Interest in BMW engine development

[jcolley]

Greetings all,

Long time lurker / occasional poster. Years ago (before buying an old house) I was looking into getting into FB and the BMW S1000RR motor fascinated me. A lot of time has gone by and I'm surprised this hasn't been figured out yet with any real showing.

Fast forward to now, post Naval retirement, and I play with BMW engines and ECUs all day. What started as a hobby with my own M5 turned into not a hobby anymore and I build and tune the last N/A BMW M3 and M5 engines. Over the last year, I wasn't satisfied with the limited reference available for the DME (digital motor electronics aka ECU) in these cars and started decompiling and disassembling the binaries. I've been inside of pretty much every subroutine that runs on the computer and pretty familiar the the Motorola MPC 563 processor it runs. Most of this was done in an effort to replicate some of the BMW routines in the Motec M1 ECU in a firmware package I'm writing.

A few weeks ago, I had the idea of looking at the S1000RR motor again and while most of the rest of the BMW DME line has moved to TriCore processors, I was pleased to see the S1000RR uses the Bosch BMSK ME9 DME based on the Motorola MPC 561 processor, almost identical to the MPC563s I've been playing with.

I've been emailing with Jake Latham a bit on his Stohr, may head up to look at it soon while it's still in Canada.

To keep it short, I'm 100% certain after looking through some code and schematics of the bikes that I can make it work. The CAN (two wire communications protocol) message framework it uses is identical to what I have seen in the cars and programmed Motec to emulate. Making the engine work in a car seems fairly straightforward with possible an external box to emulate and pass signals needed that are missing from the bike *if* they can not be coded out of the DME completely.

I'm debating picking up and ECU on ebay to play with in the next few weeks.

If anyone has any interest in this, please let me know. Not sure the rules on commercial links and forum sponsors, so won't post an email, but feel free to ask questions or PM. If allowed, I can post all the schematics for engine to the bike if there's any interest.

[holmberg]

Somebody got the S1000RR working in a race car.

https://www.tapatalk.com/groups/dsrforum/stohr-wf1-s100rr-t11533.html


Greg

[jcolley]

Somebody got the S1000RR working in a race car.

https://www.tapatalk.com/groups/dsrforum/stohr-wf1-s100rr-t11533.html


Greg

I'd seen that one, but it uses a Haltech ECU.

I was under the impression FB still required the stock OEM ECU with only reflashing allowed.

[99sh]

Right, stock ECU required. I'd be interested in your development. I lost the bubble, how many generations of S1000RR motors have there been? Essentially curious to hear if the ECU can be used across a wide spread of Model Years.

[jcolley]

Right, stock ECU required. I'd be interested in your development. I lost the bubble, how many generations of S1000RR motors have there been? Essentially curious to hear if the ECU can be used across a wide spread of Model Years.

I totally lost it as well (and haven't heard anyone outside of submarines use that term before).

Not entirely sure on the updates to the engine. From the core perspective, the crank is shared up to current production:
https://www.realoem.com/bmw/enUS/par...&q=11218536310

Similar results for rods and pistons. So the rotating assembly is of a consistent design.

For ignition coils there were some updates, but don't know if it's significant.

Fuel injectors are unchanged from start of production to now.

So, from the perspective of engine control, not a lot of differences. I believe the later versions had modified head ports and would be desirable as well.

The HP4 version is race oriented and has a lot of features that would make car installation more problematic, but in my experience in tuning the M cars, most functions of this type a configuration flag in the calibration section of the DME that allows enable/disable and some version control.

I have a couple of the binary files loaded up for the BMSK (bike) version of ME9 as well as a complete map set for the 4 cylinder car version of the ME9. I don't have a map set for BMSK yet, but looking at both in the disassembler should allow for determining what areas of code are shared from bike to car and narrow down the unknowns for the bike specific functions.

[ghickman]

I may have something constructive to add here.

In F1000 the aftermarket ecu's are not legal.

An F1000 owner (one of the cars I built) in Washington state has managed to get the BMW to run on the stock ECU. There's a long back story to this that involves some help from former Williams F1 team engineers which most of us would never have access to.

Just a reminder to those thinking about taking a newer Generation engine and back dating it with previous Generation parts I am told that this is not legal in F1000. I believe though that this is currently being done to get the GEN 5 Kawi to run by back dating the fuel injection and ECU.

[jcolley]

Thanks for the input Gary.

