PDA

View Full Version : Why no DSP for LML?



GMPX
October 12th, 2016, 02:09 PM
As most people in the Diesel tuning industry know, EFILive offers on the fly switchable tuning for many different ECM types from 2001 to current models, but the one we get asked a lot about is the E86A and E86B LML ECM, so I've decided to make this a sticky thread so anyone can find the truth behind it and why we can't offer DSP5 for these ECM's.

For any switchable tuning to work EFILive needs to modify the original Operating code of the ECM to perform the switching functionality, on the E86A & E86B (aka Bosch EDC17) the operating system code is protected with a Digital Signature. The correct digital signature for the OS data can only be generated by Bosch as they hold the private keys.

Although the OS data in the ECM is not encrypted (yet) the contents are signed to produce a unique 'digest' that is validated when the ECM is flashed via the OBD-II port.
Like the image below shows, the 'Input' is some text, the 'Hash function' alters the text to create a unique 'Digest'.

Picture from : https://en.wikipedia.org/wiki/Public-key_cryptography
https://upload.wikimedia.org/wikipedia/commons/thumb/2/2b/Cryptographic_Hash_Function.svg/740px-Cryptographic_Hash_Function.svg.png

When flashing the ECM after the OS section(s) have completed the ECM will verify the data written generates the correct Digest that Bosch would have created for that section of code. And that right there is where we hit a wall, we cannot generate the correct numbers so the ECM will accept the modified data as valid, instead it knows the data is modified and will not complete the flash process.
If you look at the image above you can see that it is impossible to take the data in the pink boxes and generate the text in Blue when we have no idea how Bosch created the Hash function (I believe they use a 1024 bit Key), but that is what we would need to be able to do for the ECM to accept any modifications to the OS code.

Sadly there is no way to switch off this security check using any tricks via the OBD-II port, but this is the only reason we've never pursued DSP5 for the LML engine. Sometimes you just can't beat technology and for what it is worth there is no way to brute force figure the digests out either.

Cheers,
Ross

30 RUM
November 17th, 2016, 05:29 AM
I know you guys do not obsess over the competition, but how do you think EZ Link is doing it? Is there anyway we can change run modes via switch, kinda a built in DSP?

Thanks

Sam

rcr1978
November 17th, 2016, 01:06 PM
I would be more interested in more tables in the current software then focus on DSP

GMPX
November 17th, 2016, 02:43 PM
I know you guys do not obsess over the competition, but how do you think EZ Link is doing it? Is there anyway we can change run modes via switch, kinda a built in DSP?
I don't know exactly but I was told it is trickery to make it run in some different modes the ECM already has in there. Definitely not modified OS code like the traditional DSP & CSP OS's are. Ben has done a good job on figuring that all out.


I would be more interested in more tables in the current software then focus on DSP
Says every user for every controller :grin:
We have no "immediate" plans to add tables to the LML, new Duramax ECM is about to drop and we have been working on some new controllers for a while now. Hopefully they will be released early next year (no, not going to say what they are).

rcr1978
November 17th, 2016, 03:16 PM
Yeah I know there is lots of new ECM's hitting the market and the name of the game is to keep all us guys :w00t: with support of the new stuff coming out. Was not to long ago my hopes of tuning a LML was very slim, then you guys came through after saying no way, keep up the good work you guys are awesome :rockon:!

Road
November 17th, 2016, 11:29 PM
I know you guys do not obsess over the competition, but how do you think EZ Link is doing it? Is there anyway we can change run modes via switch, kinda a built in DSP?

Thanks

Sam

Limit the fuel quantity by varying the resistance of a sensor.

Snipesy
November 18th, 2016, 05:28 AM
I looked into something similiar and it's really not worth doing. It won't change how the truck behaves at all, you'll just get well... A lower fuel ceiling.

Messing with the barometer would allow you to do fancier things, but then you can't adapt boost for altitude. If you care that is.

