennyt

...Any recent developments, or should I dive in with this op amp circuit?

If it is peak and hold, try the opto isolator circuit.
http://ecomodder.com/forum/showthrea...tml#post106855

how many ohms is the injector when unplugged? If it is low (< 7 ohms?) it is likely peak and hold.

[dcb]

Originally Posted by racprops

Have you seen the Peak and Hold injector adaptor by JayCar?? They sell a kit that works with their duty cycle reader kit.

RIch

it looks like they use a shunt resistor and an op-amp. Not sure how it gets dialed in. The opto works by the voltage drop across the inector (for negative switched systems).

re: resetting, you must have inherited a cold solder joint or loose wire or something, cuz there is already a diode and several caps on the prebuilt power.

[drbobwoolery]

OK, what is the current best fix for use with a Geo Metro?

[dcb]

still trying to figure that out. Can you measure the injector ohmage too?

here are the defaults FYI

unsigned long  parms[]={55ul,8208ul,500000000ul,3ul,420000000ul,10300ul,500ul,2400ul,0ul,2ul,0ul};//default values
char *  parmLabels[]={"Contrast","VSS Pulses/Mile", "MicroSec/Gallon","Pulses/2 revs","Timout(microSec)","Tank Gal * 1000","Injector DelayuS","Weight (lbs)","Scratchpad(odo?)","VSS Delay ms","InjTrg 0-Dn 1-Up"};

[drbobwoolery]

thanks for the defaults. The injector measures 0.8 ohms. This is somewhat uncertain, as the probes alone measure 0.8 ohm. There is a power resistor in the 12 V line, and the signal across that may be what we need. The low side of the injector is that resistor and the injector coil down from 12 volts. Maybe the junction of the resistor and injector (high side of injector) is what we need. There may not be as much in the way of spikes. We'll see.
bob

[dcb]

were you measuring the injector ohmage with the connector removed?

[drbobwoolery]

Yes, I worked the connector off the TB, and measured between the male pins on the TB side. The connector on the resistor was so hard and stuck that I gave up on measuring it for now.

[Superturnier]

I think I have peak and hold injectors also.
After reading this topic I realized that I have similar symptons that someone mentioned earlier. My MPGuino reacts primarly to gear changes. Pressing the gas pedal does not have much affect. Also I have the low ohm injectors.

So I measured the injector signal with a scope and it seems that I have a problem to solve. The hold part in the picture is somekind of pwm signal which should be counted also to the injector open time.

My first thought was to make a high pass filter which averages the short pikes in the signal. I think that the injector pintle cannot follow the short pikes either.