What is the shftBut field in datalogs?
Re: What is the shftBut field in datalogs?
OK, fourth gear, what I found is that if the controller selects 4th gear, but for whatever reason TCC isn't engaged then 4th just freewheels. I have read in a manual somewhere that this is the case, therefore 4th gear is sol1= on, sol2 = on, sol3 = on, TCC = on.
I've messed with the TCC settings, I've even editted the ini to allow me to lower the unlock rpm to 2000. (although I hit 1980rpm in the attached log)
All this is an attempt to stop the controller shifting out of fourth unexpectedly. I just cannot explain the downshifts, or why when it does downshift it also disengages TCC when as far as I can see full time lockup should mean it shouldn't.
On the attached log, @200.00 you can see an event where everything is as it should be in fourth, steady speed and load, then bam! dreq kicks in, it shifts to third and disengages TCC. I've checked every parameter in the datalog, but am at a loss to explain it. The problem is made worse by the fact that after the downshift hysterisis period it shifts back to fourth, but the TCC hysterisis means no TCC so the trans freewheels and the engine just races to 4300rpm.
Note: even though I lowered the TCC unlock rpm in the INI at 245.281 a legitimate shift into fourth drops the rpm to 1980 so TCC legitmately disengages. I'll alter the INI to lower the limit to 1900rpm.
Can you see something that I cannot?
- Attachments
-
- 2015-03-08_09.12.21.msl
- (937.34 KiB) Downloaded 488 times
-
- CurrentTune.msq
- (57 KiB) Downloaded 465 times
-
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: What is the shftBut field in datalogs?
The code is set up to *always* unlocked on downshifts. So it's not surprising that it unlocks on the 4-3 shift.
There are 3 things I wuld think about trying if this was mine:
1)
If that's the case, perhaps you should treat the TCC as a solenoid output (i.e. in a fixed state for a given gear), so that it is always on in 4th. BTW, have you tried the full-time lock-up settings I mentioned above? If you set the Lock at or Above setitng to any rm you are likely to see in 4th (and the unlock rpm lower) you should always have the TCC locked in 4th.I have read in a manual somewhere that this is the case, therefore 4th gear is sol1= on, sol2 = on, sol3 = on, TCC = on.
2) I don't see any activity at 200 seconds into your log. At 255 seconds there is a 4-3 downshift (and the TCC unlocks). At that time the vehicle is going about 53.2 mph and the load is approx. 68.5. You are very close to a 4-3 boundary in your table. Your target gear drops from 4 to 3, and the current gear follows a second later. You can probably fix this behavior in the shift table. Your current table is:
And you might try changing it to:
This makes the shift a function of speed on in that area of the table, so that you aren't bouncing the target gear from 4 to 3 to 4 regardless of changes in the load, etc..
3) You might also want to work with the decel mode parameters to prevent downshifts. See: http://www.msgpio.com/manuals/mshift/V5tune.html#gh Based on your log, you might want to set the "Decel Mode TPS Threshold" to about 35%, and the "Max. sLoad in Decel Mode" to about 60 or so. Then tune from there.
Lance.
Re: What is the shftBut field in datalogs?
Thanks for the reply, some good things for me to think about and try.. although I'm not 100% convinced
Not sure why the time is different, but yes the event you see at 255 is the same as what I see at 200..
This appears to contradict this:Lance wrote: The code is set up to *always* unlocked on downshifts. So it's not surprising that it unlocks on the 4-3 shift.
I 'believe' I have TCC full time mode active, I have the Lock at or Above RPM(rpm) set to 2600 and the Unlock at RPM(rpm) set to 2000.Converter always unlocks on downshift unless TCC full time mode or Jake Brake are active
Looking at the event in question the TCC locks in 3rd at 2943rpm (not 2600) but it unlocks the moment the downshift from 4th to 3rd appears, the rpm hasn't dropped below 2000 so why has TCC unlocked? Do I not have TCC full time mode active?
