Ultra-strange timing behavior with Cobb - any ideas ? Pics and logs inside

[cstavaru]

Hi,

6MT FBO RB + Meth here. Running Cobb. It begun occurring on a map in which I increased the timing by 1.5 degrees at higher RPMs (5500+) compared to a previous map that did not exhibit this issue (but I will have to re-load that map and check again - maybe now it happens on that too but I kind of doubt it).

It first occurred today on the dyno, and then when I did some circuit runs this evening it happened again. Basically the timing is completely erratic, in the log on row the timing is correct, the next row the timing is ZERO (or close to).

I have attached a log picture (it isn't pretty) and the CSV log. There are some timing corrections on some cylinders but I am sure they are not related to this problem. I have also attached a dyno picture to see how it affects the torque/power. The dyno is not related to the posted logs but the dyno log exhibits the same problem.

It seems to me like a torque limitation algorithm in the ECU or something. Any ideas on how to fix this ?

[benzy89]

Flat line problem........ What gears are you shifting into? It's a known issue with FBO 6ATs going from 3rd to 4th or 4th to 5th that you flatline. Didn't know it plagued the 6MTs also

[cstavaru]

Flat line problem........ What gears are you shifting into? It's a known issue with FBO 6ATs going from 3rd to 4th or 4th to 5th that you flatline. Didn't know it plagued the 6MTs also

Doesn't look like a flatline to me...the timing alternates between the correct timing, and 0. A flatline would be 0 timing for a second or so.

[JoshBoody]

not flatlining, but a super stutter. this happen on the street? seems to coincide with rpm

EDIT: compare closely with rpm slope in your other "good" logs. You do have some boost overshoot compensation, but this is unrelated. Also I remember you doing some heavy experimentation with various tables. What's different in your map relative to e30 other than typical boost, fuel, ign?

[rader1]

Have you tried disabling torque intervention by ignition? Although, like josh said, this looks more like it's RPM related rather than torque.

[cstavaru]

not flatlining, but a super stutter. this happen on the street? seems to coincide with rpm

EDIT: compare closely with rpm slope in your other "good" logs. You do have some boost overshoot compensation, but this is unrelated. Also I remember you doing some heavy experimentation with various tables. What's different in your map relative to e30 other than typical boost, fuel, ign?

Yes, seems to begin right after 5500 or so. Just where I increased the timing from the previous map. The stutter cannot be felt, car feels just a bit down on power, but it does keep up with a maxed out FBO+meth and stock turbos. It first happened on the dyno, but now I see that it happens on the street too. However I did not yet try again the previous map with less timing.

My base map for this is Stage 2+FMIC Aggressive, not the E30. I am runnig 100RON fuel (94-95US) and meth. Not many tables changed, only boost, load, afr, timing.

There are no error codes.

I wonder if disconnecting the battery would do anything, because it begun while the car was on the dyno with all the dash lights flashing because of ESP errors, etc.

[cstavaru]

Have you tried disabling torque intervention by ignition? Although, like josh said, this looks more like it's RPM related rather than torque.

Yes, it doesn't help.

[Josh@Cobb]

Is this occurring in all cylinders? Normally we only log Cyl 1 Timing, but it might be worthwhile to log all 6.

[cstavaru]

Is this occurring in all cylinders? Normally we only log Cyl 1 Timing, but it might be worthwhile to log all 6.

I will have to check. Will log all cylinders tomorrow.

[JoshBoody]

Yes, seems to begin right after 5500 or so. Just where I increased the timing from the previous map. The stutter cannot be felt, car feels just a bit down on power, but it does keep up with a maxed out FBO+meth and stock turbos. It first happened on the dyno, but now I see that it happens on the street too. However I did not yet try again the previous map with less timing.

My base map for this is Stage 2+FMIC Aggressive, not the E30. I am runnig 100RON fuel (94-95US) and meth. Not many tables changed, only boost, load, afr, timing.

There are no error codes.

I wonder if disconnecting the battery would do anything, because it begun while the car was on the dyno with all the dash lights flashing because of ESP errors, etc.

This is not related to your timing setpoints, but an intervention based on traction, dyno load... something with changing rpm ramp rates. Yeah, all 6 cyls will be interesting.
Another thought is that rpm slope is result of ignition, although I think its the other way ign result of rpm.

You mentioned circuit, so happened outside the dyno BUT only after dyno?

[rader1]

You're running the OEM flywheel, correct? Maybe you're seeing the first symptoms of the flywheel "chatter" that has been accused of causing misfires in high HP cars?

