Gen 2 Cobb ATR discussion

[rfinkle2]

Thanks to @Speed3eak and @indianaryan for helping me figure out my screenshot issues.

Keep an eye on things, if you haven't been hitting your boost targets before you go bonkers and wot if you decide to alter your load maps.

I hit 274 g's today, and @ around 16.5 psi @ 6500 rpms. I tapered my boost way down up top, because I wasn't sure what consequences changing the load tables would yield.

All of my throttle - req load x gear tables look this way.



I'm more confident now than ever that the boost dynamics table is keeping a lid on your throttle req-x gear maps.

If you notice the right hand side of the load dynamics table is populated and I think you will have an easier time hitting boost targets this way. (I know I did).

[Ckmazdaspeed3]

Yeah, my second gear is still manageable, and I just checked my app tables, and they are all the same. And when I increased the req load in 2nd, nothing changed (i thought it might make it unusable). So i have no idea how cobb limited the power on the ots map... not that ive bothered to look that deeply. its probably obvious and im sure someone will tell me.

Originally Posted by jwilkins88 Stupid phone won't let me edit my post. Edit to the previous post: these tables are how you control boost by gear while doing load tuning. So this is how you would detune your car were you using load based tuning. Basically, at this point, I'm not sure what the hell the boost based tuning does. It's supposed to make it so that the ecu disregards the load tables, but that doesn't seem to be the case since we still have to fuck around with the load tables.

but it seems like it does work. you just have to take the load values out of the game by setting them too high which forces the ecu to look at a different table.

[Ckmazdaspeed3]

Originally Posted by rfinkle2 Thanks to @Speed3eak and @indianaryan for helping me figure out my screenshot issues. Keep an eye on things, if you haven't been hitting your boost targets before you go bonkers and wot if you decide to alter your load maps. I hit 274 g's today, and @ around 16.5 psi @ 6500 rpms. I tapered my boost way down up top, because I wasn't sure what consequences changing the load tables would yield. All of my throttle - req load x gear tables look this way. I'm more confident now than ever that the boost dynamics table is keeping a lid on your throttle req-x gear maps. If you notice the right hand side of the load dynamics table is populated and I think you will have an easier time hitting boost targets this way. (I know I did).

if you're not hitting load values of 2.31 while logging, i wouldnt be surprised if you raised those values to 3 and saw nothing change... i was scared to raise mine at first, but as you can see w my logs, nothing bizarre or scary is happening.

obviously, if u have good control now though, you dont need to do it, but it would be interesting to see... cobb even sortuh talks about doing something like this in the help file.

[rfinkle2]

From the ATR helpfile (italics pg. 17 / 71)...



First we would like to review some of the stock MS ECU logic. The ECU will use the lowest torque target from one of the three groups of tables:
Throttle & Req. Load & X Gear (Norm or High BAT)
Throttle & Requested Load : Baro v. RPM
Throttle & Requested Load
This logic allows several different strategies to be used to calibrate these ECUs. One can set two of the tables to higher values; this would force the ECU to allow torque to be tuned by the one table (or group of tables) with the lowest torque target values. This would allow the Throttle & Requested Load :Baro v. RPM table or the Throttle & Requested Load table to be used to tune torque for all gears. On the other hand, one can set the above two requested load tables to higher values and then the torque targeting tables associated with each gear can be used. This is what we have chosen to do with our OTS maps so wecan try to map boost differently in each gear based on the differing load conditions for each gear. If you increase the desired torque values in these tables, the ECU will do what it can to make additional torque. If you decrease the desired torque values in these tables, the ECU will do what it an to make less torque


----and last but not least italics from pg.71 / 71 of the helpfile.

In order to achieve more consistent boost control, it is essential that the individual "Throttle - Req. Load
(Norm BAT)" tables fall within .05 of actual observed Calculated Load. It is also essential that the values
in the "WG Duty Cycles" are not over-aggressive. These values are a base for the Boost Dynamics system
to start from, and can cause boost oscillations if set too high.