I think it would be fairly straightforward to make the DME run any of the engines of any year range, given the similarity of the parts. At least that was the purpose of my comparison. On the street car side, we use this website a lot to determine parts compatibility. Often times with BMW, a part number change can denote little more than a change of supplier and not a design change itself. The S1000RR has a chassis code (internal production code) of K46. From the first dates of production in 9/08 to the last 2018 (4/18), part number cross references are:
9/08: https://www.realoem.com/bmw/enUS/par...0_0507%2C0517_
4/18: https://www.realoem.com/bmw/enUS/par...00_RR_17_0D80_

What I'm not sure of, is it allowed to run a 2018 ECU on a 2008 engine, perhaps someone more knowledgeable in the rules can provide some insight.

What I do know, BMW does not like to reinvent things nor remove them from the code if they are no longer needed. For example, the E60 M5 (2006-2010) used the SMG 3 transmission, yet retained all the code to run the SMG2 from the E46 M3. In fact, the E92 M3 ECU has all the code to run both SMG3 and SMG2, for which neither were an option from the factory.

While this leads to massive code bloat, it's also beneficial as it gives some hope that many of the available functions of the ECU are controlled by simple configuration switches in the calibration (tuning) area of the ECU.

Having been quite far into the mss60 and mss65 DMEs, we've been able to delete EWS (key anti-theft), enable MAP sensors on cars that never had them, delete MAF sensors (which are great until you significantly alter the intake tract), and use different transmissions.

Getting more into it, I think making the BMSK ME9 run into a car is something quite doable without great effort, at least given what I've already been doing. I'm finishing up a project transplanting the dual clutch transmission into the M5s (not offered from factory) and if all goes well, will be driving within a month. Given that, I need something to chew on next and would appreciate any guidance from the brains here on what would and would not be permissible to focus the effort.


Edit: The 2016 and up models seem to use a BMSK 17.2 DME which makes sense as that's a variant of the MEVD 17.2 used in the current range of M3, M4, M5, and M6. These would be possible as well, but more difficult as they are based on Tricore processors for which no decompiler is available. With the older (2009-2015) ME9 DME, it's power PC based and IDA Pro spits out both assembly and decompiled C source code, greatly simplifying the effort to reverse engineer (at least for me as I don't speak assembler well).

[ghickman]

Thanks for the input Gary.

What I'm not sure of, is it allowed to run a 2018 ECU on a 2008 engine, perhaps someone more knowledgeable in the rules can provide some insight.

As far as I have been told and the way I see the rules interpreted back dating the engine (new engine older ecu / injection) is not legal.

This could be a potential problem at the RO's this year if someone was inclined to push this issue via a protest. Best way to solve this is petition the CRB with a "Rules Clarification".

As I know it 2014 and newer engines are subject to CRB review and could require submission of dyno sheets to the CRB should they request it. This all goes back to the whole restrictor threat from a few years ago and the thought that engine wars might begin with competitors coming in with big money and new high HP engines.

As most know the newer M/C engines are bound be multi axis gyros and sometimes throttle by wire. Either one of these posses major hurdles to get over. I'm not a code guy and can offer zero input there.

Much more HP added to the F1000 class and we WILL be faster than WELL DRIVEN Formula Atlantic cars. Most of the time now I out qualify the entire P1 field. Where this all stops nobody knows. From my viewpoint I believe it's limited to driver talent, these cars are crazy fast. We just got back from testing at Sonoma and my times now are in the 1:30 flat, the Pro Atlantic cars were only 2 sec quicker than this back in the mid 2000's.

[jcolley]

Thanks again.

Well, I just bit the bullet and bought a 2013 ECU on ebay with no key. I'll get it connected, read out the flash memory and see what's there.

Is throttle by wire prohibitive due to rules or technical complication? I was a hold out for a long time, but now having the ability to alter pedal feel, I'm sold 100%.

The tuning tools I use for the other BMWs are from AlienTech, specifically the PowerGate3. You can pick up the flash tablet and encoder tablet for around $500 for both. From there if you have a good map set defined, you can use the freeware TunerPro to alter maps as desired. Basically, once the engine was running and the DME program modifed as needed, the individual could have full control over the DME's parameters (believe me, there are a lot...if it's anything like mss60, around 11,000 configurable items).

To me, that's huge for keeping cost down.

At this point it's all conjecture though. More to follow.

[jcolley]

Ok, I lied. I'm pretty obsessed on this at the moment.

So the ECU I bought is from a 2013. Looking at available part numbers for them, there is only one listed. This means that the options for that bike in that year for Race ABS and traction control are functions of the flash program used and the associated hardware. So, at least most of that can be deleted with the correct flash in the ECU.

