Please note, that before I started this thread I did my best to search for an answer :-)

I'm currently trying to get my boost dialed in with the GM 3-port. I used the HOW TOs and set my wastegate duty cycle with the error correction zero'd out etc, that all went per the instructions, THANKS BTW!

When I was finally ready to turn the error correction on, I realized I wasn't able to log load error, or WGDC Correction in evoscan. Some searching, and I found Tephra/Mfreds thread from ~2008. I added the tables and changed the addresses and was happy to see I could log those parameters.

Now the issue is...the WGDC correction is not following my table from the ROM! My concern is that my address for the "load error" in the table is not matching the actual load error the car is reading?

If you look through the plot and compare to the boost settings, you will notice there is never more than 0.5% correction downwards (purple line), even when the Load error (green) is up to 20%. reading the WGDC correction table, this should be giving me a downward correction of 6%.

As you can see in my settings, I did reduce some of the correction percentages as I did hear that the 3 port can tend to overcompensate. Have I just not compensated enough yet for that? I did also notice that the first spike I get during spool up is likely because the error control is trying to compensate for the actual engine load not matching the desired engine load. I will correct this for my next runs

 

Could you post up the actual log? Would be easier for most of us I think to look at that directly as opposed to just the graph where it's harder to see exactly what's going on.

 

Sure can, thanks for helping!

FYI, change the file from "xml" to "xlsx"....I'm still not sure how to post excel files on here...fail :-/

 

+ I would set max total upward correction to all zero's

 

I have considered this, but I decided I wanted some (~3%) upward correction to account for warmer weather (since i'm setting this up in ~45 degree temps).

 

That's not the right way to compensate for temp changes. Use the IAT correction map for this.

I suggest disabling max upward correction as well. And you need to use smaller changes in WGDC correction. Your correction appears to be doing exactly what it's configured to do, you're just being too aggressive with your corrections.

 

hmmm, the only IAT compensation map I have is for ignition timing. I'm still running the 96940011 (USDM 05), someday i'll switch to the 96530706 tephra v7 map.

I'll have to look around for a definition for this in the short term.

Next time the cars out (hopefully tonight), it will be with a less aggressive Correction table(s), see attached.

 

You don't want to zero out the negative load error correction, just zero-ing out the max upward correction is all you want. The guide linked in my sig should help explain the difference with upward WGDC correction between the two tables.

The less-aggressive correction is certainly better, though I think it might still be a bit too aggressive. Keep in mind that 2.5 load is a tiny difference between actual load and desired load, so I wouldn't even use any correction for that small of an error for example.

 

I don't know about the max upward correction table as I don't run Tephra. But, when I first set up my GrimmSpeed I ran a map similar to your original one. I was getting oscillating boost like you are as well. I spent some more time tuning the base WGDC then changed correction as shown in my picture. It holds great now. If you have any correction values in the lowest negative values the ECU will be adding correction to your base WGDC while you're spooling up. It will then create an overshoot, followed by rapid decrease in correction, and keep playing this game oscillating overshoots both ways.

 

Thanks for the info guys. I have adjusted my tables accordingly (I think)

Raptord, read your guide, and am struggling slightly with the logic for example 2, which as I read it, is what you and ReaperX have recommended.

"Example 2 with only Max upward correction zero-ed out: Base WGDC is 50. Boost is too high, ECU lowers WGDC to 48. Now boost becomes too low. ECU increases WGDC back to 50. Later in the pull, boost is too low again. The ECU doesn't increase WGDC because there are no more negative corrections to cancel."

The part in red is what I am struggling with. I interpreted the BEC as a lookup table - maybe i'm incorrect here. So I'm not sure why it wouldn't see boost error as being low, and in turn adjust the wastegate by the value in the table. Does it need to reset after a pull? I made an example (followed yours) with some numbers. The line in red is where my thoughts differed from the example.
I assumed "low" or "high" is -10 or +10 % boost error respectively. What am I missing here?

 

Boost control resets every time to go under the boost control threshold (~100 load), so yes, after every pull or gear change it resets.

Example 2 shows that using the upper half of the BEC table allows the ECU to add WGDC up to your base WGDC. If max upward correction is enabled, it can also increase it beyond base WGDC, but this is what causes the overshoot you were seeing during spool.

 

Ok, now it all makes sense, the "max upward correction table" allows above BWGDC, whereas the BEC table simply allows it to get back up to BWGDC if it was lowered, typically from avoiding the spike coming into boost.

Thanks guys!

had it out this morning, had a pretty steady load curve starting from 2k-7.5k locked in, but have been struggling on getting the curve correct if I say punch it at 4k. This would typically result in a small overshoot, that would end up hurting the rest of the pull because the BEC had pulled a bunch of WGDC and the upper part of my BEC table wasn't enough to allow it to get back to the BWGDC.

I've added some more "correction potential" in the upper half the BEC table to account for this, (previously it was closer to 1.5% max).

 

I don't believe you're correct. The BEC table allows your MWGDC to go up or down based on your desired vs. actual boost pressure/load depending what you're running. The MUCT stops your ECU from adding too much WGDC depending on what boost/load you're at. I will repeat that you need to take your -20, and -17.5 BEC tables to zero and work on your BWGDC to clean up your tune. You're getting the end result based on the band-aid of MUCT instead of having a solid MWGDC.

PS - Waaaaaaaaay too many acronyms.

 

No, he's correct in what he's understanding. BEC does indeed allow WGDC to correct upwards/downwards, but the upward correction will never exceed the MUCT.

I do agree that -20 and -17.5 should be zero, but zeroing MUCT isn't a band-aid; it prevents overshoot during spoolup.

 

Per Raptord's comments I believe it is correct...but I could be wrong.

Also yes you are right, too many acronyms, took me a second to figure out what "MUCT" was haha

As far as I know, there is only a Base WGDC, I read it was named "Max" WGDC earlier on. From your post, are you saying there are two? or are you using the name interchangeably?