I'm going to speculate that this why only the right hand side of the load dynamics table (authority to decrease wgdc) is populated in the ots maps.

[Ckmazdaspeed3]

Originally Posted by rfinkle2 From the ATR helpfile (italics pg. 17 / 71)... First we would like to review some of the stock MS ECU logic. The ECU will use the lowest torque target from one of the three groups of tables: Throttle &#8211 Req. Load &#8211 X Gear (Norm or High BAT) Throttle &#8211 Requested Load : Baro v. RPM Throttle &#8211 Requested Load This logic allows several different strategies to be used to calibrate these ECUs. One can set two of the tables to higher values; this would force the ECU to allow torque to be tuned by the one table (or group of tables) with the lowest torque target values. This would allow the Throttle &#8211 Requested Load : Baro v. RPM table or the Throttle &#8211 Requested Load table to be used to tune torque for all gears. On the other hand, one can set the above two requested load tables to higher values and then the torque targeting tables associated with each gear can be used. This is what we have chosen to do with our OTS maps so we can try to map boost differently in each gear based on the differing load conditions for each gear. If you increase the desired torque values in these tables, the ECU will do what it can to make additional torque. If you decrease the desired torque values in these tables, the ECU will do what it an to make less torque

exactly what i was referring to.

[jwilkins88]

Originally Posted by Ckmazdaspeed3 but it seems like it does work. you just have to take the load values out of the game by setting them too high which forces the ecu to look at a different table.

Not really. You can achieve the same thing without toggling the boost based tuning. I believe Joe (forzda) was doing this before boost based tuning was even revealed.

My understanding of the toggle was that it would ignore the load tables regardless of the anything else

[rfinkle2]

Guys, this is more relevant to the discussion now, but here is a post @wolly6973 made on the cobb forums.

Travis brought Christian in to explain how to control boost per gear , and is well aware of the year of the car, and that it is "pressure base tuned".

Boost Comp x Gear - 1st not working

In the interest of saving time, I cut and pasted Christian's answer to @wolly6973 here:

Correct, the APP can be used to limit boost in various gears. When the pressure-based boost control option is chosen, the ECU simply uses the values in the Boost Targets table in conjunction with the WG Duty and Boost Dynamics table to operate the boost control system in a close-loop fashion. The Req. Load X Gear tables can also be used to refine the boost control characteristics.

Christian.

[indianaryan]

Originally Posted by rfinkle2 Here is a 2nd - 3rd gear log I did today. The boost curve needs serious work LOL, but the OTS map was a bit rough. Any comments welcome.

I don't know if it's just me, but the spreadsheet is really messed up.

[rfinkle2]

Originally Posted by indianaryan I don't know if it's just me, but the spreadsheet is really messed up.

Its not just you.

I think I know why. Thanks. First log I ever posted.

I jacked the log file up, but you will get the point...(2nd - 3rd gear)

[xxdjfate76xx]

looks good finkle, i need to keep reading more.. im still stuck on just getting the throttle to feel smooth.. i cannot understand this boost logic/WGDC for shit

[rfinkle2]

Originally Posted by xxdjfate76xx looks good finkle, i need to keep reading more.. im still stuck on just getting the throttle to feel smooth.. i cannot understand this boost logic/WGDC for shit

We've all got a lot to learn.

I haven't done much with the throttle app.

If it helps, the boost dynamics table makes adjustments to what is seen by the ecu vs the numbers in the boost targets table by adjusting the wgdc.

The values to the top row / right of zero on the boost dynamics table represent too much boost, and the values under that (negative numbers) are the % change to the wastegate duty cycle, to lower the boost (because it overshot boost target table targets)

The values to the left (top row) of zero represent a miss in the boost target (under boost desired), and the positive numbers allow the ecu to adjust the wgdc higher to get closer to boost target table values.

[xxdjfate76xx]

Heres my 5th gear pull

[rfinkle2]

Originally Posted by xxdjfate76xx Heres my 5th gear pull

@xxdjfate76xx that log looks great. That is a lot of g/s.