It other words, it's really nothing efi live can't do already.

DURAtotheMAX
November 18th, 2016, 06:19 AM
I looked into something similiar and it's really not worth doing. It won't change how the truck behaves at all, you'll just get well... A lower fuel ceiling.

Messing with the barometer would allow you to do fancier things, but then you can't adapt boost for altitude. If you care that is. The way I and I guess Ben was going for was to instead have it recieve the signal over CAN. I guess he got it to work, certainly never told me.

It other words, it's really nothing efi live can't do already.

PM'd you.

30 RUM
November 18th, 2016, 03:14 PM
I would venture that is what's happening. I did notice they are pining the switch to the ecm. I haven't got my hands on one yet though.

GMC-2002-Dmax
November 18th, 2016, 03:37 PM
About 10 years ago ROSS wrote an LB7 os with VFO, variable fueling option, it worked pretty good at the time and used a Rheostat.

Anyway, BEN had a cool deal first, but it never made it to market.

As the saying goes, you snooze, you lose !!!

If there is demand, someone will fill it, some way, somehow.

I am ordering a bunch of units, it's not going away, it's only going to grow, so you either beat them or join them........and that is just how it is.

I'll be testing on my LML soon enough.

GMPX
November 20th, 2016, 10:21 AM
As the saying goes, you snooze, you lose !!!

If there is demand, someone will fill it, some way, somehow.

I am ordering a bunch of units, it's not going away, it's only going to grow, so you either beat them or join them........and that is just how it is.

I'll be testing on my LML soon enough.
Tony maybe I am reading in to what you wrote the wrong way as some sort of stab at EFILive for not doing something like this ourselves? After seeing how the OS is protected on these ECM's we never pursued the idea of developing a true DSP OS for the LML, so the only thing we've been 'beaten' by is Bosch.
What Ben has done is a great solution given the restraints imposed by the hardware design, there is no angst, jealously or envy going on here. As long as Ben is getting a cut from the product that I assume is now selling his idea then good on him.

GMC-2002-Dmax
November 20th, 2016, 04:13 PM
It's not a stab at you or EFI or Ben, it's just an observation, BEN figured it out but could not get it to market, you guys did not pursue it for the reason you stated but the fact is someone got it to work.

Its nothing more than that, someone decided they were going to develop it and make it work, and the demand for it is driving sales of it to them.

Its not the first time that this happened, it happens all the time in business, someone had an idea to build a computer once upon a time, now we have Apple and MICROSOFT.

I was just commenting on the fact that the idea you guys have done so well with , DSP5 and CSP drove a lot of sales, but not pursuing the LML due to not being able to write a custom OS was why you never did the switchable tuning.

On edit: I don't think Ben has any skin in that process, in fact he was going to send out his beta setup for me to test but never did, to my knowledge this is all EZ LYNK, but again I do not have all the facts yet, but Ben used the + & - tap shifter to switch, this system uses a rotary switch.

GMPX
November 20th, 2016, 05:28 PM
Tony, I had some PM's with Ben about his LML mod months ago. At the time he needed to be able to flash a modified Slave OS section via the OBD-II port (maybe that has changed, I don't know). I said it was not possible for the reasons explained in my opening post and appologised we couldn't work with him on it. I also said that requiring users to pull apart their ECM and flash a modified OS using something like Alientech's KTAG tool to get an EFILive DSP OS in the ECM made absolutely no sense from a business perspective for us which is also why we chose not to develop a true DSP OS ourselves.

Yes DSP and CSP have been very good for our sales, it is what people want and I enjoy writing them too which makes it a win all around. The E98 2.8L DSP came out long after the LML ECM was released but we knew we could do them which is why it happened, there was no technical brick wall like the LML so we 'get' that people want it.

