datalogs stop logging while running

[akana]

My tunerstudio datalogs suddenly stop logging properly - you can see it where all the resets occur (logs attached). This happened twice in a row over two days. Thought maybe it was a cable connection issue, but have never had this with MegaTune before or after the TS problem. The MS continues to run fine, it's just the datalog.

Any idea why this might happen? Anything I can add to help troubleshoot?

[LT401Vette]

So...... Looking at the first log "2008-10-31_10.10.28v85_log.zip", at record 6459 when it all went to heck... The car was still running fine but the logged values when wacky? There were no real resets occurring?
Is the above accurate?

Did the gauges go nuts too?

I think I have seen this happen, but couldn't reproduce it. I may have to do some digging here..

[akana]

Yes, the car was still running fine, there were no real resets occurring. The gauge also went nuts - so I shut down TS and started megatune and started logging again in MT, all while the car was still running. Unfortunately, I didn't try restarting TS. I'll give that a shot next time.

I know there's not a lot to go on here - I'll post back if I observe anything further. Willing to help if you can suggest any tests I can run.

[dave]

i've noticed a few resets in my logs as well, but cannot confirm if they are real or not.

[LT401Vette]

I haven't seen this in a while, but if you have it happen you will know. Resets are easy to confirm, look at secL

It isn't resets, resets are fine if the data still looks normal. TunerStudio identifies resets when SecL doesn't follow the proper pattern, count up second by second until it rolls over.

Looking at these logs, all data goes nuts at a point in the log. There are a bunch of resets, but those are just because secL is gone wacky too.

I can think of a way this is possible, but if akana confirms it, we have proof

I will look at the logs to see if maybe there is a byte shift that is throwing it all off.

[Kurt A]

Phil & akana,
That log appears to be the identical issue that I was dealing with this summer. The log just goes nuts along with the gauges in TS. I posted this summer about it and have seen it about 6 times now. I have found no way to duplicate it or can I predict when it will happen. The only thing that I noticed unusual in TS prior to one of these occurences was that TS was showing 2 accel inputs 1 record apart when only one accel was applied. Phil now that we have 2 issues is it possible that it may be a hardware problem (maybe we are both running the same usb adapter or the same laptop)? akana just so that you know, if you restart TS all will be well again till the next time, at least that is the way mine has been reacting.
Kurt

[LT401Vette]

The only thing that I noticed unusual in TS prior to one of these occurences was that TS was showing 2 accel inputs 1 record apart when only one accel was applied.

Good input.... There may be something causing a second thread to kick off.

[akana]

Thanks for posting Kurt, glad to hear a TS restart works for you - I'll be sticking with TS to try to replicate and troubleshoot. Didn't happen last night though. I do have a generic DB-9 to USB converter, running to a Dell laptpop.

The only strange thing I can pick out of the logs is that the delta Time between the last good record and the first wacky record is exactly 0.188 sec in both logs. I realize that delta Time fluctuates between most records, maybe it's just coincidence.

Sorry to trouble you with this Phil, sounds like it's a rare occurrence. I can live with the occasional restart.

[LT401Vette]

No trouble.... I am grateful for any details! There is certainly something happening here, but because it is so rare it is hard to fix.
So far I think both of you are on MS1 and I am pretty sure that is where I saw it happen too.

[Kurt A]

Yes I am running a MS1 unit Phil, my hardware is a IBM T-61 with 1GB Ram using a Belkin USB to serial adapter. Running Win XP pro O/S.