EDIT: didn't see the part about it not happening until after throwing errors on the dyno. Give disconnecting the battery a try, see what happens.

[Terry@BMS]

I would guess you're fluttering between a torque limiter of some sort there.

[cstavaru]

This is not related to your timing setpoints, but an intervention based on traction, dyno load... something with changing rpm ramp rates. Yeah, all 6 cyls will be interesting.
Another thought is that rpm slope is result of ignition, although I think its the other way ign result of rpm.

You mentioned circuit, so happened outside the dyno BUT only after dyno?

Circuit...well...street It happened only after the dyno, but it is a new map for the dyno (timing increased by 1-1.5deg. after 5500rpm), never tested before the dyno and only tomorrow I can revert and test with the previous one.

[dzenno@PTF]

you're hitting a torque limit...log torque limit active

[themyst]

are you running race logic aka boosa?

[benzy89]

I was looking at the boost plot

[JoshBoody]

you're hitting a torque limit...log torque limit active

If this is the cause, instead of altering the tuning, what table(s) does he need to amend for a higher limit? Nice tidbit on the log channel; I tried output and that seemed sorta useless.

It’s a little surprising, since e30 and stg2+ maps have the same table values in the limit section IIRC. And OPs load, timing is very similar or lower than others running e30.

[themyst]

If this is the cause, instead of altering the tuning, what table(s) does he need to amend for a higher limit? Nice tidbit on the log channel; I tried output and that seemed sorta useless.

It’s a little surprising, since e30 and stg2+ maps have the same table values in the limit section IIRC. And OPs load, timing is very similar or lower than others running e30.

I had that exact same timing activity on race code aka boosa. Which is why I asked the question. I've never seen that kind of activity on non-boosa code.

[cstavaru]

I had that exact same timing activity on race code aka boosa. Which is why I asked the question. I've never seen that kind of activity on non-boosa code.

And how did you solve it ?

I am running regular ATR software.

[cstavaru]

If this is the cause, instead of altering the tuning, what table(s) does he need to amend for a higher limit?

This would be a great information

Not related to the problem in this thread, I also noticed that when in spool mode the timing targets are not reached, the ECU has a logic of its own and limits the timing to lower values than the set target (with no corrections logged). Probably other torque limits reached. I just noticed that the "disable torque intervention by ign" makes the low-end much snappier, I think the timing target is now reached, I will have to log and see.

What's strange is that when I log the "Requested Torque" it is always a fixed value of 510 or so, while the "Actual Torque" goes over 700 sometimes. However the "Torque Limit Active" stays at "0". (but I did not log these values lately, this is from previous experience).

[654]

This would be a great information

Not related to the problem in this thread, I also noticed that when in spool mode the timing targets are not reached, the ECU has a logic of its own and limits the timing to lower values than the set target (with no corrections logged). Probably other torque limits reached. I just noticed that the "disable torque intervention by ign" makes the low-end much snappier, I think the timing target is now reached, I will have to log and see.

What's strange is that when I log the "Requested Torque" it is always a fixed value of 510 or so, while the "Actual Torque" goes over 700 sometimes. However the "Torque Limit Active" stays at "0". (but I did not log these values lately, this is from previous experience).

Please log all those Torque related values now, Torque Limit Active included.

[cstavaru]

Very interesting findings today. I did some logs with the "problem" map, with the previous working map, and with a map derivated from the "problem" map but with the torque limit increased by 30% in certain areas in the "Load to Torque Limit" map.

The good news is that the previous working map, still works perfectly.

You can see that in the log where I increased the torque limit, the timing drops are more rare and of less amplitude. Which leads me to believe that if I increase the torque limit even more, I may get a working map with the increased timing after all. It remains to be seen at the next logging session.

Attached you have the log pictures from the "bad" map, from the "so-so-torque-increase" map, and from the "good" map, in this order.

[cstavaru]

Here is a more complete log from the map where I increased the torque limits in the "Load to Torque Limit" table. You can see how the boost TARGET (not the actual boost) oscillates too !

This leads me to believe that the torque limits may need DEcreasing instead of INcreasing to provide timing and boost target stability ! But this is just a supposition of mine...

[themyst]

And how did you solve it ?

I am running regular ATR software.

well, I solved it by stacking a piggyback for boost control, but cobb-only I never had that issue from running regular ATR mapping. it could definitely be a torque limit issue like dzenno said if you're not running boosa.

[dzenno@PTF]

@cstavaru, how are your load actual and requested in this last log posted? Can you give it a bit more resolution/scaling as well?