If you've 'ordered a bunch of units' then surely you understand how it works to invest so much? Is it really switching in multiple commanded fuel, injection time, driver demand, rail pressure and timing maps like you think it must be? The very things you tuners know needs to be done for it to work right, or is it just fooling a sensor with tuning to match? I don't know, I'm happy to be educated on it though.

GMC-2002-Dmax
November 20th, 2016, 11:32 PM
Unfortunately it makes no difference how it works in the end, it apparently does work good enough to make the customer happy, and just as I pointed out, we have an option that is driving sales.

I can either ignore it and suffer sales losses to my peers and competitors, or begin to support it and capture some of those sales, its strictly a business decision and not a personal one.

In the grand scheme of things you know that the demand drives what you do as a business owner, not what you want to do necessarily, but what your customers want.

IdahoRob
November 21st, 2016, 07:56 AM
I think there is some confusion. EZ Link is not Ben, EZ Link is a cloud based tuning product that has switchable LML tuning.

As Tony mentioned, it's what the customers are asking for. Like the LML, a competitor started tuning them and customers flocked to them because there wasn't a good competitive product. Then EFI Live decided to market tuning for the LML, seemed like a simple business decision.

I see the EZ Link becoming a force, especially if they do not have competition that offers switchable tuning. I'm not discrediting anything anyone has done or pointing fingers, just stating what is happening in the trenches by listening to the customer's feedback. Just information from a company that listen's to many inquiries a day on this subject.

GMPX
November 21st, 2016, 09:35 AM
Far out guys, I don't know what to say.

GMC-2002-Dmax
November 22nd, 2016, 07:22 AM
Far out guys, I don't know what to say.

Say you might revisit it........;)

S Phinney
November 23rd, 2016, 05:52 AM
I don't know enough about how you guys at EFI Live run things internally but, with that said why wouldn't you have different divisions of the company that keeps developing the released items and another that works on all the future items that's coming out? Having continued support and development if the current controllers is very important to the guys who tune. As we learn more and more about what and how things work and what could be beneficial for us all including you guys to just makes good business sense for that to happen. You guys are almost in a class by yourself but not quite. These other guys develop or try to for the things that people want and can't get through EFI Live. I appreciate what you do and would really like to see a more open minded approach to satisfying the customer demands.


Sent from my iPhone using Tapatalk

GMPX
November 23rd, 2016, 09:49 AM
I don't know enough about how you guys at EFI Live run things internally but, with that said why wouldn't you have different divisions of the company....

It is a two man show on the software side of things, no divisions to delegate resources too (did you realise there was only two of us doing everything in development ?)

S Phinney
November 23rd, 2016, 09:59 AM
No sir I didn't realize that. That's us why I stated I didn't know how the inner workings was there. I do applaud you for what you do. I like many others do feel that the existing items are in need of refining. If I were smart enough and had the time and monies I would gladly join you in helping accomplish that.

Admin Edit:
I moved posts that were not related to DSP for LML to a new thread here:
https://forum.efilive.com/showthread.php?27232-LML-wishlist

catman3126
December 12th, 2016, 04:45 AM
SO it looks like we will get DSP tuning after all? I see D tuner released an update on this today?

DURAtotheMAX
December 12th, 2016, 04:55 AM
SO it looks like we will get DSP tuning after all? I see D tuner released an update on this today?

:mrgreen: :rockon::coool:

GMPX
December 12th, 2016, 09:05 AM
Let me just nip this one in the bud before the rumours start. EFILive had no idea this was being worked on for use in our software (well, on our hardware), we had nothing to do with it at all. Looks like they cover both E86 ECM types which is great news for everyone out there.

Road
December 12th, 2016, 09:37 AM
Who is D Tuner?

Mitco39
December 12th, 2016, 09:49 AM
Duramax Tuner - Nick.

Road
December 12th, 2016, 10:11 AM
Duramax Tuner - Nick.
That's what I thought. Don't see anything about it on his website.

catman3126
December 12th, 2016, 12:25 PM
Let me just nip this one in the bud before the rumours start. EFILive had no idea this was being worked on for use in our software (well, on our hardware), we had nothing to do with it at all. Looks like they cover both E86 ECM types which is great news for everyone out there.

