Live mapping 850CSi

[Omega man 1969]

Update:

Today I received this in the post :-)



I wanted to see what fault was being thrown by the EML when I changed the values in the sport mode pedal map. I uploaded the modified file to the EML and then ran DIS......EML INTERNAL ERROR . I am assuming this is due to an incorrect checksum......I'm still not sure which are is used for calculation.

While I was at it I thought id check the pedal pot. I uploaded the standard CSi file back into the EML and ran DIS pedal pot test. This test gives pedal angle and voltage. Here is screen shot with the pedal flat to the floor.


Not good but this does explain a few things. I thought maybe it was just the adjuster that needed screwing in some more to allow more pedal movement. In the end I removed the stop all together, higher reading but still not right.



I removed the pedal linkage from the pot and it would still only max out at 80 deg. So I have been driving around at just over half throttle opening until now. Has anyone else here check their pedal values?

One thing I noticed with the pot when turning it by hand was that upto around 70 degs it turns smoothly then it sort of hits a stop where you need to turn it really hard for the last 10 degrees...is this normal? feels a bit like a kick down on an automatic car.

So even with full pedal movement the pot will only give 80 deg and around 4v value to the EML. Is there a specified value for these pots? I guess the EML could be going strange.....maybe some internal resistors have drifted in value.


As you can see on page 27 a valid angle would be up to 100.8 degrees

http://www.e32-schrauber.de/bmw/date...me-1_7-eml.pdf

- - - Updated - - -

Another interesting detail, here is a screen shot of the EML with the Alpina file loaded



And again with the CSi file loaded.

[ArnoNym]

Hi Omega man,

Congratulations to the DK20/DIS set. If I see this right, this is a true UNIX installation, not running on VMware. Is it running faster as on VMware?

I’m going to borrow a PWG from a friend to do some measurements. Will meet him on Sunday, so it takes a while having results.

Regards
Joachim

[Omega man 1969]

Hi, Yes I bought the whole set including the laptop running a Linux installation for a very reasonable price. It runs at a good speed, much better than my other laptop running a virtual machine DIS.....I could never get that thing to work properly.


Thanks for helping out.

I'd really like to see another 8 with DIS plugged in testing the pedal voltage....I see the Bentley E32 manual says 3.1v is acceptable for a manual and pedal stop is adjusted differently for an auto to allow the kickdown part of the travel.....I can't help but think this must reduce throttle plate angle though. I will setup a video cam on the throttle shaft to see how far it rotates at full throttle

Hi Omega man,

Congratulations to the DK20/DIS set. If I see this right, this is a true UNIX installation, not running on VMware. Is it running faster as on VMware?

I’m going to borrow a PWG from a friend to do some measurements. Will meet him on Sunday, so it takes a while having results.

Regards
Joachim

- - - Updated - - -

I'm thinking that the EML knows if it's a manual or auto so somewhere in the code it must have a value that tells it to accept 3.1 volts as full throttle.......that would mean using the Alpina automatic code in a manual car with a manual setup throttle pedal wouldn't give 100 percent throttle opening.

[Timm]

I don't have any experience of the V12 pedal - but, the pedal on the M62TUB 740i has a pseudo kick-down action at the end of travel (where it gets harder to press). This is done to mimic the mechanical switch which is used under the cable-driven throttle on the M62B 740i

[Omega man 1969]

I don't have any experience of the V12 pedal - but, the pedal on the M62TUB 740i has a pseudo kick-down action at the end of travel (where it gets harder to press). This is done to mimic the mechanical switch which is used under the cable-driven throttle on the M62B 740i

Thanks Tim, the Bentley manual does note that the pedal adjustment screw is adjusted for more travel for the auto.

[Fatandre]

So a laptop has to have the so called serial port, right?

[Omega man 1969]

So a laptop has to have the so called serial port, right?

No,, if you have the yellow head it uses the RJ45 Ethernet port.

[ArnoNym]

Ok! Back from the garage now.

PWG - Pedal untouched
PWG_Values_1.jpg

PWG - Pedal pressed down
PWG_Values_2.jpg

DK - Pedal untouched
DK_Values_1.jpg

DK - Pedal pressed down
DK_Values_2.jpg
.

[ArnoNym]

EML interpretation of the PWG potentiometer

I think the EML interprets the max. pressed pedal correct, even if the PWG delivers only a value less than the max. possible angle. See the hardcopys…

German/English:
LL = Leerlauf = Idle
VL = Vollast = Wide Open Throttle

EML - Pedal unpressed
EML_LL_VL_1.jpg

EML - Pedal inbetween
EML_LL_VL_2.jpg

EML - Pedal pressed down
EML_LL_VL_3.jpg
.

[Omega man 1969]

Thanks for you help Joachim, I can rule that out now. I think if I had used INPA it would have shown more clearly the throttle opening on the DME screen.

I'm going to try some more checksum adjustments to see if I can get my modified file to run without throwing an error.

I've started to disassemble the EML code using IDA pro but it's a long road ahead.....I need to learn a lot about the 8051 processor. There are so many tables and functions within the code, I'd really like to understand the entire system rather than just a few areas.

