MoTeC Global Forum

Hi I have a subaru ver8 jdm spec with a m8oo plug in ecu.
The engine has cosworth cams and a efr turbo.
The problem I am having is a ref sync runt error at over 6400 rpm, the engine also skips a beat at this time.
I am really scared of the engine losing sync at this high rpm, and the cam timing getting mixed up and bend valves ruining pistons and all the other bad stuff that will happen.This is in a road race car , and I need 7200 rpm on the front straight .I am of the belief that lots of people are running up to 8500rpm. I have tried to do a recording at high rpm and the crank (ref) signal seems to go funny. I have seen a couple of webnairs that say the traces are not reliable at high rpm?
I am at a real loss .
Any input would be a great help.
I hope I am not the first with this problem.

Ive had lots of these errors on a friends car, M800 pnp.

Not sure exactly when they pop up, but it is very regular. It's never had any impact on how the engine performs though. Always pulls smoothly to 8500rpm

Changing the crank trigger wheel may help.

Chain wrote:...
I have tried to do a recording at high rpm and the crank (ref) signal seems to go funny.
...

I have used a lap top before to collect the crank signals.
9000 RPM is 150 rot/sec, so if it is a 36 tooth you would need a high data rate.
900x36= ~32k, and you to sample at least 2x, and really more is better (4x)

If it was a 12 tooth it would work at 44100, but you probably need a 98000 or 192000 ingest rate.
Also you need a divider (couple of resistors) to not be clipping the signal at WOT.
But if could collect the signal it makes it easier to know what to do based upon it.

The ref/sync mode is sensative to changes between the cam & crank sensor relationship.

Have you modified the block/head relationship or dialed the cams in a different position?
If you have then the cam belt strech at hi RPM is moving the sync pattern enough to create the problem putting the engine out of sync.

People have replaced the sync magnetic sensor with a GT101 hall sensor to get the edges back to where they need to be.

Hi all thanks for the replies, sorry l haven't gotten back sooner but I was away.
Holmz,
Are you of the opinion thatthe ac voltage created by the crank sensor becomes too high at high rpm for the m800 , and the ecu needs to chop the voltage (maybe the ecu can't keep up with the crank sensor) and adding a couple of resistors would drop the voltage and give the ecu an opportunity to handle the voltage chopping?
Stevieturbo,
Have you experianced a differance in timing belts (amount of stretch) between manufactures? The belt on the engine is new oem ( came in a Subaru box from a dealer here in Canada). I don't know if it is the same as the oem in Japan.
The block and heads are factory height.
Thanks again for the replies, looking forward to see everyones input.

Chain wrote:Hi all thanks for the replies, sorry l haven't gotten back sooner but I was away.
Holmz,
Are you of the opinion thatthe ac voltage created by the crank sensor becomes too high at high rpm for the m800 , and the ecu needs to chop the voltage (maybe the ecu can't keep up with the crank sensor) and adding a couple of resistors would drop the voltage and give the ecu an opportunity to handle the voltage chopping?
Stevieturbo,
Have you experianced a differance in timing belts (amount of stretch) between manufactures? The belt on the engine is new oem ( came in a Subaru box from a dealer here in Canada). I don't know if it is the same as the oem in Japan.
The block and heads are factory height.
Thanks again for the replies, looking forward to see everyones input.

Magnetic sensors can be sensitive to loading and distance to wheel. In some cases there is a second overshoot/zero crossing during tooth gap that can be interpreted as another tooth. Adding series resistance reduces the level of this false zero-cross (to hopefully below the threshold to trigger a tooth) AND loads the VR sensor less; both helping the situation.
I may still have some waveform captures on 60-2 that can illustrate.
YMMV...

Chain wrote:Hi all thanks for the replies, sorry l haven't gotten back sooner but I was away.
Holmz,
Are you of the opinion thatthe ac voltage created by the crank sensor becomes too high at high rpm for the m800 , and the ecu needs to chop the voltage (maybe the ecu can't keep up with the crank sensor) and adding a couple of resistors would drop the voltage and give the ecu an opportunity to handle the voltage chopping?
Stevieturbo,
Have you experianced a differance in timing belts (amount of stretch) between manufactures? The belt on the engine is new oem ( came in a Subaru box from a dealer here in Canada). I don't know if it is the same as the oem in Japan.
The block and heads are factory height.
Thanks again for the replies, looking forward to see everyones input.

Ive only done a couple of engines with AVCS, and baselined their cams both with engine warm after the engine was built.
So as to how much any belt would stretch or change after that I dont know. Havent actually monitored raw positions under load either.

If there are any changes, not really a lot you can do about it. Would generally only ever use Gates or the new Kevlar belts on powerful builds anyway.

When I had other issues a few years ago, I did scope traces at circa 6000rpm. Signal was perfectly clean for both crank and cam.
The traces might even be on the forum somewhere when I did the query about my problem.....must be some 5-6 years ago. It turned out to be massive over-fuelling and a crappy wideband
The ecu clips any voltage over 10v anyway, and that's happening from around 1500-2000rpm, so that isnt an issue. That's just standard for a mag sensor.

Totally unrelated to these runt errors though

enclosed, small screenshot of ref/syn on 2001+ crank/cam trigger taken at 5000rpm

Is that a tiny bit of cross talk between cam and crank?