Oh ok, wel when they post "w efi live" you can see how that can imply you guys had something to do with it.

GMPX
December 12th, 2016, 01:07 PM
Oh ok, wel when they post "w efi live" you can see how that can imply you guys had something to do with it.

Do people even read what we post? I try to stop the rumours and it still isn't good enough, go ahead and ask Nick if we developed it.

SASDakota
December 13th, 2016, 07:53 AM
Oh ok, wel when they post "w efi live" you can see how that can imply you guys had something to do with it.

I didn't take it that way at all. It is very clear that they did some work on their base files and are simply providing it using the efilive platform.

PPEI - Kory was asking on facebook whether he should port his version into efilive and lightly explained why he had not up until this point.

There is so much going on in the market, you really need to pay attention to the details.

GMPX
December 13th, 2016, 09:18 AM
That is correct SASDakota, PPEI decided to put their LML switching tune exclusively on EZ Lynx. Once it was released with installation instructions it became clear to some people how the system may have worked so others have now done it too. The key difference is that Nick decided to stick with EFILive when it came to distributing the tunes and now it looks like PPEI will also port their EZ tunes on to EFILive as well.

Mitco39
December 13th, 2016, 10:53 AM
I am just excited to see our industry moving forward in the face of increasing security and the threat of it putting a hard stop to what many of us have been doing for years :)

Lets just hope (fingers crossed) that this will continue with the L5P!

Dmaxink
December 13th, 2016, 11:28 AM
That is correct SASDakota, PPEI decided to put their LML switching tune exclusively on EZ Lynx. Once it was released with installation instructions it became clear to some people how the system may have worked so others have now done it too. The key difference is that Nick decided to stick with EFILive when it came to distributing the tunes and now it looks like PPEI will also port their EZ tunes on to EFILive as well.

Ah thats the easy way! I decided to go full force and just CAXd in all the changes. Whew.. ready to go home.

DemonDMAX
December 14th, 2016, 04:49 AM
No legal battles fought or being fought at all.. and if it came to that - I wouldn't say a word. The truth of the matter is - it simply looked very bad seeing KrazyOn and then Whitehorse having the same tune levels/descriptions especially after the convo I had with them. Mike sounds like a solid stand up fella, KrazyOn just by their piss poor wording and assumptions of ezlynk I have no respect for.. especially since we made a post about it and people come up stating he said we were getting shut down, don't order, OEM is going after tuners and other silly things. Anyways, this isn't the place for it... but everyone be cautious and always on the lookout! I have been stolen from more than I can count and get plenty royalties these days from catching folks.

I won't discuss anymore here, thanks guys.

I did not say you were getting shut down. I simply shared public document found here.
https://mega.nz/#!R4FDARAA!hsyhZ3eHGRQY1GcwTytwnJ5Q2D9bb1JJ0ou3JnN AOCs

I don't know why you continue to slander my company name. I have not said anything negative about yours.
I do sell others products. I also have a large dealer network. Not only are you defamating my company and causing us to lose sales.
Your also doing it to our dealer network.

"It didn't offend me, just showed your intelligence" - Kory Willis

GMPX
December 14th, 2016, 08:32 AM
If you guys want me to close this thread and remove what probably should be removed just say the word.

Snipesy
December 14th, 2016, 09:18 AM
I vote yes. Delete the whole thread.

I think IP theft is a big issue and should be discussed. But not with cannon fire. The Efi live forums shouldn't be a place where everyone is just denouncing eachother.

GMPX
December 14th, 2016, 09:23 AM
Ok, by request I have hidden the posts that went off topic, sorry guys I don't like censoring things but it was getting a little bit emotional between you all. In the end the arguing was not even to do with EFILive products directly.
If anyone involved in the discussion that had been hidden is not happy about this then send me a PM.

