[DWR]

I'm trying to figure out what we can do in the data logging to answers these kinds of questions. Really appreciate the dialogue. It makes me think ...

[DWR]

Originally Posted by DWR

I'm just grasping at straws, and revisiting things I've already deciphered.

In my logs, SA_WERT it is the same in D or S. So, not directly related to maps.

Similarly, STAT_SCHALTDIAGRAMM_AGS_WERT = XE, even in S. I don't think it ever changes, just as others have said. So, how can the trans always be in extreme economy mode?

Originally Posted by _TB_

My 325d is in xe when I. D and in s when in ... S

Oops, data logging error. I see it now.

[Hoooper]

Originally Posted by DWR

So, you are saying now that the map lookup is NOT by current gear and OSS.
It can be in 6th gear and make a shift based on a 5>4 downshift. Since we know the trans can skipshift, why didn't go into 4th? Not trying to be argumentative, just trying to nail down the logic. Wouldn't it have hit the 4>5 upshift first?

Maybe it did hit the 4-5 upshift and that triggered 5th? I would be surprised that it would downshift at the upshift line but since the programming is obviously not set up to deal with the type of scenario you get by flatlining the 6-5 down I would say anything is possible.

FWIW I was thinking it looked like it spiked RPM more on the second test than the first, suggesting a lower gear. It must be in your log but on a scale that doesnt make sense, how can you be sure it was only shifted to 4th? I can see it at the bottom there but the scale is way too small to tell, obviously you have the raw data so would know.

[Mik325tds]

Originally Posted by Hoooper

Maybe it did hit the 4-5 upshift and that triggered 5th? I would be surprised that it would downshift at the upshift line but since the programming is obviously not set up to deal with the type of scenario you get by flatlining the 6-5 down I would say anything is possible.

FWIW I was thinking it looked like it spiked RPM more on the second test than the first, suggesting a lower gear. It must be in your log but on a scale that doesnt make sense, how can you be sure it was only shifted to 4th? I can see it at the bottom there but the scale is way too small to tell, obviously you have the raw data so would know.

Good discussion guys. That's exactly what this thread was made for. Hoooper I see what you're suggesting now. Assuming the strategy is to always find the correct spot in the map for any given combination of OSS and Throttle input, my 70% throttle input could have caused a downshift to 4th and immediate upshift to 5th. However, according to the traces and map that should have already happened at around 50%.
Given the interest in the data, I massaged the plots a little bit more, deleted the torque as currently irrelevant and matched the scaling. Since OSS and wheelspeed are closely related (diff ratio), I calculated the OSS for the second log and also displayed that. Data is attached in the .zip file for anyone interested.

[Mik325tds]

Actually, Hoooper might be right. The downshift occured exactly where the throttle input and OSS hit the 5>4 downshift line. Except that it didn't shift to 4th but 5th.
I guess the only way to find out if the 5>4 line caused the shift is to edit that one too. But that'll have to wait a couple of weeks.

[DWR]

Originally Posted by Mik325tds

Actually, Hoooper might be right. The downshift occured exactly where the throttle input and OSS hit the 5>4 downshift line. Except that it didn't shift to 4th but 5th.
I guess the only way to find out if the 5>4 line caused the shift is to edit that one too. But that'll have to wait a couple of weeks.

Mik325tds, throttle scale is from a Ford map. We don't know scale on the BMW map, do we? Also, it should have 13 data points. Not sure that changes the hypothesis, though.

[Mik325tds]

Originally Posted by DWR

Mik325tds, throttle scale is from a Ford map. We don't know scale on the BMW map, do we? Also, it should have 13 data points. Not sure that changes the hypothesis, though.

I hate to be imperfect. Here's the corrected map. At least we're not lacking pretty pictures these last pages.
According to the trace, I crossed the 5>4 line with OSS 1914 rpm and throttle 78% about a 100ms before the STAT_SA invoked the shift. That is still a pretty good fit.
If this hypothesis is true, it should not shift until 100% throttle input if I'm driving faster than 110km/h (OSS=2715rpm).

[_TB_]

Originally Posted by DWR

Mik325tds, throttle scale is from a Ford map.

Perhaps you can use some of the Bosch factors from EDC16/17?

Throttle (%) = (0.012207*[INPUT-EEPROM]/1)
RPM = (1.0000*[INPUT-EEPROM]/1)
Temperature (C) = (0,1000*[INPUT-EEPROM]/1)+(-273.1)
Time (uS) = (0,976*[INPUT-EEPROM]/1)
Engine Torque (NM) = (0,1000*[INPUT-EEPROM]/1)

[DWR]

I've plotted a point (star) at @1900, 78% throttle. Unless I've made a mistake, it is not near any shift point, on this map.

[Mik325tds]

Originally Posted by DWR

I've plotted a point (star) at @1900, 78% throttle. Unless I've made a mistake, it is not near any shift point, on this map.

Well, yes, it is below the 5>4 shift line, thus it would determine to shift. Just like if the OSS would fall below the 6>5 line.

[Mik325tds]

Originally Posted by _TB_

Perhaps you can use some of the Bosch factors from EDC16/17?