If you are running a 93 octane map, you can lean the car out a bit to 11.6. - 11.8

I've updated my throttle - req. load x gear tables (I left 1st gear as is (OTS) and adjusted the remainder to these values.

[Ckmazdaspeed3]

What was your reason for lowering them?

[xxdjfate76xx]

Thanks @rfinkle I'll try that.. This thing is a blast I've been stuck on it all weekend!!!

[rfinkle2]

Another good reason to log calculated load: The max ignition tables are populated with load values, and if you are adding timing, even @ high rpms, the load drops off way under the max load value (2.0) in the table.

e.g. if you look @ rpms and the corresponding load in my logs, my car is in the 1.7's and 1.8's nearing redline (i.e. 5500, 6000, 6500 rpms)

Originally Posted by Ckmazdaspeed3 What was your reason for lowering them?

The other tables are set to unachievable values, but I did quite a few logs, and I wanted to get them to the recommended within .05 of actual calculated load values.

I am still experimenting. LOL.

[rfinkle2]

Originally Posted by xxdjfate76xx Thanks @rfinkle I'll try that.. This thing is a blast I've been stuck on it all weekend!!!

I'm addicted too!

I started a gen1 vs gen2 ATR thread to help clarify some of the differences in tables and their effects here:

Gen1 vs. Gen2 ATR tables

[xxdjfate76xx]

Excellent, yea I've noticed that some of the ranges are different

[Ex_WRX_driver]

Ugh, just read ATR for the 2011s wont be out yet till testing is completed and the AP is "officially" released for the 2011s.

[indianaryan]

Originally Posted by Ex_WRX_driver Ugh, just read ATR for the 2011s wont be out yet till testing is completed and the AP is "officially" released for the 2011s.

That's coming up pretty soon isn't it? I'm anxious to see what changes have been made from the v100 to v101 maps.

[Ckmazdaspeed3]

Originally Posted by rfinkle2 The other tables are set to unachievable values, but I did quite a few logs, and I wanted to get them to the recommended within .05 of actual calculated load values. I am still experimenting. LOL.

Let me know how this works for you. I tried it, but it didn't work for the ebcs at least. Anyway, the problem I see is that load varies so much anyway. When I set it to within .05 one day, if I go out and the next day happens to be much cooler/warmer, everything changes.

Cobb set your preivous table correct?

Put your ebcs in god damn it!

[rfinkle2]

Originally Posted by Ckmazdaspeed3 Let me know how this works for you. I tried it, but it didn't work for the ebcs at least. Anyway, the problem I see is that load varies so much anyway. When I set it to within .05 one day, if I go out and the next day happens to be much cooler/warmer, everything changes. Cobb set your preivous table correct? Put your ebcs in god damn it!

Those values are slightly above those that I've seen from my logs.

I am all about the grimmspeed LOL, but if that is even more difficult to tune, I want to make sure that I am able to get the boost curve I'd like with the factory ebcs.

[jwilkins88]

I think there's an obvious solution Keith.

Disregard K04, Acquire GT28

[PapaSmurf]

Originally Posted by jwilkins88 I think there's an obvious solution Keith. Disregard K04, Acquire GT28

Im def. going to get a decal that says that when i go BT

"Disregard K04, Acquire GT35r"

[rfinkle2]

After looking @ various tables in ATR, more specifically the timing tables, it is getting more difficult to say the load tables do not effect a boost based tune.

The only way the ecu finds ignition advance is based on a load and rpm based axis table.

Maybe the more important question is where ignition advance values come from when seeing loads over 2.0, given that is the max load on the table.

See @Ziggo 's thread as he has attempted to model ignition advance of loads > 2.0, and his mathematics are both less rusty, and likely more advanced than mine.

[rfinkle2]

This thread is courtesy of @BlueStreak , and is particularly relevant.

I permalinked to the last post of his results.

Now we are making progress.

http://www.mazdaspeedforum.org/forum/foru...38/#post815932

[rfinkle2]

I really need clarification on the throttle - req load x gear tables so I put this q. that was never truly answered thoroughly for @wolly6973 out there on Cobb's forums.