SASDakota
December 14th, 2016, 09:33 AM
Even though I appreciate your opposition to censoring, that was the right move.

Dmaxink
December 14th, 2016, 10:48 AM
Ok, by request I have hidden the posts that went off topic, sorry guys I don't like censoring things but it was getting a little bit emotional between you all. In the end the arguing was not even to do with EFILive products directly.
If anyone involved in the discussion that had been hidden is not happy about this then send me a PM.

I actually thought we were in the private area.. Sorry. :-)

DemonDMAX
December 14th, 2016, 12:15 PM
Ok, by request I have hidden the posts that went off topic, sorry guys I don't like censoring things but it was getting a little bit emotional between you all. In the end the arguing was not even to do with EFILive products directly.
If anyone involved in the discussion that had been hidden is not happy about this then send me a PM.

how do I pm?

GMPX
December 14th, 2016, 01:20 PM
Your post count is probably too low to send PM's, but I don't really want to be drawn in to this to be honest as the topic was more in line with issues relating to another tuning product totally unrelated to EFILive.
DemonDMAX I have hidden any comments that were made mentioning you.

dirtygerman
February 14th, 2017, 12:35 PM
I did not say you were getting shut down. I simply shared public document found here.
https://mega.nz/#!R4FDARAA!hsyhZ3eHGRQY1GcwTytwnJ5Q2D9bb1JJ0ou3JnN AOCs

I don't know why you continue to slander my company name. I have not said anything negative about yours.
I do sell others products. I also have a large dealer network. Not only are you defamating my company and causing us to lose sales.
Your also doing it to our dealer network.

"It didn't offend me, just showed your intelligence" - Kory Willis


Maybe I misunderstood this but 1&0's are hard to keep people from stealing or copy and pasting just pointing that out.

GMPX
February 14th, 2017, 03:22 PM
Maybe I misunderstood this but 1&0's are hard to keep people from stealing or copy and pasting just pointing that out.
Well that is true, but the same logic might be applied to mp3's and movie files, they are only 1's and 0's too.

dpc
February 27th, 2017, 02:50 PM
Just to clarify LML switchable tuning (OS Upgrade) is not available to the public? I've been watching this thread for awhile hoping I would be able to do my own LML switchable tuning. Can one of you tuners send me your CAX file so I can do my own tuning...:mrgreen:

GMPX
February 27th, 2017, 03:07 PM
Just to clarify LML switchable tuning (OS Upgrade) is not available to the public?
There is no such thing as an 'OS Upgrade' for switchable tuning on LML, it is just manipulation of a limiter table based on a switch input. We allow you to edit the table used but I don't know what else sits in the background to make it work.

dpc
March 18th, 2017, 04:55 AM
There is no such thing as an 'OS Upgrade' for switchable tuning on LML, it is just manipulation of a limiter table based on a switch input. We allow you to edit the table used but I don't know what else sits in the background to make it work.

Which limiter table is being used? Thanks.

DURAtotheMAX
March 20th, 2017, 08:23 AM
which limiter table is being used? Thanks.

b0457

dpc
March 21st, 2017, 01:57 PM
b0457

Fuel Temp reduction factor? What equals tune one, two, three, four, five? Thanks.

GMPX
March 21st, 2017, 11:19 PM
Well it isn't a true switchable tune, you are simply applying a limiter based on a false sensor input reading (now your switch).
There really is no Tune 1, 2 etc like in the past which is why we certainly don't refer to this trick as 'DSP', because it really isn't.

Road
March 21st, 2017, 11:34 PM
Fuel Temp reduction factor? What equals tune one, two, three, four, five? Thanks.
There is a lot more than just changing one table to make switch control the fuel rate. You don't have to use fuel temp sensor either you can use IAT, EGR temp, etc any sensor that ecm needs less would be best. You will need to edit several axis and disable some bits and then turn off diagnosis for the said sensor. Then you will need to have a switch that is capable of controlling the resistance and no one switch works with every method. So that being said there is many ways to control fuel rate and a working setup takes a lot of time to make it work and be reliable. There is way more involve than most know and requires little more than a cax file that's why you only see the best of the best offering it and very seriously doubt they will share their hard work.

