Uneven pulse lengths in tach input

[the_bluester]

I am looking at temporarily transferring my C127 (Which has IO enabled) to my brother in laws car to gather some data. To do it I will need to build myself a patch loom to get information in to the dash. The car has an old ECU and no CAN data available.

I am hoping to parallel off the crank position sensor to get pulses to count for a tach input but I can not use the existing OEM dash tach input as it uses coil negative pulses into the dash and the signal may peak over 100V or more, and there are secondary pulses produced from coil ringing which may also end up counted.

I do not expect any problems with the C127 loading the OEM crank position signal down (Nissan optical) but the Nissan optical trigger (RB30) use different length windows on the TDC triggers to identify different cylinders in the firing order. Will the unique/different pulse lengths cause me any issues in producing a tach signal? If so I will have to buy or build an interface circuit to produce a TTL or 12V square wave from the factory tach input.

My data is going to be a little inaccurate as I will most likely have to parallel off existing ECU inputs (Water temperature ETC) but it will be far more accurate than the factory dash.

[adrian]

The Dash Digital/Speed inputs simply measure signal edges to determine pulse width/frequency etc. There is no way to configure uneven pulse spacing so this signal will not be suitable for a tacho signal into the Dash.

You will need something with even teeth or a frequency output directly related to the RPM.

[David Ferguson]

Have you considered using an injector signal as the Engine Speed input. That is nominally a 12v signal and the frequency varies with the engine speed.

[the_bluester]

Unfortunately given Adrian's response, the injector signal would suffer from the same issue.

Adrian, looking at the specs, the speed inputs are falling edge and other digital inputs rising edge. Is there anything to stop me using the appropriate input? One edge of the pulses is in a fixed position WRT TDC on the crankshaft, I would just need top scope it to work out which edge.

David, given Adrians initial reply above, I assume I would run into the same issue with an injector signal.

The other option I was considering was to pick off the signal on the ECU side of the ignition coil driver transistor. That would have the advantage of being a fixed dwell time. I just need to scope it to make sure there are no peculiarities there with the signal.

[adrian]

Yes injection timing would move the signal around and cause the RPM reading to be inaccurate. You essentially need a fixed frequency to RPM ratio. So for example if it was from an ECU or something similar you would expect 1 pulse per RPM or similar. If it is off a physical disk on the crank then it needs to have teeth that are evenly spaced otherwise it will read faster for teeth closer together and slower for teeth further apart.

If one of the edges is evenly spaced around the disk then that should work fine if you pick the correct input (dig/speed) to match the edge.

[the_bluester]

OK, I should be able to work with that. I will scope the signal and work out if the trailing edge (Which is a timing mark in the ECU so consistent) is rising or falling and use the appropriate input on the dash. I could probably also use the injector pulses in the same way but from memory in some circumstances the R31/VL turbo ECU used additional injector events for acceleration pump, that would obviously lead to spurious readings.

I would just pick up the coil driver output of the ECU except for the very small chance of failure in just the right mode dumping back EMF from the coil primary into the dash and maybe frying the input used. Small chance I know as it would have to fail in a very specific way, but knowing my luck!

[the_bluester]

Just in case anyone else wants to do this with a Nissan or similar sensor.

I have wired it in using the Nissan 120 degree signal from the CAS and a digital input as the rising edge is evenly spaced for each cylinder, the trailing edge of this signal varies depending on cylinder number. No apparent impact on the trigger signal to the stock ECU, it revs cleanly to the cutout still.