Datalogging has me pulling my hair out! MPScan & Logworks Plugin

[wheming]

I posted some datalog anomalies some time ago when first setting up MPScan to try and get logs to be sure I am not seeing spark retard due to knock. The issues I was seeing were the logged sensor values were sometimes failing, showing at the max of the range. Speedo would say 200mph, tach would say 8000 rpms, Ret1-Ret4 would say -31 deg, etc. I tried a config file from Rob, thinking my dashboard gauges weren't set up correctly or something. It did the same thing at times as well.
I went through the setup for Logworks plugin, hoping to use it and check to see if I have a car wiring problem or if it is software for the datalog/cable connection issues. I set it up, but when hitting start on the plugin I get a fail to initialize the com port message. I went and checked the USB advanced settings and set the latency timer to 1 (which it wasn't). Tried it again and still got that fail to initialize message. Restarted computer and reset connection and tried again, and still get same message. So I can not get the logworks plugin to connect to my ecu. I have logworks3, but I'm not even at the point to open it. The plugin needs to be started and connected first.
So back to MPScan, and tried it with the changed latency timer. Still get those failed values at times.

I've included some logs. Log_1150_8-20-13 check around marker 866 that's when the funky stuff starts happening. During this log, I did not see any CEL KnockEnable. But during one of my logs I did see it flash.
The folder Slug_Log_Rob_Config has a log using Rob's dashboard. You can see the same failed readings at around marker 668

I really need to get something workable so i can get some logs. I need to be sure i am ok and running safe, so I can start cranking things up and hopefully race her before the end of the year. I do occasionally see a CEL flash which appears to be the CEL on KnockEnable (I have no codes), which worries me a bit. I am running what I think you all would agree is pretty rich as well, 10.7 AFR @ ~18psi boost. So If I am truly getting some knock retard, that seems kind of weird.

Anyone have thoughts or suggestions? I'm ready to use my lifelines!

[ShelGame]

This looks like a bad connection to me. Like a vibrating wire or something...

[wheming]

Bad connection on the data cable? This is one I purchased, I didn't want to bother making one myself.
Or something of the car wiring?

If I didn't see the log, I would say she was running pretty good. My Dawes gauge was showing green (that should mean around .92-.95v) and the wideband was reading like 10.7 afr. She was pulling good. But under similiar circumstances I did see the CEL.
It also bears pointing out since sdac I've been running with 10 deg base timing since I was seeing timing retard while testing at sdac. So I just have a hard time swallowing seeing knock at these lower boost levels and retarded timing. I can't hear any detonation, but does not necessarily mean there isn't.

I can try disconnecting and recoonecting the data cable. It doesn't appear that anything is loose on the end. The little board is gooped up with that sealing stuff.

[ShelGame]

Bad connection on the data cable? This is one I purchased, I didn't want to bother making one myself.
Or something of the car wiring?

If I didn't see the log, I would say she was running pretty good. My Dawes gauge was showing green (that should mean around .92-.95v) and the wideband was reading like 10.7 afr. She was pulling good. But under similiar circumstances I did see the CEL.
It also bears pointing out since sdac I've been running with 10 deg base timing since I was seeing timing retard while testing at sdac. So I just have a hard time swallowing seeing knock at these lower boost levels and retarded timing. I can't hear any detonation, but does not necessarily mean there isn't.

I can try disconnecting and recoonecting the data cable. It doesn't appear that anything is loose on the end. The little board is gooped up with that sealing stuff.

Could also be electrical noise. Where does the USB cable run? Anywhere near the high tension cables (plug wires)?

Another possibility might be a loose distributor or the shutter hitting the HEP. Your RPM goes to 7k, but it drops to more than 0. So, it's not really like an electrical problem. The Advance follows the RPM trace, which is probably what's setting off knock. The RPM reads 7k, advance goes up, but since the RPM isn't really 7k, you get knock.

EDIT: Nope, probably not a loose shutter. All channels are 'noisy'. Even coolant temp. This is a logging error of some kind. But, I have no idea why it happens only at high RPM. I've had the occasional bad data point, but nothing like this...

[wheming]

Nope. It goes through the firewall near the brake booster AA body. And the diagnostic connector is on the drivers side near the strut tower.

I just had a log running while I wiggled and jiggled the connection and didn't see anything funky. The connection seems solid.


It appears though this wackiness only happens when the log is important. Meaning I'm wanting to log my higher boost. So stuff like tps and map are supposed to be high up on the scale. Somehow that is related.

[ShelGame]

Ooops, see my edit above...

Does it always happen at the same time into the log? IE, ~40+ seconds? Maybe it's some kind of USB buffer overrun?

[wheming]

I saw the logged battery volts drop out and peg also. Yes, it pretty much happens to all the logged sensors.
On your log config I saw all those bits you log click on/off also.

So that's why I went to see if I could get logworks to work. Try something else and see if I can get good data. But that isn't connecting.
Btw, thanks for the help. I really wish I wasn't so far behind the curve on this stuff!

[ShelGame]

With LogWorks, I think it won't let you load a plugin unless you actually have an Innovate product hooked up, too. There's a modified DLL you can install that will let you do it; somewhere. That's just by memory, I never really tried to get the Logworks plugin working myself.

[wheming]

I do have an innovate mtx-l wideband.
But in the logworks plugin step by step it has you run the plugin first before starting logworks.
I guess I could try at least having the wideband connected to the comp and see if it them initializes the ecu com port.