effect of degrading CAS on data logs?

mrfred · Oct 27, 2009, 01:54 AM

I've been doing a bit of tuning for someone, and in his EvoScan WOT logs, I've been seeing instances where the RPM jumps backwards and then forwards. An example is attached below. It becomes more prominent at higher engine loads. I'm kinda thinking its being caused by a degraded CAS sensor. Has anyone else seen this before and figured out what is causing it?

Fast_Freddie · Oct 27, 2009, 01:59 AM

I have seen it on my car also, but attributed it to logging 2 byte RPM... however when the RPM drops load also drops as well, so I am not so sure...

Ceddy · Oct 27, 2009, 03:11 AM

I have heard of AfterMarket Ignition Systems causing interfere with the Hall-Effect CAS. With DSMs we would switch to the older optical CAS when this happens, but you don't have that option. Could there a be missing the ground wire to the block or does he have a upgraded ignition?

linuxman51 · Oct 27, 2009, 06:26 AM

mine does this as well, I asked about it elsewhere and got crickets. I have not replaced either sensor yet, but I noticed it starting to happen after I turned the logging speed up & attributed it to that, the car does not behave like it's losing sync anywhere.

I've got completely stock (& more or less original) ignition on mine.

mrfred · Oct 27, 2009, 08:48 AM

I'm pretty sure he has a stock ignition. I may ask if he's interested in installing a new CAS.

03whitegsr · Oct 27, 2009, 08:51 AM

Is the car missfiring at all?

I've never seen that type of behavior on my EVO, but my cam sensor is only like 8 months old.

Jorge T · Oct 27, 2009, 09:05 AM

I don't know if the sample rate of evoscan is doing this, but I get the same and it is very annoying.

I remember with zietronix data log software the raw sample rate was so high per channel (over 70 smples per channel)that when the data is displayed as a scatter diagram it would get the same reverse and jumping of rpm, boost and AFR, if I displayed as such graph type. The same behaviour was seen on datalogs from the xede.

Logworks uses a moving average at 12hz per channel and this does not happen (just the skipping (out of phase?)) of 2 byte load.

mrfred · Oct 27, 2009, 11:27 AM

Quote:

Originally Posted by 03whitegsr

Is the car missfiring at all?

I've never seen that type of behavior on my EVO, but my cam sensor is only like 8 months old.

No mention of misfire from the owner. You mentioned CAM sensor. Isn't the CAS sensor used for RPM measurement?

Quote:

Originally Posted by Jorge T

I don't know if the sample rate of evoscan is doing this, but I get the same and it is very annoying.

I remember with zietronix data log software the raw sample rate was so high per channel (over 70 smples per channel)that when the data is displayed as a scatter diagram it would get the same reverse and jumping of rpm, boost and AFR, if I displayed as such graph type. The same behaviour was seen on datalogs from the xede.

Logworks uses a moving average at 12hz per channel and this does not happen (just the skipping (out of phase?)) of 2 byte load.

Interestingly, when I was tuning low boost (15-20 psi), it didn't happen it all. It started happening as I approached 30 psi. If it were caused by high data acquisition rate, it should be more prominent at low boost where it takes longer for things to happen.

linuxman51 · Oct 27, 2009, 11:29 AM

Mine primarily does it on spool up long before it hits peak load, and periodically when I use nlts

03whitegsr · Oct 27, 2009, 03:42 PM

Oh, sorry about that. Read it and thought about all the cam sensor issues with the AEM and just thought "cam sensor" instantly.

inco9nito99 · Oct 27, 2009, 05:35 PM

I've seen something like this before as well. RPM jumps around 5k where it advances a few hundred RPMs and then kicks back and forth. This is seen under 100% throttle above 25psi with meth, (not a failsafe issue). IPW is shown to act goofy by getting down to around ~1.xx ms as well. Car almost feels like its fuel cutting.

0xDEAD · Oct 27, 2009, 05:36 PM

I think david buschur went through one or two in his or some customers evos. Maybe get him to chime in here and help out.

**MR Turco** · Oct 27, 2009, 06:07 PM

what sort of mileage on the car? I also believe buschur has seen this on numerous occasions and it can cause catastrophic issues if/when it fails especially at WOT.

MattS00 · Oct 27, 2009, 06:25 PM

Can you log the CAS through evoscan? I'm away from my laptop now.

Ceddy · Oct 27, 2009, 07:07 PM

Quote:

Originally Posted by MattS00

Can you log the CAS through evoscan? I'm away from my laptop now.

You could edit the MUT Table to log TCAS(Time between CAS interrupts)
TCAS goes on to make RPMs, they are the inverse of each other.

I believe the CAS analog signal is completely handled by hardware, you would need a scope to view it.