GMPX
March 22nd, 2017, 02:08 PM
100% spot on Mike, we don't know how it is done from start to end, it was not our idea to mimic or claim to be our own but we already had the axis points in the software (just hidden from user modification), that since got opened up for editing. If there is some other 'hidden' things to make it work then I can't tell anyone what they are because I don't know. And those that made their own .cax files I would expect wouldn't make them public.

dpc
March 22nd, 2017, 02:09 PM
There is a lot more than just changing one table to make switch control the fuel rate. You don't have to use fuel temp sensor either you can use IAT, EGR temp, etc any sensor that ecm needs less would be best. You will need to edit several axis and disable some bits and then turn off diagnosis for the said sensor. Then you will need to have a switch that is capable of controlling the resistance and no one switch works with every method. So that being said there is many ways to control fuel rate and a working setup takes a lot of time to make it work and be reliable. There is way more involve than most know and requires little more than a cax file that's why you only see the best of the best offering it and very seriously doubt they will share their hard work.

I did some more searching on the B0457 table and it sounds like that is that table being used. We can currently buy the switch with instructions on how to pin into the ECM. I'm just not sure of the axis within that table for each "tune". I now understand that it is simply a limiter and not an actual DSP tune.

I did ride in an 2016 LML with the switch and a well know tuner's file. It does feel like it has 5 different tunes in it, not quite like my 03 LB7 or 06 cummins (both have DSP and CSP switches) but if you didn't know any better you would think it had a DSP5 switch.

DURAtotheMAX
March 23rd, 2017, 12:31 AM
I did some more searching on the B0457 table and it sounds like that is that table being used. We can currently buy the switch with instructions on how to pin into the ECM. I'm just not sure of the axis within that table for each "tune". I now understand that it is simply a limiter and not an actual DSP tune..

Thats what you'll have to figure out, trial and error....unless you can find someone to tell you.



I did ride in an 2016 LML with the switch and a well know tuner's file. It does feel like it has 5 different tunes in it, not quite like my 03 LB7 or 06 cummins

Thats because its only pseudo "tune switching". It's somewhat "close enough" to give most people the impression that its "switching tunes"....but for more discerning people who know better, they can tell that its just modifying max allowed mm3.

The version of "SOTF/DSP" that I made used CAN messages and actually changed/commanded which pedal to torque and fuel maps the ECM used...so it was actual mode switching, albeit near as expansive and fine-tunable as true EFILive DSP. :)

But my method was much much more complex to implement, thats why I never was able to fully release it besides the dozen or so trucks running around with it. Its too bad I ran into mass-production snags though, because I did like how it was all controlled via the driver info center and factory tapshift buttons.

Ben

DURAtotheMAX
March 23rd, 2017, 01:00 AM
Also, on the fuel temp sensor SOTF...

There are several other changes required that are outside of what EFILive has mapped in their tuning tool.....

1. You have to enable "fuel temp sensor 2" on the 2012+ OS's. Because only 2011 trucks had two separate fuel temp sensors from the factory. 2012+ it was unused.

2. You have to change the scaling of the measured-voltage-to-temperature table for the fuel temp sensor #2. Basically what B0212-B0213 does in an 01-05 LB7/LLY tune file. Note I said 01-05 LLY. You wont find B0212-0213 in an LML tune because EFILive does not have it mapped.

