Deriving ignition pulse from automotive 12v supply?

[indecided]

This is in reference to my earlier LM2907 shift-light project, which utilized a seperate third wire to the negative ignition coil of the car (or the tachometer output) to get the pulses. After looking around, I find that there are commercial shift lights which do not use a 3rd tachometer wire; instead, the ignition pulses can be derived from the power lines, which is all they use. Hence this makes the device much more portable.

However, I have NO inkling how to retrieve and filter out this square wave pulse off the 12v lines.. does anybody have any experience with this? Assuming that the ignition simply carries it, what circutry will be needed to bandpass the pulses alone?

 

[Klaus]

The 12v wires normally carry DC current. If there are any pulses superimposed you should be able to pick these up via transformer action. The secondary would have the pulses only, the DC stays in the primary.
What kind of transformer? You'll have to experiment, do consider the DC component which could saturate the core.
Good luck
Klaus

 

[stevez]

I was trying to solve an "alternator whine" problem and put a scope on the cigar lighter in my 95 Chevy Beretta. It was quite an interesting display that included some very high voltage peaks that were likely ignition coil or fuel injector related. They were clearly way above the other stuff and in time with engine RPM. The highest amplitude pulses were easy to filter - but it would seem you might use them - just the peaks - to use for RPM. I don't think I'd consider this to be as reliable as a more direct connection (or via inductive probe on coil primary or injector). At 5000 RPM you are making 83 revolutions per second and would see 333 pulses per second with an 8 cyl engine. Is that too fast for a simple comparator? Set for amplitude that's well above the DC and other noise but within amplitude of the stable pulses.

 

[indecided]

Klaus : I have taken apart a commerical unit which picks up signals from the power lines, and there is no transformer or anything - it uses a motorola 86 microprocessor, a bunch of resistors, an open air coil, diodes, a couple of pots and really nothing else unordinary. However I had to return the unit shortly and was unable to dissect the PCB and the contents (the PCB construction would have made it impossible anyways.

Stevez : thanks for the info, I suppose a simple circut to filter out the unwanted areas would do some luck. The unit mentioned is a Pivot Shift-Light X which performs extremely reliably. Afterall, this is a "frequency switch" and not a 'frequency display' so it's probably set to trigger at a certain voltage..