I'm still stuck on the effect of these tables and their ability / inability to effect a pressure based tune.




Please provide us some clarfication

[rfinkle2]

OK. I don't think Christian would mind me posting the reply here:

his reply is in italics.

Join Date Jun 2003
Posts 66,666
rfinkle2, we have seen that some cars respond to the settings in the Req. Load X Gear tables. I know...it is odd that not all respond in this manner. I have personally tuned several cars that did not respond to any setting changes for these tables. Being that we cannot consistently confirm that these tables have an affect on all cars, we are looking into the ECU logic to see if we can get a more definitive answer.

We will be sure to post up some additional details once we've spent some more time in discovery. At this time, we are focused on getting the 2011 AP out first, then we will look into this further. Thanks for bringing this up, this is something on our radar. We've found that setting the Req. Load X Gear tables similar to what the car achieves negates their effect on the boost control. This is the best that we can suggest for now.

Thanks,
Christian.

[indianaryan]

Originally Posted by rfinkle2 OK. I don't think Christian would mind me posting the reply here: his reply is in italics. Join Date Jun 2003 Posts 66,666 rfinkle2, we have seen that some cars respond to the settings in the Req. Load X Gear tables. I know...it is odd that not all respond in this manner. I have personally tuned several cars that did not respond to any setting changes for these tables. Being that we cannot consistently confirm that these tables have an affect on all cars, we are looking into the ECU logic to see if we can get a more definitive answer. We will be sure to post up some additional details once we've spent some more time in discovery. At this time, we are focused on getting the 2011 AP out first, then we will look into this further. Thanks for bringing this up, this is something on our radar. We've found that setting the Req. Load X Gear tables similar to what the car achieves negates their effect on the boost control. This is the best that we can suggest for now. Thanks, Christian.

First of all 66,666 posts!

Second of all, very interesting find. Let's get the 2011 AP rolled out so we can get some of these issues cleared up!

[rfinkle2]

Here is an example of a stage 2 map being load capped.

If you look closely @ calculated load, you will recognize repetitive numbers in the 1.92 area.

This car is absolutely following the throtte - req load 3rd gear table.

This is an excellent example.

BTW, I hijacked this from First Stage 2+ logs

datalogs courtesy of @Todd98SE

[Speed3eak]

Originally Posted by rfinkle2 Here is an example of a stage 2 map being load capped. If you look closely @ calculated load, you will recognize repetitive numbers in the 1.92 area. This is an excellent example. BTW, I hijacked this from First Stage 2+ logs datalogs courtesy of @Todd98SE

IMO load cap =/= what you're describing.

Load cap hasn't been a problem for a number of years now.

What seems to be happening is his req load x gear tables are set to the ots values, and the computer is targeting those loads for whatever reason.

I think it was you that posted about an e-mail/response you got from cobb about some cars still using req load values and others ignoring them.

FWIW, the only way I could tell my req load x tables were having any effect was seeing my logged throttle position change after setting them all at unattainable levels.

The pressure based tuning maps seem to still pay attention to load, but the ECU doesn't seem like it can affect anything strongly enough to prevent the load curves from staying pretty smooth.

I'm sure if Todd were to change his req load x tables, he wouldn't see the ecu trying to limit power anymore.

[rfinkle2]

Originally Posted by Speed3eak IMO load cap =/= what you're describing. Load cap hasn't been a problem for a number of years now. What seems to be happening is his req load x gear tables are set to the ots values, and the computer is targeting those loads for whatever reason. I think it was you that posted about an e-mail/response you got from cobb about some cars still using req load values and others ignoring them. FWIW, the only way I could tell my req load x tables were having any effect was seeing my logged throttle position change after setting them all at unattainable levels. The pressure based tuning maps seem to still pay attention to load, but the ECU doesn't seem like it can affect anything strongly enough to prevent the load curves from staying pretty smooth. I'm sure if Todd were to change his req load x tables, he wouldn't see the ecu trying to limit power anymore.

