Mike Simard

I'm stumped badly and need help.
My car will suddenly cut off during random times but always the worse time possible like on the first lap of a big race.
It just shuts off but will restart after stopping. Logging shows a sync change at the cutoff point so it must involve the crank and/or cam sensors.
This has been so random that I can't pinpoint it to something common. The only original 928 parts here are the crank and cam sensors.

Some thoughts so far:
What are the rising/falling edge setting supposed to be for these? I do know that the cam sensor will work with either.

The cam sensor may need to be moved to the 5-8 bank. The current bank sees belt fluctuation from both banks so maybe being closer the the crank will help?

How should the sensors be shielded? Do the Bosch sensors handle that internally or should I be termination the shield lead somewhere or leaving it open?

Any ideas appreciated.

blown 87

They should be grounded on the shield wires, and the ground points should be the same for the ECM grounds. (pretty sure on that as most cars are set up like that)

Can you data log the wave forms on these to see what is going on when they stop working?

I have a portable, recording, 4 channel, digital scope in my Snap-On scan tool if you need me to bring it out one weekend to help you track it down. (Modis)

I also have a 12 volt powered digital Snap-On scope that does not record that you are welcome to use for a extended period if you want it, it is a little big and bulky, but I have used in in cars on the go before to track down issues like that before I got the newer one. (Counselor 2)

john gill

Hello Mike

Have you checked that the setup for the sensors in the software , that there is not something quirky in the setup ie the crank is a 60 tooth (1 or 2 missing) .
THe stock porsche srtup is not as common as you think .

ANother thing to try is to load an older ecu file that you know is ok , so as to eliminate the issue of corruption , this happens more then you think .

You can also email the file to me so I can have a look at it in my software if you like .

john gill

I re read your post again , most ecu system have the fuel pump activated of the crank angle sensor , ie the ecu has to sense that the engine is running above a certain speed befors it will allow the fuel relay circuit to to be energised , this actually sounds what may ( I say may could a dozen other things) .

The software should clearly allow you log all inputs and outputs , so you should be able to see what is being cut .
WHat is the type of ecu m400 or 800 ?

wildguy

Double check the sensor figuration in your files with old files that worked.
The sensor wires should be shielded and wired to 0V. i always use the 0V.

Lizard928

Add a 1k ohm resistor to the + line off the reference sensor.
it is likely that the sensor is generating too much voltage (common problem with this sensor), and it goes outside the functioning range or the aftermarket brain.

Mike Simard

Greg, you could be a big help, we need to get together somehow.

John, it's an M48 with IEX. The only thing visable in the log is a sudden change in 'relsync' that corresponds to the cutout.
I just looked at the logs and the last cutout timeframe was happening at part throttle through a turn with shakey TPS readings from the bumpy ride. Aside from the sudden sync change there's nothing obvious. It does report a sync error.

Thank you lizard and wildguy.

GregBBRD

Mike:

I'm trying to recall what your flywheel looks like. Did you use a "stock" style "counter" ring, or make a smaller, custom one?

I ask, because Porsche Motorsports had an issue with the Bosch triggers and went to a completely different style crank sensor. Seems like they started to have problems when the "diameter" of the "trigger wheel" (teeth) got smaller....Turned out that the sensors couldn't keep up with the "acceleration" of the counting ring....the teeth essentially turned into a "solid" piece of metal, as far as the trigger was concerned.

Are the problems always related to very fast engine acceleration....like when you "stand on it hard" on the first lap of a race?

John Speake

It does sound like a sudden loss of synch between speed sensor and ECU. If it generally happens when your revving hard then maybe the amplitude of the sensor signal is too great. (output amplitude of the sensor rises as rpm's rise)

Is the sensor new ? When hot they can fail, then work again when they cool, only to fail again when hot.

blown 87

And one thing Mike's engine will do is rev, I have heard very few, if any motors that were on gasoline have the throttle response that Mikes does, it is not fuel quick, but it is damn fast to rev.

Mike Simard

It's a stock size 60-2 wheel in the original location made from mild steel. The tooth shape is identical.

Both recent incidents do have recent rev activity. The first was right after a throttle blip downshift. My engine does rev at a very high rate.
The second case didn't have that but does show rapidly changing TPS inputs from the shakey ride and trying to apply part throttle during a sweeping turn.

Mike Simard

So the quick acceleration will cause a stronger signal with a mag sensor?
Could harmonic vibrations and shuddering also cause an occasional tooth to sail by the sensor at a high enough rate to cause spikes?
Would Lizard's resitor cure this or is it in the nature of the mag sensor and the ability of the ECU to 'see' signals?

Lizard928

Mike,

Adding the resistor will hurt nothing and will not affect the signal to the brain (confirmed with oscilloscope).
I have seen this problem with mega squirt and VEMS. I have also seen reports of this on other systems too.
The stock reference sensor generates a TON more voltage than any other reference sensor used by other manufactures.

I would open the harness right by the brain and add it.
The other question is to confirm that the sensor is secured down by a bolt.
What block did you start with? S4? Or did you have to add a spacer to hold the sensor up? Could this be varying the space on the sensor?

justaguy

Mike
You have guys helping with way more knowlege than me but I have had inexplicable shut downs with a Motec M84. It will also shut down on Low voltage below 10.8v and high temp over 205 Deg F. I will watch this thread with intrest.

John Speake

Rate of acceleration of rpm won't increase the signal any more than it will just due to increase of rpm at a lesser rate.

The variation of the signal amplitude/plus spikes sounds like poor continuity between various ground points. These may be enough to cause a loss of synch at the gap in the flywheel teeth. Once it looses synch, the ECU will shut down, it will only regain synch at cranking speed.

"Loosing" the odd tooth signal won't cause a cutout, it's the loss of the synch at the missing tooth that is doing that. So that's probably why it appears to be random.

I don't have any experience of Colin's mod, but I don't think it will help with the problem you are experiencing.