Also, this bike had exhaust flaps used to tune pulses in in the primaries. If this is problematic with regulation, it could be deleted as well.

If not, could make use of it for tracks which have noise restrictions, all inherent to the OEM design.
https://www.realoem.com/bmw/enUS/sho...diagId=18_0860

[ghickman]

Ok, I lied. I'm pretty obsessed on this at the moment.

So the ECU I bought is from a 2013. Looking at available part numbers for them, there is only one listed. This means that the options for that bike in that year for Race ABS and traction control are functions of the flash program used and the associated hardware. So, at least most of that can be deleted with the correct flash in the ECU.

Also, this bike had exhaust flaps used to tune pulses in in the primaries. If this is problematic with regulation, it could be deleted as well.

If not, could make use of it for tracks which have noise restrictions, all inherent to the OEM design.
https://www.realoem.com/bmw/enUS/sho...diagId=18_0860

This is where the whole back dating thing comes in. If part numbers for ECU's are the same for multiple years then who cares.

Everything you brought up here is a real gray area.

In the case of the GEN4 GEN5 Kawi the ECU's are different and same goes for the throttle body etc. With the BMW not sure about this I don't own an F1000 with a BMW. I do own 2 F1000's one with Suzuki power the other with GEN4 Kawi power.

Not sure if throttle by wire is legal.

[jcolley]

I'll keep posting here with stuff as I collect info. If nothing else, may prove useful down the road for someone else should I abandon ship.

Wiring diagrams for the DME.

BMSK F36 SSP.pdf
BMSK F53 SSP.pdf
BMSK F61 SSP.pdf
BMSK F60 SSP.pdf
Attachment 79907
BMSK F68 SSP.pdf
BMSK F70 SSP.pdf

If this belongs in another subforum or isn't desired here, mods please let me know.

-Jim

[JakeL]

subscribing because it's interesting

-Jake

[jcolley]

Not much to update, busy week so far, but received a 2013 DME. Hoping to get it opened up early in the week and get a read out of the flash memory.

I did find a few stock files around the internet for comparison, but haven't loaded them into the disassembler yet.

[Jnovak]

As far as I have been told and the way I see the rules interpreted back dating the engine (new engine older ecu / injection) is not legal.

This could be a potential problem at the RO's this year if someone was inclined to push this issue via a protest. Best way to solve this is petition the CRB with a "Rules Clarification".

As I know it 2014 and newer engines are subject to CRB review and could require submission of dyno sheets to the CRB should they request it. This all goes back to the whole restrictor threat from a few years ago and the thought that engine wars might begin with competitors coming in with big money and new high HP engines.

As most know the newer M/C engines are bound be multi axis gyros and sometimes throttle by wire. Either one of these posses major hurdles to get over. I'm not a code guy and can offer zero input there.

Much more HP added to the F1000 class and we WILL be faster than WELL DRIVEN Formula Atlantic cars. Most of the time now I out qualify the entire P1 field. Where this all stops nobody knows. From my viewpoint I believe it's limited to driver talent, these cars are crazy fast. We just got back from testing at Sonoma and my times now are in the 1:30 flat, the Pro Atlantic cars were only 2 sec quicker than this back in the mid 2000's.

i have heard that there is a west coaster with a late model engine with an early model fuel injection throttle body. This is clearly illegal imo.

[JRMarchand]

i have heard that there is a west coaster with a late model engine with an early model fuel injection throttle body. This is clearly illegal imo.

Hi Jay that's totally understandable, he had to find a way to make it run.

I was so happy to see this thread a few weeks ago and didn't want to attract much attention. So about 30 minutes ago I was able to finally after 3 years, I was able to get spark ,injector pulse and fuel pump under cranking, the engine is ready to fire up. I'm on a pre 2012 with all the factory harnesses, sensors, instrument cluster, yaw sensor, and magic key. So yes, it can be done, took me 3+ years and about 14grand in Buying engines, ecus, parts calibration software upgrade, the list is long but its all downhill from here. Im waiting on Devins to get back from runoffs, He is selling me a scan tool to run the adaptations for the three throttle sensors and will be able to view live data.
The engine is already in the car and that had a bunch to do with how long it took to convert. I'm using the NovaDimmer conversion, Jay I used 75% of the engine brackets for the k5 on the s1000. Cheers.

[Jnovak]

Hi Jay that's totally understandable, he had to find a way to make it run.

