Subaru WRX Forum banner

Cobb AccessPORT + AccessTUNER race vs Open Source

15K views 44 replies 7 participants last post by  SubaruWRC 
#1 ·
I know there are a bunch of posts about the cobb AP vs Open Source, but from what I have found it seems like the information is outdated? I have seen that Cobb can't do data logging (which I believe it can now?) and you need a tuner to tune it for you.

From what I can gather with the AccessTUNER software this gives you essentially open source capability with cobb support? Is this true? Also how hard is it to learn AccessTUNER?
 
#2 ·
#4 ·
SubaruWRC said:
So the fast polling that from raider can do. Is this a big deal? That is the only thing other than the price that open source does better than the Cobb ap right?
Resolution is important; it allows you to get a better picture of what's going on. There are also a bunch of tools/utilities that also help with map-creation / log analyzing, but I know Cobb offers some (I'm just not familiar enough with them to comment on how extensive/useful they are).
 
#6 ·
I think a big advantage for the COBB system (and why most tuners use it) is the real time/live tuning option. For most people doing their own cars the open source method of flash, log, adjust, and repeat is fine, but if I was doing it all the time for a business or something, I would want it real time. Especially on a dyno where time is money!

Opensource has a good community behind it and a lot of people that calibrate the engines themselves. The AP can be used for this too, it just costs more.
 
#8 ·
SubaruWRC said:
Does anyone have experience with both open source and accesstuner?
I've modified maps with ATR on other peoples computers (I'm unable to install ATR on my computer because I don't have an AP).
 
#11 ·
SubaruWRC said:
So someone with open source tuning experience could help me with my ap?
Yes. It may take a little bit longer because some tables may be named differently or setup with different measurement units, but the concepts are the same. Someone familiar with Cobb should be able to give you an OpenECU tune, and vice versa.

The lead designer of OpenECU is now employed by Cobb, for what it's worth.
 
#12 ·
Okay So I Flashed Stage 1.... Its amazing how quickly the turbo spools, I can feel it start to kick in at 2500 RPM. Its really pretty amazing now I want to get to modifying this map, but I don't know how to do that. I attached the files to for everyone to examine to see if it looks correct? Also if it checks out where could I start looking to make improvements?
 

Attachments

#13 ·
SubaruWRC said:
Okay So I Flashed Stage 1.... Its amazing how quickly the turbo spools, I can feel it start to kick in at 2500 RPM. Its really pretty amazing now I want to get to modifying this map, but I don't know how to do that. I attached the files to for everyone to examine to see if it looks correct? Also if it checks out where could I start looking to make improvements?
You've got some FBKC in the first log when you stabbed the throttle; this sometimes happens, and didn't happen in the second log, so I wouldn't worry about it too much just yet. Both logs show some FLKC in them.

Remove AF Learning, Dynamic Adv (keep DAM), and the two MAF columns, so we'll have a bit better resolution. If you can upload the logs as CSV files (native file extension) to Google Docs, it makes it a little easier to read/manipulate a bit. Just make sure to make them public, so we can read them.
 
#14 ·
FBKC is feedback knock? And what does this mean? And what is FLKC? Fine Knock Learning? What does this mean? And what do you mean better resolution? just on the actual document so it is easier to see? Also is the boost supposed to spike like that? it hits 14.9 and I think the boost target is 14.7 which i think is okay because I'm sure it's not that accurate but it drops to 12.5 right after? Is this supposed to happen? Also does it make sense that the wastegate duty only hits 50%?
 
#17 · (Edited)
SubaruWRC said:
And what do you mean better resolution? just on the actual document so it is easier to see?
There's only so much bandwidth in which Cobb can read data from the ECU.

Let's have an example...

Say the bandwidth limit for Cobb is 50 cells of information per second. If you log 10 parameters, that means that each second worth of data will contain five instances of information. If you cut the log down to five parameters, each second worth of data will contain 10 instances of information. As long as the log gives you the proper information via parameters that you need, which log would you rather have in front of you?

