Timing does NOT follow tables bug - 2010 MS3 ATR

[Lex]

I've been running against an issue that is probably a software bug.

After changing both the OL/High throttle tables (Knocking and Not Knocking) to the same values along with the Unknown timing map, the actual timing of the vehicle does not match these command tables at WOT. Attached are the map file along with the log file.

If I increase the timing in these tables, it also has no effect on actual timing indicating this is not an offset issue.

The ATR version is: 1.9.1.0-3024
The AP FW version is: 1.6.5.0-2992

The vehicle is a 2010 MS3

[Lex]

Bumpity bump COBB.

Also can anyone else running ATR for the 2010 (newest version) verify this?

[rfinkle2]

Originally Posted by Lex Bumpity bump COBB. Also can anyone else running ATR for the 2010 (newest version) verify this?

I didn't look @ the map, but looked @ the logs.

The only time I've run into a problem similar to what you are describing is when I had the closed loop max load tables set pretty high.

One of the closed loop max load tables definitely has an effect on timing.

[Lex]

Originally Posted by rfinkle2 I didn't look @ the map, but looked @ the logs. The only time I've run into a problem similar to what you are describing is when I had the closed loop max load tables set pretty high. One of the closed loop max load tables definitely has an effect on timing.

They're all set to 1.0

[rfinkle2]

Originally Posted by Lex They're all set to 1.0

Nevermind! LOL

[Dano]

Lex I have a 2010 in my station and don't recall this becoming a problem with the latest AP/ATR, so I'll take a closer look at his logs when I get the next round back...hes up in the east coast and getting wetter and wetter all the time so it may be a while.

[Zbrit00x]

Thanks guys, perm tuning ftw

[rfinkle2]

Is this zbrit's car @Lex?

[superskaterxes]

is the unknown timing map, MAX A/B timing? maby i missing something because this is a genpu only thing but im sure you have your max tables set accordingly right?

[Lex]

Yes, this is zbrit's car. His max timing values are much above what the car is commanding.

Dano, take a look closely at the car you're tuning - it may not be obvious and if you get a chance the map I posted up. Maybe something I overlooked.

Hi timing is not terribly low, it just won't move up. I will make a map with a lot lower timing to see if it moves down shortly.

[Lex]

Uploaded the timing tables instead of the entire map to make things a little easier to see. It seems that overall timing is between OL and CL maps (the knocking and not knocking tables are identical).

I just sent him a map with OL and CL maps being identical. We will see if this resolves the issue. In the Gen1 speeds, at WOT only the OL timing tables are used.

[rfinkle2]

Originally Posted by Lex Uploaded the timing tables instead of the entire map to make things a little easier to see. It seems that overall timing is between OL and CL maps (the knocking and not knocking tables are identical). I just sent me a map with OL and CL maps being identical. We will see if this resolves the issue. In the Gen1 speeds, at WOT only the OL timing tables are used.

Gen2 maps have identical open loop knock and no knock as well as closed loop knock and no knock.

I'm 99% sure all 4 of those tables are the same in all ots maps.

[Lex]

Originally Posted by rfinkle2 Gen2 maps have identical open loop knock and no knock as well as closed loop knock and no knock. I'm 99% sure all 4 of those tables are the same in all ots maps.

Seems silly to call them OL and CL then doesn't it

COBB has gone through several versions of the timing maps and how they dealt with them over the various maps.

[Lex]

Hey COBB, I figured out your problem

Your 2010 Gen2 timing tables are scaled to the Gen1 axes.

So, the 2.19 load value corresponds to the 2.00 value on the gen1 table.

If you look at the log, at 6500 RPM he is running 1.61 load.

In the Gen2 table he should be running somewhere around 16* of timing and he is actually running 13.5*

If you however use the axes of the Gen1 maps, you will see that the 1.61 load value is between row 7 and 8.

Now back to the Gen2 map, row 7 and row 8 at 6500 RPM are 13.5 and 14*

You can play this game with other values. at 5000 RPM he is running 2.0 load and 6*.

He should be running 8.5*. However the bottom most row (2.19) is 6* which lines up with the bottom most row on the Gen1 which is 2.0 load.