I was so happy to see this thread a few weeks ago and didn't want to attract much attention. So about 30 minutes ago I was able to finally after 3 years, I was able to get spark ,injector pulse and fuel pump under cranking, the engine is ready to fire up. I'm on a pre 2012 with all the factory harnesses, sensors, instrument cluster, yaw sensor, and magic key. So yes, it can be done, took me 3+ years and about 14grand in Buying engines, ecus, parts calibration software upgrade, the list is long but its all downhill from here. Im waiting on Devins to get back from runoffs, He is selling me a scan tool to run the adaptations for the three throttle sensors and will be able to view live data.
The engine is already in the car and that had a bunch to do with how long it took to convert. I'm using the NovaDimmer conversion, Jay I used 75% of the engine brackets for the k5 on the s1000. Cheers.

can't wait for you to hit the track. Let me know and i will crew. I cannot help with the engine stuff but chassis development that i can do.

[J Leonard]

i have heard that there is a west coaster with a late model engine with an early model fuel injection throttle body. This is clearly illegal imo.

You must be referring to a Kawasaki.

[JRMarchand]

Hi all, just a short video of the BMW engine on the dyno yesterday. It was the first, full throttle pull, it was nerve wrecking to say the least. only got to 12.8 rpm on the video, today i went to the rev limiter a few times, no problems above 14k. All i can say is that it is making all the power, no nannies cutting in, no shortness of breath, it's all there if you know what i mean. Hope it stays good long enough to play around Gingerman, not racing it for a year or so. Cheers
http://youtu.be/OnljvUtSui4

[JRMarchand]

Here is the pull to the rev limiter, no issues so far, not posting up numbers cause it's too early to tell, though a formidable gain over the previous K6 engine, leave to your imagination. I hope some racers want to swap into BMW, I would be glad to support. https://youtu.be/ZDSYr6jTTw8

[fieldsrb]

Won't be for an F1000, but have much more than a passing interest in putting this drivetrain into another road course application. Will likely be in touch...

[JRMarchand]

Won't be for an F1000, but have much more than a passing interest in putting this drivetrain into another road course application. Will likely be in touch...

Would be glad to support at any capacity. This one with all of the factory switches, harness, instrument cluster, key and other sensors and BMW Race ECU just took me too damn long to figure it out, now the has been FB is a great new frontier class, with open ecu and hopefully more layers of anti protrusion.
If I had to do it again, a standalone ecu is an option, although I gotta tell you, BMW did an excellent job with the egas or electric throttle body and how it let's the engine breath through the entire powerband. Standalone ecu is the way to go at the end of the day, I can't think of how many engines lost in FB due to lack of safeguards, oil pressure safeguards, water temp safeguards, afr safeguards. It wouldve saved a lot of engines , cars on fire.... ya know. Open ecu is a good thing, a good investment and a good way to insure your investment has protections and nannies for the engine.
Just my unprovoked two cents. )

[david oleary]

Would be glad to support at any capacity. This one with all of the factory switches, harness, instrument cluster, key and other sensors and BMW Race ECU just took me too damn long to figure it out, now the has been FB is a great new frontier class, with open ecu and hopefully more layers of anti protrusion.
If I had to do it again, a standalone ecu is an option, although I gotta tell you, BMW did an excellent job with the egas or electric throttle body and how it let's the engine breath through the entire powerband. Standalone ecu is the way to go at the end of the day, I can't think of how many engines lost in FB due to lack of safeguards, oil pressure safeguards, water temp safeguards, afr safeguards. It wouldve saved a lot of engines , cars on fire.... ya know. Open ecu is a good thing, a good investment and a good way to insure your investment has protections and nannies for the engine.
Just my unprovoked two cents. )

I would be interested also in switching to BMW from Suzuki, if you can email me some information on year, model, HP, torque
and pricing of a package with ECU doleary@rohrich.com

[JRMarchand]

I would be interested also in switching to BMW from Suzuki, if you can email me some information on year, model, HP, torque
and pricing of a package with ECU
doleary@rohrich.com

I can do that, it may take a week or so, I want to test the ignition cut on the paddle system with this car in the dyno. No sense in trying to sell something that can't shift unless you go geartronics and motec. I do support ProShift. I don't expect the Bosch BMW ecu to freak out when coils are interrupted. We'll see. I will email anyone interested, I'll get my thoughts together as well.
So i want to see and abuse the engine for a little while, for weeks, under load on the dynojet.

[JRMarchand]

BMWs1000rr VD on the dyno, ProShift system working really well. BMW ecu is happy. Glad I'm recording stuff, that header is glowing like mad. Enjoy
https://youtu.be/ffTdrZz25sk

[99sh]

Considering the new forthcoming rules for FB/FA any final updates? [emoji16]

Were you able to get a dyno plot?