[revtor]

I think it's already clear the EML interprets the pedal travel correctly, but just for reference here's another 850CSi readout with the pedal fully depressed (stop still present):

Hi, Yes I bought the whole set including the laptop running a Linux installation for a very reasonable price. It runs at a good speed, much better than my other laptop running a virtual machine DIS.....I could never get that thing to work properly.

The main problem with running DIS in a virtual machine is the lack of compatible mouse drivers in SCO OpenServer (Unix). The result is a frustrating experience where the mouse gets slower and slower over time, does not register clicks unless you move the mouse afterwards, and sometimes even freezes or gets stuck on the right side of the screen. There is however a workaround: Fixing poor mouse performance of DIS in a virtual machine. This will not improve overall performance of DIS - a native install will still have shorter loading times between the screens - but it completely eliminates all mouse issues and makes working with DIS feel as snappy as if it was a regular Windows application.

[Omega man 1969]

Thanks Revtor. It's interesting that the same pedal pot is used for the manual but the stop is set so that the Kickdown isn't reached. Information like that will be useful for anyone attempting an Auto to manual conversion. A CSi EML file and pedal stop adjustment would need to be added to list of things to do for a conversion.

[Fatandre]

But does that not mean then a different harness needed?

[Omega man 1969]

But does that not mean then a different harness needed?

The EML hardware appears to be the same for the Auto and Manual. The software determines if certain features are used like communication with the gearbox.....if you run the Manual software in an Auto car EML those inputs/outputs will be ignored.

[Fatandre]

I understand.
So as I bought a harness from a auto 850 it should work on the CSI engine?

[Omega man 1969]

Did your car not have a harness already ?

[Omega man 1969]

More progress today. I figured out which area of the EML file is used to calculate the checksum, my modified CSi file works perfectly now.

i whipped the general module out for a cap change this afternoon as my batteries keep going flat, once I've sorted that I will try some more changes to the EML file to figure out what the other 11 x 11 maps.

[ArnoNym]

I figured out which area of the EML file is used to calculate the checksum...

Hi Omega man,

Applies this also to the DMEs? Did you also found where the checksum is located?

Regards
Joachim

Update: Just now, the borrowed PWG is measured. Now I have to prepare the documentation of the readings.

[Fatandre]

and does that make the car run better than previous?

[Mark in SBA]

Keep in mind the variable red-line. There's a side calculation going on there to determine which gear the car's in.

[Omega man 1969]

11x11 map csi.jpg

Here is one of the 11x11 maps........maybe one axis is RPM the other road speed and the value the allowed throttle opening. I will have to drive the car with the data hit trace running and see if the road speed makes the pointer move.

Hi Omega man,

Applies this also to the DMEs? Did you also found where the checksum is located?

Regards
Joachim

Update: Just now, the borrowed PWG is measured. Now I have to prepare the documentation of the readings.

I didn't find the routine for the EML yet but I did find the space for the start and end address, the DME is calculated for the entire file which is 0000h to 7FFFh.


No Andre the EML is only response, you can't change performance by modifying it, only the DME's can do that.


Mark, thanks for the tip......I did have a suspicion the 11 x11 map was road speed related......there are very few inputs to the EML....Pedal position, Engine RPM, Road speed and engine temp being the main ones. It's odd because the 11 x 11 Maps have the same I.D descriptors as the Pedal x RPM 9x9 map, I would have thought road speed would have had its own descriptor. The variable RPM limit is in the DME though.....The Vmax is in the EML right?

- - - Updated - - -


http://www.youtube.com/watch?v=OUDNJ...e_gdata_player


Here is a clip of the data hit tracing in Tunerpro. This is the sport mode.... pedal x RPM map.

[Omega man 1969]

Revtor, if you're reading this......was EWSII installed on 94 cars ? I didn't think my car had that fitted but the DME screen shot says otherwise

[revtor]

Omega man 1969: This is a bit vague but according to the EWS documentation EWS II was not introduced until January 1995. However, another member on this forum said his October 1994 850Ci does have EWS II. The coding changelog for chassis E31 first mentions EWS (without version number though) on September 2nd 1994 and EWS II is mentioned for the 850CSi on November 11th 1994. So it's possible EWS II was not introduced in January 1995 but a few months earlier (and perhaps not on all models and/or market versions at the same time).

[Omega man 1969]

Ok thanks, it's odd because my key doesn't have a chip in it. I'm pretty sure there is no ring on the ignition barrel....ill check .

- - - Updated - - -

The GM checked out ok. For the record only the 10uf cap needs one leg lifting in order to test, all the others test fine in circuit.

[revtor]

Omega man 1969: The EWS documentation is quite clear on this: EWS (I) starting from 01/1994 and EWS II starting from 01/1995. I would not have reason to doubt this information if it wasn't for this forum member with an October 1994 850Ci who claims his car has EWS II. Maybe he is wrong and means EWS I but then I can't explain why DIS says your DME modules have EWS II. What BMW part number do your DME modules have? 1 404 270 or 1 404 739?