Thread: Crank signal valid limits

Hi,
I'm still circling........ you have obvioulsy landed........... what's it like?

Originally Posted by sierrafery

Hi, anyone knows(OffTrack maybe?) what's the lowest crank signal limit for the engine to start? i'm interested in the already converted value not the voltage cos in my experience it's somewhere around 280rpm but i dont have the skills nor the tools to read within the ECU's brain... eventually how this whole crank signal monitoring is achieved? e.g. what would trigger a "high speed crank invalid" fault code? Happened to me twice to get below 280rpm(260 -265) while cranking on friend's vehicles and the engine didnt start then compared with mine which was around 300(with pump relay removed to be able to read it well otherwise it would have started too fast) ... i can read the voltage with oscilloscope but i dont see it's relevance as RAVE is very vague in this area, they say 2-3V while cranking

I have the crazy ideea to build some kind of crank signal emulator only for diagnostic puspose to be able to rule out the crank signal in a no start issue on autos when that strange starter interference thing happens...for manuals is easy as it can be bump started but it's very hard to diagnose that on an auto without actually replacing the starter for test

I look occasionally and put the code back into the "too hard" file. I'm happy just knowing the RPMs.

On the PCB the +ve and -ve signal go into a low noise dual op-amp. I've traced some of the circuit but it's a bit tricky on a 4 layer board. I have a vague recollection that the other half of the op-amp is configured as a schmitt trigger which converts the signal to 0-5V logic levels. The section of the MC68336 the crank signal connects to is dedicated to time based functions.

My understanding is the crank sensor is generating a timing signal based on the missing holes in the flywheel. I don't believe the voltage is especially important, providing the levels are above/below the thresholds of the schmitt trigger.

As previously mentioned, the crankshaft is drilled at 10? intervals and features 5 missing holes. Because the missing holes are positioned at uneven intervals around the circumference of the flywheel, the ECM is able to calculate the exact position of the crankshaft within each 360? of rotation.

In addition to detecting the crankshaft position, the ECM is able to assess relative acceleration and deceleration of the crankshaft, caused by a pistons compression cycle and firing strokes.

So the important thing will be the intervals between the missing holes which occur at 0/360, 80, 130, 230 and 290?. The locking hole is at 150?.

My guess is the voltage noted in RAVE is a diagnostic for mechanics to check proper functioning more than anything else.

Originally Posted by OffTrack

.....
My guess is the voltage noted in RAVE is a diagnostic for mechanics to check proper functioning more than anything else.

Thanks for your input, that's what's buggering me cos the tester shows the rpm not the voltage and a multimeter is useless in this case as the frequency of the signal is not 50Hz

Originally Posted by sierrafery

Thanks for your input, that's what's buggering me cos the tester shows the rpm not the voltage and a multimeter is useless in this case as the frequency of the signal is not 50Hz

You'll need to use an oscilloscope.

This is was a image someone posted to a forum I used to run showing the crank signal.
The peaks occur at 10 degree intervals, the "missing holes" appear as the gaps in the signal.

I've seen the waveform in various situations with my oscilloscope, actually the one above is a very "clean" one... what i'm interested is where's the limit or what must be wrong in that signal for the ECU as to not let the engine start or to log a fault code.

SF,

Lets work through what we already have in the thread...

The TPU module that processes the CKP sensor is not concerned with voltage, it works by looking for transitions on logic signals and comparing these with timers.

My assumption would be that logic signal must contain the correct number of transitions for the 10 deg spacing and missing holes for it to be valid. That would mean an invalid signal is one that has any corruption of the pulse sequence.

There is likely to be some tolerance for occasional missed transition. The various fault codes will relate to different levels of error.

The +/- CKP signal goes into a differential amp, which will remove common-mode noise, and output a singled ended voltage that is the difference between CKP+ and CKP-. The signal is then converted to logic levels using a schmittt trigger arrangement.

You can verify the circuit arrangement by pulling the lid and spending some time tracing from the CKP pins. The differential op-amp is within 2 cm of the plug so it's not too difficult.

The op-amp is a TLC2272A. When I originally traced this circuit I found that the CKP signals went to 1+ve and 1-ve, and the output from the Schmitt trigger was 2out.
https://en.wikipedia.org/wiki/Schmit...mplementations

If you really want to know the minimum voltage feed a test signal to the CKP pins on the ECU connector and monitor the schmitt trigger output on the 2272A with a scope. You'll see soon enough at what voltage the trigger stops latching.

Thanks, that makes sense, though i dont like to work on any circuit without having it's clear diagram. When i made measurements i inserted needles in the ECU connector from behind and hooked the oscilloscope to them... next time when i'll have the chance to get near one with noisy crank signal current code or non starter i'll record what i get and maybe then we can figure things out

Originally Posted by sierrafery

Thanks, that makes sense, though i dont like to work on any circuit without having it's clear diagram.

Let me know when you've got that clear circuit diagram.
Not much point discussing further if that is your biggest obstacle.