SubaruWRC said:
Also is the boost supposed to spike like that? it hits 14.9 and I think the boost target is 14.7 which i think is okay because I'm sure it's not that accurate but it drops to 12.5 right after? Is this supposed to happen? Also does it make sense that the wastegate duty only hits 50%?
It's typical for the boost to spike up to target, then taper off towards redline. According to Cobb map notes, their S1 map targets 13.5 PSI, so what's happening is that when it shoots up past 14, it recognizes the overboost, at which point the Turbo Dynamics tables kick in. They pull WGDC to bring boost back down to where you should be. On both maps, you're a little on the low side, but nothing I'd really see as an issue, according to Cobb specs. Because of the cold weather, you may need to datalog the car come warmer months to see if the underboost gets worse; you may need to run the HWG map in order to hit target when temps go up, depending on how aggressive their IAT compensations are set. For their target boost, it doesn't surprise me that WGDC is so low.
 
#18 ·
#19 · (Edited)
SubaruWRC said:
So here are some more. Just look for the highlighted portions because I took more data in these. In datalog 4 during the highlighted pull the engine really bogged down and was really bad. Let me know if this is an easier format.
https://docs.google.com/file/d/0B-4z3mdp6vfvMEpfQTlOT1JkSHM/edit
https://docs.google.com/file/d/0B-4z3mdp6vfvc0xtNk9PS2h6T2c/edit
I prefer Excel format when looking at logs. What I will say, however, is that you should be a bit more selective in what you're logging. Multiple short logs are easier to deal with than long ones.

You're hitting a lot higher boost than you did before. Did anything change between last log and this one?

Datalog4.xlsx:
Much of the FBKC in this log is caused by driving at too low of an RPM or sudden throttle changes.
2740: Significant FLKC in this section, caused by FBKC in previous areas of the log. 4-4.5° of ignition timing is way too low
2837: Your AFRs are off pretty bad @ idle.
2880: You're pulling significant timing via FLKC. A drop that significant of timing will be felt (4° of timing is probably 30-40hp)

Datalog6.xlsx:
You're pulling timing across the board due to your DAM being too low. This is likely caused by the stuff I brought up in my D4 section
 
#20 ·
How do I use google docs to send excel format?

So the FBKC is not a big deal?
and why would the ignition timing be off how would I correct that?
Again for the AFR how do I deal with that?
And I did notice it the car really bogged down on that pull.

I guess I know my map isn't performing at its best.... It is 15 degrees here so that could be it? And no I didn't change anything from the last datalogs.

Also how do I know if the resolution is bad? The AP told me I picked too many so I decreased it until that didn't come up would that be okay then? Or should I try reducing more parameters?

Thank you for your Help EJ257
 
#22 · (Edited)
So the FBKC is not a big deal?
Not true at all. It depends on the amount of FBKC and in which load cells the timing is pulled.

and why would the ignition timing be off how would I correct that?
This IS the tune. Total ignition timing = base timing + FLKC + FBKC + IAM*(KCA max) + IAT Compensation + ECT Compensation + other compensations

Again for the AFR how do I deal with that?
Again, this is all in the map. Frankly, there's not much outside of closed-loop fueling that you can adjust without a Wideband O2 sensor.

Also how do I know if the resolution is bad? The AP told me I picked too many so I decreased it until that didn't come up would that be okay then? Or should I try reducing more parameters?
As EJ257 mentioned, the ECU has a set bandwidth, and as the number of logging parameters increases the resolution decreases. It's good practice to log only the parameters that you need to observe.
 
  • Like
Reactions: SubaruWRC
#21 ·
SubaruWRC said:
How do I use google docs to send excel format?
It was in Excel format (.xlsx).

SubaruWRC said:
So the FBKC is not a big deal?
Don't lug the motor. Sudden throttle changes can cause FBKC; it's inevitable without getting into things like tip-in, but generally, not really an issue.

