RomRaider

It caps the value from the 'Max Wastegate Duty' table (post-compensation) which would have the same effect.


The final wgdc by Max WG Duty (which itself is limited by the "global" cap). That is, the TD corrections are just other variables in the calculation along with initial wg duty. The sum of these variables will always be compared to max wg duty and therefore capped by it, regardless of the individual values.

I don't remember seeing anything that would indicate that TD prop takes priority over TD integral in a way that you described, although I would be interested in seeing logs to this effect. Perhaps I missed something.

Here are two logs.

In the log labeled, RomRaiderlog_20070924_3rd.csv, I had the proportional too high (14%), you can see that below 2864 rpm, the TD integral couldn't windup because the TD proportional was blocking it. Maybe because the TD proportional was calculated first? IDK.

In the second log labeled, RomRaiderlog_20070926_3rd.csv, I reduced the maximum proportional to 8.5% and you can see that the TD integral is now allowed to windup but only by as much as allowed by WGDC init. Ignore the blip at 2578 rpm, I believe that was due to the WGDC by rpm table.

If this is true, it seems like a good way to control TD Integral Windup.

Allright, I think I know what is going on, but I'll have to check some stuff further, so the following is from what I remember from messing with some of the LGT's unique logic.

With the roms like the LGT that actually make use of the initial wg duty alternate table, a constant is sometimes added AFTER being capped by max wg duty. Most roms, this is zero, however with the LGT and some others, it is 0.2. This is why you are seeing a max of 90.2% in these logs. This is important because TD integral positive will not be used if current wgdc > max wg duty (this is something I thought wasn't important before because with most roms this will never occur). Note that it will not clear TD integral cumulative if this is the case, just not add to TD integral cumulative if boost error continues to be positive and wgdc continues to be above max wg duty. The dip in the log dropped wgdc below max wg duty and therefore, TD integral positive was now used briefly until the next line where wgdc was 90.2% again. It held this value because TD integral pos table was not being utilized again. Not until wgdc drops below 90.2%, do we see TD integral ramp up to its max 15% as it should (note: the parameter values are slightly out of sync which is normal due to the way SSM logging is handled by the ecu).

Thanks for the explanation. That really explained why the 4.91 was added on the second log. I wanted to make sure I understood it rather than guessing.

BTW, do you know what the time interval is for the TD intgral cumulative to be updated?

TD integral cumulative will be updated at the same time as TD proportional. That is given that both are being used during the same execution (TD integral also has to meet thresholds for boost error besides the issue mentioned above). But I don't have a number as far as how many times this code is executed over time.

What values did you have in the alternate wastegate duty table for those logs?

What do you mean alternate? Are you talking about the WGDC by rpm table or are you talking about WGDC initial table?

This was the last rom that BigHonu posted. later, he changed the max value in TD proportional from 14 to 8.5.

http://www.RomRaider.org/download.php?id=3539

from this thread:

http://www.romraider.com/forum/viewtopic.php? ... &start=105

I need to clarify something about where I discussed td integral positive being disabled a few posts back. I said that if current wastegate duty is greater than max wastegate duty, then further td integral positive is not allowed (this would only potentially occur with certain roms).

This is true, however I just realized that it also is disabled when current wastegate duty is equal to max wastegate duty. This would apply to all roms. I've updated the original post with this detail. I also had updated it last week to make everything else a little clearer.

Thanks for helping people like me figure out my ecu.