pcmhacking.net

Hey guys,
I'm playing around with the idea of building a custom dash cluster for my HZ. I plan to build my own pcbs and use an atmega or another arduino compatible microcontroller to control aircore gauges with a MLX10407 air-core driver IC, an also have an LCD odometer and maybe even digital speed readout etc.
It's kind of a long term project so I'm playing around with individual bits at a time before I try the whole project (Ive already raided some clusters for air cores and ive got 10 mlx10407 ic's on the way so I can accidentally destroy a few).
Getting most of the input signals seems relatively straight forward. The vehicle has a VT fuel pump module so I'll be using the sender on it. Oil pressure and coolant temp should be relatively easy even if I end up getting new senders. Vehicle speed I can tap into the 12v square wave vehicle speed sensor.

However I'm not too sure how to go about rpm. I know I can tap into the hall effect distributor or the signal from the ignition module into my '808. I'm probably just making it hard, but that seems too easy and I want to do it from the coil negative. Then I would have the understanding to make it useable on vehicles without electronic ignition, and I could even directly measure dwell from the square wave.
There doesn't seem to be a lot of good information about how to do this from my google searches.
I understand the coil - has a quite high flyback voltage of several hundred volts when the "points open" (module switches off coil). I also know I dont want to pull much current out of that flyback pulse and tie the voltage down or I will induce a missfire.
I'm not looking to pick up on that flyback voltage, but rather filter the very noisy signal and end up with a 0-5v square wave to feed into the microprocessor if thats possible.
I'm imagining I'll need some sort of opto isolation and a low pass filter to prevent noise getting through. But ive got no idea how to set this up.
I'm also worried about coil negative voltage if the ignition module is current limiting, but I'm thinking this isn't a problem on a VN V8 bosch module?

Thanks in advance

You could easily just use a high value resistor and a zener diode for whatever voltage you are wanting to clamp to. also a low value capacitor may be good to filter out noise and the flyback voltage. You don't want too high of a value for the capacitor that it start distorting the signal.

I agree with fl0wfl0w - there is a such a design documented here: http://www.dgp.toronto.edu/~mccrae/proj ... ignal.html
Its probably overkill for you but a clear writeup for the less experienced so ive mirrored the content here in case it disappears.

Thanks for the help guys.
I was thinking something like that would be enough, but pretty well everywhere ive seen online has used an optoisolator. Is that just overkill?

I spoke with my sparky mate and we both come up with;

Coil -ve -------> 100K resistor ------> 5v Zener diode ---------> to PCM

will be testing once i get the materials, using a VS PCM (16176424) with 4L60e standalone.

My understanding is the zener diode needs a minimum amount of current flow to function, so even though its a tiny amount and for a short time, the PCM could see well over 5v until the zener breaks down and starts conducting. I'm no expert so I could be wrong

I'd suggest a TVS diode and 5w resistor, that's worked well for me in the past. Zeners are designed for regulation, TVS (transient voltage suppression) diodes are designed to clamp high energy spikes from things like ignition coils.

Another method that is better in some regards is to wind ~100 turns of thin enamel wire on a small bobbin core, and slip it over the #1 HT lead to make a 100:1 torroidial transformer.
Add a rectifier, and play with the number of windings to get a pulse voltage level that suites your application.

pman92 wrote:My understanding is the zener diode needs a minimum amount of current flow to function, so even though its a tiny amount and for a short time, the PCM could see well over 5v until the zener breaks down and starts conducting. I'm no expert so I could be wrong

Not really, you'll just get an under-voltage, not an over-voltage, if the zener current is too low.

Joe.

Any luck with getting your circuit going? I still waiting on Jaycar to get zeners in stock

it might be quicker to grab your diodes from ebay. shouldnt take too long if you buy from an aussie sellers, and lots of options there.