3. You have to disable range-extension of the fuel temp sensor. The ECM has a forward-biased pullup resistor inside it with a little FET in there that can change the value of the pullup resistor on the ADC input of the fuel temp sensor circuits. Obviously all ADC inputs to ECU's have to have either a somewhat high-impedance pull up resistor to +5volt reference, or a pulldown resistor to low-reference. Otherwise the ADC input would float erratically and not give accurate readings. Its also buffered with an op-amp in voltage-follower configuration...but I digress.

Anyway, the IAT and ECT sensor inputs have this feature too. Its to give the sensor more accuracy throughout its range of -40* to 300* or whatever. Think of it like high range vs low range in a transfer case. Around 130* or so, the ECM triggers the FET to change the pullup resistor value so now instead of the thermistor referencing a 10k ohm bias to +5v, its using say 5.6k pullup value.

If you dont disable the range switching, when you change tunes to the point of the ECM reading around 130*, it'll switch that internal FET and the readings from your DSP switch will then get all screwed up and it'll jump from like 130* to 200* or something. Then your B0457 table becomes all wrong again.

Or I guess you could also just carefully tweak sensor scaling range 1 table and sensor scaling range 2 table independently so even when it switches ranges, the measured temp value is overlaid/corresponds properly despite the shift in pullup bias....but thats probably more of a pain.


The rest of the stuff you need is right in EFILive. B0457, and the axis modifier tables....because B0457 only goes from like 220-260* or whatever, you need to broaden the range to give a greater difference in the fuel limiting factor value.

2015-2016, there are a couple extra things you have to do....you cant directly copy-pasta all changes from 2011-2014 OS to 2015-2016.

Its not rocket surgery, I messed around with this "DSP" method like a year and a half ago and it worked alright...but I didnt know as much about working in the LML binaries back then because I didnt own an LML until october 2015, so Id never worked on them before....so I wasnt able to figure out how to disable the sensor range-extension feature....so I had to make a complex DSP switch with some extra logic in there to compensate and change its outputs at the same time the ECM changed the pullup value. If I had all of the LML documentation and A2L, it would have been easy-peasy. The concept is simple as hell.

But I abandoned the EGR temp/fuel temp sensor based "DSP" because I never liked the idea of sensor-fooling just because its not true mode switching, I think its kinda ghetto/hack...but thats just my own stupid opinion. So I pursued my CAN-based method that actually commands the ECM to change modes and put all my eggs in one basket with that. Yeah, it was probably dumb of me to do that, should have kept the EGR/fuel temp based "DSP" and just said "screw it, good enough, uses existing DSP switches, inexpensive, works alright"....but whatever, thats because Im a true out-of-the-box tinkerer/engineering-design nerd and NOT a business man. :)

Ben

Snipesy
March 23rd, 2017, 08:06 AM
So on CAN...

If there is a way to control the mode it would be done with the quote unquote 'Device Control'. Which for the LML's ECM would be a command that looks like 7E0 08 AE aa bb bb bb bb bb, where AE is the mode, aa is the the control mode, and the bbs are the data for that control.. Also the length is always going to be 8 bytes, because Bosch. You know it did something if you didn't get a 7F.

Additionally you will need to send this to to the ECM about once every .5 seconds or so, else nothing will work. 7E0 01 3E(just the ecm), or 101 FE 01 3E (everything).

Also sending 7E0 02 10 02 or 101 FE 02 10 02 after you start sending the 3E would be a good idea as well. When you send the 3E signal trouble codes get disabled among a bunch of things, the 10 signal will undo most of that.

I doubt the ECM listens for a mode change on anything other than 7E0 08 AE blah blah blah. Wouldn't make sense really. They use this same system in the OEM tuning of the engine.

DURAtotheMAX
March 23rd, 2017, 08:10 AM
So on CAN...

If there is a way to control the mode it would be done with the quote unquote 'Device Control'. Which for the LML's ECM would be a command that looks like 7E0 08 AE aa bb bb bb bb bb, where AE is the mode, aa is the the control mode, and the bbs are the data for that control.. Also the length is always going to be 8 bytes, because Bosch. You know it did something if you didn't get a 7F.

