944 Turbo Knock Counter

[whalenlg]

Here's the waveform from LTSpice

[Tom]

Is the blue line the KLR knock pulse, and the purple line the output of this circuit? If so, I'm not sure what's being accomplished by the circuit, since the duration of the output pulse seems to be the same as the input's duration? Or are you solely trying to filter out super short pulses to debounce the hobbs switch? If so, does the simulator show it filtering out very short pulses? Or am I totally missing the point?

Edit: scratch that. I think you are just deboucing the hobbs switch and not trying to filter the KLR output at all? So my concern about the long RC isn't important since a quarter second isn't a big deal when deciding when you are above or below 4psi. But why is the output still generating pulses when the hobbs switch (green) is low?

[whalenlg]

Correct - it's debouncing the Hobbs switch.
The idea is that once the pressure switch output goes high the 1st time, we want to pass the pulses until it settles, so there has to be some hold time longer than multiple pulses until it settles to 12v, but the spice model for the pulses didn't allow settling to a on 0 value.
Attaching a longer simulation that shows the full picture better - inputs on top, output below.

[whalenlg]

Revision 2 - I got a model for the pressure switch working and saw that the cap wasn't draining fast enough. Difference was using a voltage source that returns to ground vs. a switch that just opens are much different.
Added a draining resistor & tweaked other resistors to get the fully off voltage way below the 0.7V threshold of the counter.

See schematic, waveform graphs and pictures of the final build. Still needs a road test.

[Tom]

Should be interesting to see how it does. If I follow that correctly, the KLR signal has no real filter, so will essentially pass through whatever is on the line -- is that right? Will be curious if the ignition gets picked up since that counter picks up such short pulses...?

[whalenlg]

Yes - the KLR Knock detect signal from the diagnostics block gets passed through if the boost pressure exceeds 4psi in my setup.
I don't worry so much about noise coming through since:
(1) It's only about a foot of wiring and there's a metal firewall between engine and the KLR + the counter electronics
(2) The KLR is driving from a low impedance source, so it's harder for noise to affect it
(3) The transistor in my circuit will reject some spurious noise from it's input anyway

[Tom]

Yes - the KLR Knock detect signal from the diagnostics block gets passed through if the boost pressure exceeds 4psi in my setup.
I don't worry so much about noise coming through since:
(1) It's only about a foot of wiring and there's a metal firewall between engine and the KLR + the counter electronics
(2) The KLR is driving from a low impedance source, so it's harder for noise to affect it
(3) The transistor in my circuit will reject some spurious noise from it's input anyway

Will be interested to hear how the road test goes. If the KLR produces pulses less than half a millisecond, you may pick up more genuine KLR pulses than the Curtis counter too. That means your results may not be directly comparable to mine, but if you get a base line for the amount of pulses you see when your car is healthy, that would be your best reference point. For what it's worth, my car will see at least 1 or 2 pulses per gear at just about any boost level at WOT -- though some of that is during transition periods (between gears, when first flooring it or letting up, etc.). I have LED's on the dash that flash when I get a pulse (blue for on-boost pulses, red for off-boost pulses) and I can also report that not 100% of the pulses generated by the KLR are actually knock. I've had the off-boost LED flash going over pot holes.... That said, most of them are true knock I assume. After about 19psi, I get progressively more knock counts while flooring it, so tend to keep it there or below.

[whalenlg]

Road test was great - I drove to work mildly and boost never went about 7 psi - 0 knocks

Drove home and pushed the boost to ~15 psi for a few brief periods and saw 10 knocks total.

Calling this project done.

[Tom]

Shameless plug for a 3D-Print connector that plugs into the diagnostic port behind the glove box. With this connector, you can wire in the knock counter without splicing a single wire.

viewtopic.php?t=3626#p40330

[vacobra]

Here is a downloadable DIY guide showing how to make and install a knock counter on a 944 Turbo. A knock counter (and wideband) are invaluable tools for modified cars!


Porsche 944 Turbo Knock Counter DIY.pdf




knock counter pdf pic.jpg

I’m having trouble opening this pdf is anyone having issues or is it just me?