Addicted,
I am glag it is working for you (mostly). I think you are right that a bit more fine tuning ought to polish out most of the remaining oddities. If you find you aren't able to tune them out, please post here and we can have a look too.
Lance.
Trigger hysteresis using avload instead of load?
-
Bernard Fife
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: Trigger hysteresis using avload instead of load?
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
-
Bernard Fife
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: Trigger hysteresis using avload instead of load?
Martin,
BTW, I don't see that I mentioned it anywhere in this thread, so I thought I ought to say that the major change in the most recent codes has been to remove the load hysteresis entirely.
This was done because it was defeating the speed hysteresis function, and messing up the shifting, as you know.
There are sound theoretical reasons for this. When the engine speed (rpm) drops after an upshift, but the throttle opening remains the same, the MAP (in kPa) will rise (approximately in proportion to the decrease in rpm). The converse happens on downshifts. This is despite no significant change in the throttle position or vehicle speed (so no change in "driver demand"). This change in load does two things:
The second issue is best dealt with by using a higher percentage of TPS (with appropriate modifications to the TPS/rpm table if necessary). TPS is probably a much better measure of "load" for most applications. In the upcoming 5.100 code, the default value will be 25% MAP (& 75% adjusted TPS).
The latest code still had the load hysteresis parameters in place, but these will be removed for the upcoming 5.100 code. This may generate a one time error in TS if you load a previous MSQ, but don't worry about this, you can ignore such errors.
Also, the default TPS/rpm adjustment table in 5.100 has been edited to more closely match what was suggested earlier in the thread.
Lance.
BTW, I don't see that I mentioned it anywhere in this thread, so I thought I ought to say that the major change in the most recent codes has been to remove the load hysteresis entirely.
This was done because it was defeating the speed hysteresis function, and messing up the shifting, as you know.
There are sound theoretical reasons for this. When the engine speed (rpm) drops after an upshift, but the throttle opening remains the same, the MAP (in kPa) will rise (approximately in proportion to the decrease in rpm). The converse happens on downshifts. This is despite no significant change in the throttle position or vehicle speed (so no change in "driver demand"). This change in load does two things:
- it immediately cancels the hysteresis (the speed and load hysteresis are AND'ed, so there is effectively no hysteresis at all, even though we really require speed hysteresis as well).
- it moves the load index to a new part of the shift table.
The second issue is best dealt with by using a higher percentage of TPS (with appropriate modifications to the TPS/rpm table if necessary). TPS is probably a much better measure of "load" for most applications. In the upcoming 5.100 code, the default value will be 25% MAP (& 75% adjusted TPS).
The latest code still had the load hysteresis parameters in place, but these will be removed for the upcoming 5.100 code. This may generate a one time error in TS if you load a previous MSQ, but don't worry about this, you can ignore such errors.
Also, the default TPS/rpm adjustment table in 5.100 has been edited to more closely match what was suggested earlier in the thread.
Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
Re: Trigger hysteresis using avload instead of load?
Funnily enough I was running the load hystersis at between 30-50kpa to effectively disable it with the previous code..
I WAS running the TPS/Load at 90% and then experimented with reducing it, the performance improved each time. I settled at 50%.. maybe I'll try going a little lower.
I've found that I've also had to redo my shift table as I was finding myself running in areas I previously hadn't hit.. for example running in fourth at 30mph and 1800rpm
I WAS running the TPS/Load at 90% and then experimented with reducing it, the performance improved each time. I settled at 50%.. maybe I'll try going a little lower.
I've found that I've also had to redo my shift table as I was finding myself running in areas I previously hadn't hit.. for example running in fourth at 30mph and 1800rpm
-
Bernard Fife
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: Trigger hysteresis using avload instead of load?
Addicted,
It sounds like we are thinking along the same lines.
Just to be clear for anyone reading through this thread, the "MAP versus TPS for LOAD" parameter is in 'percent MAP' so lower percentage mean more TPS. So the new code sets the default for this parameter to 25% (i.e. 75% TPS and 25% MAP). High TPS percentage values (i.e. lower MAP%) are achieved by moving the slider to the left.
And yeah, unfortunately changing the load set-up will affect the shift tables quite a bit. So it's helpful to set the load first (hence the change in the default so that people start with a value closer to what we now know is optimum). I'm sorry this process results in additional tuning work for you though.
Lance.
It sounds like we are thinking along the same lines.
Just to be clear for anyone reading through this thread, the "MAP versus TPS for LOAD" parameter is in 'percent MAP' so lower percentage mean more TPS. So the new code sets the default for this parameter to 25% (i.e. 75% TPS and 25% MAP). High TPS percentage values (i.e. lower MAP%) are achieved by moving the slider to the left.
And yeah, unfortunately changing the load set-up will affect the shift tables quite a bit. So it's helpful to set the load first (hence the change in the default so that people start with a value closer to what we now know is optimum). I'm sorry this process results in additional tuning work for you though.
Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw
Re: Trigger hysteresis using avload instead of load?
Lance, I reckon I'm pretty much there. I'm even considering screwing the lid on the MSGPIO and tucking it away out of sight as I don't need access to it any more.
I'm using the car almost daily now and the transmission is now very OEM like. I think I mentioned somewhere else, but on the 4th May I'll set off on a 3000 mile round trip to Worthersee in southern Austria in this car. All being well I'll post up a 'success story' on my return with some photos of the car in the alps..
I'm using the car almost daily now and the transmission is now very OEM like. I think I mentioned somewhere else, but on the 4th May I'll set off on a 3000 mile round trip to Worthersee in southern Austria in this car. All being well I'll post up a 'success story' on my return with some photos of the car in the alps..
-
Bernard Fife
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: Trigger hysteresis using avload instead of load?
Martin,
I'm glad it's working for you now. I am really grateful you stuck with it and made me realize what needed to be done; your persistence and patience will pay off for many other users.
Your trip sounds wonderful! Though I see the Rockies (in Canada) fairly often, it's been quite a while since I have seen the Alps, so I am envious. Have a fun and safe trip!
Lance.
I'm glad it's working for you now. I am really grateful you stuck with it and made me realize what needed to be done; your persistence and patience will pay off for many other users.
Your trip sounds wonderful! Though I see the Rockies (in Canada) fairly often, it's been quite a while since I have seen the Alps, so I am envious. Have a fun and safe trip!
Lance.
"Never wrestle with pigs. You both get dirty and the pig likes it." - George Bernard Shaw