SubaruWRC said:
and why would the ignition timing be off how would I correct that?
It's off because the engine is pulling timing. See previous statement, and Subaru Knock Strategy link Zach posted earlier.

SubaruWRC said:
Again for the AFR how do I deal with that?
AFR Correction is going to constantly swing. Since AFR Learning hasn't really changed all that much, I wouldn't be too concerned right now. If AFR Learning starts to get pretty high @ idle, then it's cause for concern.

SubaruWRC said:
And I did notice it the car really bogged down on that pull.
As to be expected.

SubaruWRC said:
Also how do I know if the resolution is bad? The AP told me I picked too many so I decreased it until that didn't come up would that be okay then? Or should I try reducing more parameters?
I told you what to do with parameters based on your first log. You didn't listen :)
 
  • Like
Reactions: SubaruWRC
#25 ·
So I tried to listen.... But I guess my problem is I don't know what I need. I was looking at Brandon's tuning thread and trying pick the stuff he had but I am not 100% sure what opensource stuff correlates to cobb terminology. Is there a translator somewhere?

So the reason I am getting FBKC is my driving? Too hard on the throttle tip in? Driving with the RPM's too low? right? I know it's not ideal driving, but is it hurting my car?

And I did just start reading zacks post about knock control and that is where the different romraider vs cobb definitions gave me a hard time understanding what was going on
 
#26 ·
SubaruWRC said:
So I tried to listen.... But I guess my problem is I don't know what I need. I was looking at Brandon's tuning thread and trying pick the stuff he had but I am not 100% sure what opensource stuff correlates to cobb terminology. Is there a translator somewhere?
I don't think I could be more clear in my post...
EJ257 said:
Remove AF Learning, Dynamic Adv (keep DAM), and the two MAF columns, so we'll have a bit better resolution.
SubaruWRC said:
So the reason I am getting FBKC is my driving? Too hard on the throttle tip in? Driving with the RPM's too low? right? I know it's not ideal driving, but is it hurting my car?
I can't say for sure. Sub-2K RPMs is usually lugging the engine. Large changes in throttle can cause false knock. False knock can cause timing to be pulled for no reason. Too little timing can cause knock, as can too much timing. There's a certain balance you need to retain.

SubaruWRC said:
And I did just start reading zacks post about knock control and that is where the different romraider vs cobb definitions gave me a hard time understanding what was going on
The only difference in naming convention is Dynamic Advance Multiplier vs Initial Advance Multiplier. Feedback and Fine Learning are the same.
 
#27 ·
Only difference that I've noticed is COBB uses DAM(Dynamic advance multiplier) and OpenSource IAM(ignition advance multiplier). Makes googleing a huge pain in the ass IMO.

As for FBKC is your ECU detecting knock and pulling timing. Knock at low RPM isn't as bad in the high revs. Not saying its ok at the low revs, just not as harmful. You need to tune this out.

Posted from a fone.
 
#28 ·
O EJ257 you were right that was really simple lol. Also what is Knock correction advance on that ROM raider post? I don't see that in any of the Cobb parameters? Also I think I may try to save some money and return cobb and try Tactrix cable using a map from OSecuroms & XPT Tuning • Index page is this the best place to get maps? Also how does the Tactrix connect to the computer do I have to buy a cable separately?
 
#30 · (Edited)
Knock Correction Advance (KCA) is a table in the map. See my earlier post: total timing = base timing + KCA + IAT correction + ECT correction + other corrections

Sent from my Nexus 7 using Tapatalk HD
 
#29 ·
Knock correction advance is part of the timing map. It is the amount of timing advance the ECU is allowed to use for FBKC/FLKC. So if the base timing is 20 degrees and you have 5 degrees KC advance the total timing map is 25 degrees but the ECU can only has 5 degrees of correction. This is seperate from IAM/DAM. It is an "allowance" in other terms.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top