I must be completly misunderstanding the manual though as I noticed the shift from 3rd to 4th that occurs 6 seconds earlier occurs at 48mph and 55kpa, looking at my table that should have happened 42mph.. what is it that im not seeing?
-
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: What is the shftBut field in datalogs?
A shift can be delayed for a number of reasons, including hysteresis and a number of other things. If you look at the datalog, you'll see that the shift happens (first tGear than cGear go from 4 to 3) as soon as the hysteresis conditions ('hyst' in the datalog goes from 1 to 0) are met. Also, remember that the shift table is based on the average load (avload), not the instantaneous load.48mph and 55kpa, looking at my table that should have happened 42mph.. what is it that im not seeing?
That's not quite right. The converter always unlocks on downshift. It unlocks on upshifts only if full time TCC is not engaged and it is set to unlock in the particular shift's settings. I'll clarify this in the manual.Converter always unlocks on downshift unless TCC full time mode or Jake Brake are active
There was a small but functionally fatal bug in the full time TCC lockup code, and I have fixed it (hopefully) in the attached code and INI. Thanks for spotting this.
Lance.
Re: What is the shftBut field in datalogs?
If TCC does always unlock on downshift then I will I have to review my install from a fundamental point of view. If I'm driving in fourth, brake for a corner, it drops to third, I then accelerate out of the corner so its shifts back to fourth then I'm out of luck as it will have unlocked TCC, it won't then be able to engage fourth due to TCChyst and will just freewheel until the TCChyst period is over. I suppose I'll have to experiment with reducing the TCChyst period.. Also TCChyst doesn't act like normal hyst, rather than having a value of 1 for a given period, TCChyst appears to have a value of 7.5 (if set to 7.5secs) that then decays back to zero.. but doesn't always take 7.5 seconds to decay back to zero.. I've seen examples of it taking much much longer
If I decide to treat the TCC solenoid as a shift solenoid could I somehow use OUTPUT 6? (PAD03?) EDIT, PAD03 is used as a shift lever input, does that mean I have no way of configuring TCC as a shift solenoid?
I could then configure mshift as effectively a 5 speed: 1, 2, 3H, 3M & 4M (I suspect Audi may already do this)
Also, remember that the shift table is based on the average load (avload), not the instantaneous load
I need to bear this in mind and stop refering to instant load, I knew this, but for some reason ignored it!
-
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: What is the shftBut field in datalogs?
I could put an option in the code to "always apply TCC in top gear" if you like. It shouldn't take too long to implement this. Let me know if you want this and I will get to work.
Lance.
Re: What is the shftBut field in datalogs?
If you wouldn't mind I'd love that LanceLance wrote:Matin,
I could put an option in the code to "always apply TCC in top gear" if you like. It shouldn't take too long to implement this. Let me know if you want this and I will get to work.
Lance.
-
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: What is the shftBut field in datalogs?
Okay, give me a little time to work on it (perhaps surprisingly, the TCC control is one of the most complicated parts of the code), and then I'll post some code here for you to try.
Lance.
-
- Posts: 1696
- Joined: Fri Apr 04, 2008 1:28 pm
Re: What is the shftBut field in datalogs?
Here is some code to try, it has the first cut of code to lock the TCC in the highest gear. You'll find the setting under "TCC Settings -> TCC Control Limits -> TCC Application in Highest Gear' where the choices are:
- Normal TCC Control Limits in High Gear (this is the default, and behaves as in previous codes), OR
- ALWAYS Apply TCC in High Gear
This seems to work in limited testing on my bench, and hopefully I haven't screwed a lot of other things up. If you want to try it, that would be great.
Lance.
Re: What is the shftBut field in datalogs?
Something changed with the speed. It is over reading, it WAS spot on.
Datalog attached, it appears your intended change works but its undriveable because of the speed.
- Attachments
-
- 2015-03-10_18.06.34.msl
- (845.63 KiB) Downloaded 479 times