Check on his thread. I asked him to post his max load a table.

Take a look @ the max load A table in a stage 2 tune map when you have the opportunity. (if he posts it up.)

My apologies @Speed3eak, but if the car is following a strict load value, isn't that a very good sign it is being "load capped'?

I am @ work now, and cannot see atr, but one of those load maps has a throttle position around 73, and a load # 1.92 across the board.

RE-EDIT: the throttle position values (corresponding to the load values of 1.92) are in the 62 area for some reason.

[Speed3eak]

See, that's why I think the load dynamics table is changing throttle position to "cap" the load as you so call it.

The reason I didn't think calling it "load cap" was appropriate was b/c there was an entirely different problem the gen 1 guys had to work through in the early years of the cobb AP that went under that name. From what I've read, it was a big problem and caused more than a few tossed rods. I just figured a different name would be appropriate to avoid confusion.

[rfinkle2]

Originally Posted by Speed3eak See, that's why I think the load dynamics table is changing throttle position to "cap" the load as you so call it. The reason I didn't think calling it "load cap" was appropriate was b/c there was an entirely different problem the gen 1 guys had to work through in the early years of the cobb AP that went under that name. From what I've read, it was a big problem and caused more than a few tossed rods. I just figured a different name would be appropriate to avoid confusion.

I'm with that 100%. Something is holding that tune back in the load tables IMO.

It would be easier to see graphed I'd imagine.

I've not changed any throttle settings, because I'm trying to tackle 1 thing @ a time, but maybe it should be deemed a throttle position cap or some shit.

It is no coincidence that the right side of that load dynamics table is the only one populated.

[Ziggo]

Originally Posted by rfinkle2 I'm with that 100%. Something is holding that tune back in the load tables IMO. It would be easier to see graphed I'd imagine. I've not changed any throttle settings, because I'm trying to tackle 1 thing @ a time, but maybe it should be deemed a throttle position cap or some shit. It is no coincidence that the right side of that load dynamics table is the only one populated.

Only the right side is populated because it is still necessary to limit load elsewhere in the tune. Even when boost tuning the the max load tables are still active, and the boost control loop needs the authority to reduce WGDC if this limit is exceeded.

The left side being unpopulated will probably cause strange behavior if you are not running a boost based map.

*edit
Removed "cap" and replaced with "limit" terminology FTW

[rfinkle2]

I apologize to all for using "cap" terminology.

There is some "limit" that needs to be addressed. I am only interested in the progression of atr and the ms3, and can certainly admit to using the wrong terminology.

I have nothing but respect for both Ziggo and Dano, but Christian does quite a bit of tuning, and also sees that there is a phenomenon that needs addressing.

[rfinkle2]

Below are my targets, wgdc table and boost dynamics table.

This combination of wgdc and boost dynamics table has made the stage 2 tune as close as I need it to be to achieve targets in the boost targets table.

[PapaSmurf]

Finkle do u have a log with your current tune? I wanna see how your wgdc is acting since u tweaked it.

[Ckmazdaspeed3]

Ok, so I'm repeating what I just posted in Danos boost tuning thread, but for anybody looking into 3port boost control, I haz answers!

So when I tuned, the first thing I did was figure out wgdc, but because the other tables weren't correct, the wgdc would not work either, so I ended up adjusting it higher than necessary.

After the other tables were adjusted, I had no idea that it was too high, and it caused major boost oscillation.

Today I checked out Danos boost tuning thread, and he figured (I should say, he already knew most likely) it out in a second. So after I did about 13 maps, I pretty much dialed it in.

I get a tiny dip in 4th, but no spikes and no major dips! fuck yeah!

Anyway, the point is, 3 port definitely helps me run higher boost than the stock solenoid (which is good, because I was too lazy to be gentile when I removed the hose that connects to the tip, and I broke it; whoops).

Here are a couple logs, and my tables in case someone else needs assistance with 3 port tuning.

[Ckmazdaspeed3]

Ok, here are the logs... computer was being a retard.