[rfinkle2]

Originally Posted by Lex Hey COBB, I figured out your problem Your 2010 Gen2 timing tables are scaled to the Gen1 axes. So, the 2.19 load value corresponds to the 2.00 value on the gen1 table. If you look at the log, at 6500 RPM he is running 1.61 load. In the Gen2 table he should be running somewhere around 16* of timing and he is actually running 13.5* If you however use the axes of the Gen1 maps, you will see that the 1.61 load value is between row 7 and 8. Now back to the Gen2 map, row 7 and row 8 at 6500 RPM is 13.5 and 14* You can play this game with other values. at 5000 RPM he is running 2.0 load and 6*. He should be running 8.5*. However the bottom most row (2.19) is 6* which lines up with the bottom most row on the Gen1 which is 2.0 load.

Can you guys clear something up for me?

In the latest version of atr, 3 additional lines of resolution were added in the ignition tables.

i.e. 2.06, 2.13 and 2.19.

Did the latest version of atr for gen1 add these (or 3 additional axes)?

Thanks for the research Lex.

[Lex]

Originally Posted by rfinkle2 Can you guys clear something up for me? In the latest version of atr, 3 additional lines of resolution were added in the ignition tables. i.e. 2.06, 2.13 and 2.19. Did the latest version of atr for gen1 add these (or 3 additional axes)? Thanks for the research Lex.

No, then gen1 ATR only has the y-axis go up to 2.00 load. Rob, is this the first version to introduce the higher resolution? If this is the case, I would be even more suspicious and if you're tuning a 2010+ car would be careful until the issue is resolved. The saving grace is that you end up running less timing than you want instead of more.

[silvapain]

Originally Posted by rfinkle2 Can you guys clear something up for me? In the latest version of atr, 3 additional lines of resolution were added in the ignition tables. i.e. 2.06, 2.13 and 2.19. Did the latest version of atr for gen1 add these (or 3 additional axes)? Thanks for the research Lex.

No. The GenJuan tables only go to 2.00 for CL and OL. Only the Max A&B tables go to 2.19


Tapadatass

[Lex]

Can someone please change the thread title fail to:

Timing does NOT follow tables bug - 2010 MS3 ATR

[rfinkle2]

Originally Posted by Lex Can someone please change the thread title fail to: Timing does NOT follow tables bug - 2010 MS3 ATR

@Raider will (fix the thread title).

I'm going to post up my high throttle knocking table.

It contains up to 2.19 load values. (This latest version of ATR is the first to do so)

[Lex]

Originally Posted by rfinkle2 @Raider will. I'm going to post up my high throttle knocking table. It contains up to 2.19 load values. (This latest version of ATR is the first to do so)

You have to post up the table along with a datalog. Also I assume your knocking and not knocking tables are identical?

[rfinkle2]

Originally Posted by Lex You have to post up the table along with a datalog. Also I assume your knocking and not knocking tables are identical?

Correct.

I change all of my tables to identical values and incorporate the max tables to prevent table hopping.

@Lex, and @silvapain, I attached a datalog to the post containing my ignition table.

This is from the testing being done for an e85 mix (with Cobb's e85 scaling).

[dooderek]

Originally Posted by Lex Can someone please change the thread title fail to: Timing does NOT follow tables bug - 2010 MS3 ATR

yes sir

[Lex]

Originally Posted by rfinkle2 Correct. I change all of my tables to identical values and incorporate the max tables to prevent table hopping. @Lex, and @silvapain, I attached a datalog to the post containing my ignition table. This is from the testing being done for an e85 mix (with Cobb's e85 scaling).

Yup, yours does the same thing. The timing values at 2.19 load are used when your actual load is 2.00. Point COBB to this thread as well when you reply to them about the scaling.

[rfinkle2]

I just realized this (my example) may not help much due to the .25 degree increments between load values that I use.

EDIT: Lex sees what he needs to see. LOL

[Lex]

Originally Posted by rfinkle2 I just realized this (my example) may not help much due to the .25 degree increments between load values that I use. EDIT: Lex sees what he needs to see. LOL

At 6k RPM you should be running 19* and you are running 18*. Thanks for confirming by posting up your map and log.

[rfinkle2]

Originally Posted by Lex At 6k RPM you should be running 19* and you are running 18*. Thanks for confirming by posting up your map and log.

Of course.

You caught a bad habit that I think a lot of us may have... tuning from the last load row vs. tuning with actual load values.

[atvfreek]

Those extra 3 rows caused havoc with my tune and kr. When you resave the map in atr it shifts the whole table down three rows. So I had my timing table set at 2.0 load and vertically interopolated. Then once my map resaved it was running close to 2 more degrees of timing across the board. Because it was still on the 2.0 axis that is now 3 rows up.



