Low impedance injectors, scope trace

[russell mcdonnell]

Hi Folks,

I've got a Freeduino in mostly working order and just need a decent VSS and INJ signal. I have broken out the injector signal from the Motronic DME in my 1989 Porsche 944, and I thought I'd get a few ideas about suitable conditioning. I've attached some scope traces at idle.

So a pulse goes low from about 14V to close to 0V when the injector is fired, and this pulse is around 10us low (0V), 20us high (+14V) repeated for around 1.5mS when the pulsing turns off (second trace). The big spike is back emf as the injector turns off, I'm guessing it takes a further 500uS until the pintle shuts.

I'm assuming I need to turn the low and fast pulsing for 1.5ms into something that ideally is a single pulse high for 1.5ms providing +5V as an input for the INJ signal to the Mpguino, and this signal needs to track any injector pulse width changes.

I have some idea of using a retriggerable 555 timer, to give me a pulse of say 30us, that is retriggered by any further low going pulses. Any other ideas, methods or suggestions, accepted wisdom etc.

Cheers
Russell

[dcb]

do you think this scheme has a chance?

[msc]

Can you get a scope trace for that expands the first part of the injector signal? Often the first pulse is longer then followed by a series of short pulses to hold the injector. Also play with the trigger level and slope to see if there is another part of the pulse you might be missing. It's interesting the way the trace curves up a bit at the far left of the trace. Looks like it might be coming off another part of the total injector signal rather then being steady at rest prior to the part of the trace shown.

In this thread posts 80 and 82 show some other scope readings for comparison.
Gallons Per Hour Error? Later in the thread there are some scope traces showing an inverted but similar signal. In that case either the high side of the injector is being switched by the ECU or the person had the scope signal inverted somehow.

A re-triggering timer pulse is the way to condition this signal so your on the right track. I expect it could be done with a 555 timer but I prefer to work with one-shot timers such as CD14538 or 74HC123 for greater flexibility on selecting the trigger edge or inverting the output. There are many circuit options for the function, select the one you prefer to work with.

With a signal that is normally high and has the initial longer pulse I would set up to trigger on the rising edge and use a pulse width that is about the length of the initial longer pulse. The length of the timers pulse is not critical because you will be fine tuning the pulse length in software anyway. The limits are it must be longer then any gaps between the pulses and shorter then the shortest injector total pulse. When selecting components for the RC part of the timer circuit I would use 1% resistors and NPO or other temperature stable capacitor to have a timing pulse that is stable over a wide temperature range.

Mike