Throttle (%) = (0.012207*[INPUT-EEPROM]/1)
RPM = (1.0000*[INPUT-EEPROM]/1)
Temperature (C) = (0,1000*[INPUT-EEPROM]/1)+(-273.1)
Time (uS) = (0,976*[INPUT-EEPROM]/1)
Engine Torque (NM) = (0,1000*[INPUT-EEPROM]/1)

Not sure what to do with that. What value is EEPROM?

[DWR]

Originally Posted by Mik325tds

Well, yes, it is below the 5>4 shift line, thus it would determine to shift. Just like if the OSS would fall below the 6>5 line.

Yes, and it shifts back up to 6th when it drops below 25% throttle.
Just don't understand why it did not shift down to 4th. It met the criteria. It was solidly there, not just a tip in.

If you gave it more throttle, would it go down to 4th when it crossed the 4>3 shift line? If so, that means the shift lines are not gear specific. That would be very interesting.

[_TB_]

Originally Posted by Mik325tds

Not sure what to do with that. What value is EEPROM?

That is the value found in the file. (16bit HiLo)

[Mik325tds]

Originally Posted by DWR

Yes, and it shifts back up to 6th when it drops below 25% throttle.
Just don't understand why it did not shift down to 4th. It met the criteria. It was solidly there, not just a tip in.

If you gave it more throttle, would it go down to 4th when it crossed the 4>3 shift line? If so, that means the shift lines are not gear specific. That would be very interesting.

Interesting logging session this morning. I tried some 100% throttle inputs at various speeds to see which gear it ends up in. The first try then revealed no shift at all. It stayed in 6th gear. However, the trace of STAT_SCHALTDIAGRAMM_AGS_TEXT revealed that I was in 'HM' mode - whatever that is. My only guess is that it means "Heating mode" as I had only driven 4 miles so far.
More traces later.

[DWR]

Originally Posted by _TB_

That is the value found in the file. (16bit HiLo)

You are correct, no doubt there are values in the file for the axis. Unfortunately, we are uncertain what those values are.

[DWR]

Originally Posted by Mik325tds

Interesting logging session this morning. I tried some 100% throttle inputs at various speeds to see which gear it ends up in. The first try then revealed no shift at all. It stayed in 6th gear. However, the trace of STAT_SCHALTDIAGRAMM_AGS_TEXT revealed that I was in 'HM' mode - whatever that is. My only guess is that it means "Heating mode" as I had only driven 4 miles so far.
More traces later.

That's interesting, I assumed HM was Hot Mode.

[_TB_]

Originally Posted by DWR

You are correct, no doubt there are values in the file for the axis. Unfortunately, we are uncertain what those values are.

Yes that is some of the hard part. I'll see if the headers are normal Bosch headers.

As for the maps, they (the 3d and 2d) are pretty straightforward to find.

if we look at the file in 2d, 16bit HiLo:

View post on imgur.com

The pattern is pretty obvious - lets look at the values.. (disregard the ols map identifying for a second..)

View post on imgur.com

The two first values are 8 and 10 - that means we have a 8x10 3d map. The 8 first values after 10 is the X axis, and the next 10 are the Y axis. Then the next 80 values are the output values.

3d view of the map:

View post on imgur.com

The next task (for me) is to compare a D3 file with a B3 file since they are almost identical - apart from being for a gasoline and a diesel engine. This will give us vital information about the shift maps.

(The map in question here is not changed between the D3 and the B3 file)

[DWR]

Originally Posted by _TB_

Yes that is some of the hard part. I'll see if the headers are normal Bosch headers.

As for the maps, they (the 3d and 2d) are pretty straightforward to find.

if we look at the file in 2d, 16bit HiLo:



The pattern is pretty obvious - lets look at the values.. (disregard the ols map identifying for a second..)



The two first values are 8 and 10 - that means we have a 8x10 3d map. The 8 first values after 10 is the X axis, and the next 10 are the Y axis. Then the next 80 values are the output values.

3d view of the map:




The next task (for me) is to compare a D3 file with a B3 file since they are almost identical - apart from being for a gasoline and a diesel engine. This will give us vital information about the shift maps.

(The map in question here is not changed between the D3 and the B3 file)

I believe the reason that WinOLS does not recognize the shift maps is because they do not take that form. However, you just reminded me of how I found the axis scales before. Here they are, first column is the value in the file, second is a scaled axis.
0 0
40 4
180 16
330 30
480 44
550 50
620 56
630 57
920 84
960 87
970 88
1000 91
1100 100

[DWR]

Here is my inventory of potential maps:


The shift maps are not in this inventory because I did not originally search for maps with axis greater than 13. I found those by visual inspection.

Has anyone identified what any of these maps are for?

It is a great idea to compare various cal file. That's what I am doing also. Still, it is a tremendous task.

[dave205t]

Originally Posted by DWR

Here is my inventory of potential maps:
It is a great idea to compare various cal file. That's what I am doing also. Still, it is a tremendous task.

Hi Guys,

I should be able to help out there. If you have a single confirmed start address of a shift map for me, i will be able to produce a list of maps belonging to that same shift program.

Best regards, Dave

[DWR]

Dave, info on the way.

[Mik325tds]

Here are the rest of the logs from yesterday. This one was to confirm that there would be no shift if I go faster than 110 km/h (OSS>2715rpm) and roll in the throttle to almost 100%. I did this twice with different roll-in speeds. Both instances resulted in shifts to 5th. That would confirm my theory of the external modifier since the map clearly didn't pass a shift point.