Tappin

[Lex]

Originally Posted by atvfreek Those extra 3 rows caused havoc with my tune and kr. When you resave the map in atr it shifts the whole table down three rows. So I had my timing table set at 2.0 load and vertically interopolated. Then once my map resaved it was running close to 2 more degrees of timing across the board. Because it was still on the 2.0 axis that is now 3 rows up. Tappin

Did you bring this to COBBs attention? I don't want to work around this because I assume eventually it will be fixed and then the user will be hitting much higher timing values than I intended when they upgrade their ATR and that's not safe.

[atvfreek]

Originally Posted by Lex Did you bring this to COBBs attention? I don't want to work around this because I assume eventually it will be fixed and then the user will be hitting much higher timing values than I intended when they upgrade their ATR and that's not safe.

No I haven't yet, but I do see this being an issue, I'm literally having everybody I tuned with a gen2 get me logs so I can make sure their timing tables are all set. PITA

Tappin

[Lex]

Originally Posted by atvfreek No I haven't yet, but I do see this being an issue, I'm literally having everybody I tuned with a gen2 get me logs so I can make sure their timing tables are all set. PITA Tappin

I would match their actual loads with the timing table and wait for COBB to fix it because when they do, timing will jump up for these guys and cause KR/issues.

[rfinkle2]

@atvfreek, how far do you bump timing up per decrease in load value?

i.e. .5 degree per line of load resolution needs to be addressed, but hopefully won't cause too many problems.

[atvfreek]

Originally Posted by rfinkle2 @atvfreek, how far do you bump timing up per decrease in load value? i.e. .5 degree per line of load resolution needs to be addressed, but hopefully won't cause too many problems.

Yes typically .5 per row. With most of the tunes I have out there, it shouldn't be an issue, as I extend it up a row or two, and I don't push their cars as hard as mine.

Tappin

[G26]

I noticed this issue In my car also. My car doesn't run the timing commanded, all my tables are the same except the max tables which are higher.

I am logging 10 deg always at 6k my tables command 11.5, although no matter where I set it, it always runs 10

Hopefully this issue is fixed as I am leaving a bit of HP on the table.

[Zbrit00x]

Any word from cobb yet? I feel like going in to Cobb SoCal and notify them too!

[Lex]

Originally Posted by Zbrit00x Any word from cobb yet? I feel like going in to Cobb SoCal and notify them too!

I emailed and PMed them but nothing yet.

If you can get in touch that would be great as well - after all - the more people that raise the issue the better.

[Cobb Tuning]

Originally Posted by Lex Hey COBB, I figured out your problem Your 2010 Gen2 timing tables are scaled to the Gen1 axes. So, the 2.19 load value corresponds to the 2.00 value on the gen1 table. If you look at the log, at 6500 RPM he is running 1.61 load. In the Gen2 table he should be running somewhere around 16* of timing and he is actually running 13.5* If you however use the axes of the Gen1 maps, you will see that the 1.61 load value is between row 7 and 8. Now back to the Gen2 map, row 7 and row 8 at 6500 RPM are 13.5 and 14* You can play this game with other values. at 5000 RPM he is running 2.0 load and 6*. He should be running 8.5*. However the bottom most row (2.19) is 6* which lines up with the bottom most row on the Gen1 which is 2.0 load.

It's a question again of dynamically generated axes in the ECU. I'll have to look and see if that calculation has changed from gen1 to gen2.

-David@COBB

[Lex]

Originally Posted by David@COBB It's a question again of dynamically generated axes in the ECU. I'll have to look and see if that calculation has changed from gen1 to gen2. -David@COBB

Thanks David. Also note that it is the latest release of ATR that has exposed this issue. Previous versions have the Y-axis only going to 2.00 load just like in the Gen1 software.

[Cobb Tuning]

Originally Posted by Lex Thanks David. Also note that it is the latest release of ATR that has exposed this issue. Previous versions have the Y-axis only going to 2.00 load just like in the Gen1 software.

Yep, just looked into it. Gen2 has a larger load axes (more values), and it got accidentally shared with the calculation of the Gen1 math for that axes.

It will be fixed in the next release.

-David@COBB

[Lex]

Originally Posted by David@COBB Yep, just looked into it. Gen2 has a larger load axes (more values), and it got accidentally shared with the calculation of the Gen1 math for that axes. It will be fixed in the next release. -David@COBB

Thanks David, when should we expect the next release/fix? I have a few people with Gen2 cars waiting on timing changes and I don't want to put the higher timing values in the higher load sections to compensate for this in the current version/

[G26]

Awesome, should this affect those with very similar to OTS timing already? As in cause KR issues etc? Or can I just copy/paste my values over to the new table when the fix comes out?