Additionally you will need to send this to to the ECM about once every .5 seconds or so, else nothing will work. 7E0 01 3E(just the ecm), or 101 FE 01 3E (everything).

Also sending 7E0 02 10 02 or 101 FE 02 10 02 after you start sending the 3E would be a good idea as well. When you send the 3E signal trouble codes get disabled among a bunch of things, the 10 signal will undo most of that.

I doubt the ECM listens for a mode change on anything other than 7E0 08 AE blah blah blah. Wouldn't make sense really. They use this same system in the OEM tuning of the engine.

My method does not use $AE device-control messages.

DURAtotheMAX
March 23rd, 2017, 08:13 AM
Snipesy the stuff you are referring to is basic OBD scan tool protocol, the $AE, $7E0, rejected request, $01 #3E, etc. Its not anything special. Every vehicle uses that stuff for UUDT/USDT scan tool communication.

And the fact that the messages are 8 bytes is "because CAN".....not "because Bosch".....

Snipesy
March 23rd, 2017, 08:51 AM
Snipesy the stuff you are referring to is basic OBD scan tool protocol, the $AE, $7E0, rejected request, $01 #3E, etc. Its not anything special. Every vehicle uses that stuff for UUDT/USDT scan tool communication.

And the fact that the messages are 8 bytes is "because CAN".....not "because Bosch".....

Hey Im trying to help. It is special GM specific nonsense. Yes it is the GM 'standard scan tool protocol' but is that not what you want?

If you want more you'd need dedicated boxes to repeat other modules. Let the ECM operate in your own world. Which hey, if you can do that lemme just say there are many other fields which could use you.

DURAtotheMAX
March 23rd, 2017, 09:28 AM
Hey Im trying to help. It is special GM specific nonsense. Yes it is the GM 'standard scan tool protocol' but is that not what you want?

If you want more you'd need dedicated boxes to repeat other modules. Let the ECM operate in your own world. Which hey, if you can do that lemme just say there are many other fields which could use you.

Im not sure what you're getting at? Kinda confused, sorry haha.

I am not sending the ECM "scan tool commands" to make it change modes, its something else that I figured out.

Believe me, I wish there was a simple scan tool command ($AE) that would make it change modes! Maybe there is, I dont know...maybe you could command a service regen and that would put the ECM in a different mode or something, and you could tune the service regen tables? No idea....

Snipesy
March 23rd, 2017, 09:47 AM
Believe me, I wish there was a simple scan tool command ($AE) that would make it change modes! Maybe there is, I dont know...maybe you could command a service regen and that would put the ECM in a different mode or something, and you could tune the service regen tables? No idea....

Tried that. But when you 'disable' the dpf, it also disables the the AE command. Which is interesting. There's alot of other issues with that though. The ECM won't even do it unless the vehicle is stopped anyway.

I am certain an AE command exists, but Its probably locked behind a dev switch which... All I know is once that switch is disabled it is pretty much impossible to reset it. Not with the equipment I have anyway. Maybe there is a way to edit the tune to make that not happen, but again haven't found anything like that.

GMPX
March 23rd, 2017, 09:54 AM
It is a shame the ECM's OS is as protected as it is (thought it is worse on newer controllers), Ben's concept was very good but would have required each ECM to be opened up and have the modified code (OS) JTAG's in because nobody but Bosch has the signing keys for the OS to be flashed via OBD-II :yucky:

dpc
March 23rd, 2017, 01:30 PM
Thanks for the help, I'll report back when/if I get it working. Didn't realize you were Ben, The standalone harness you built me for my race truck still works great. :grin:


Also, on the fuel temp sensor SOTF...

There are several other changes required that are outside of what EFILive has mapped in their tuning tool.....

1. You have to enable "fuel temp sensor 2" on the 2012+ OS's. Because only 2011 trucks had two separate fuel temp sensors from the factory. 2012+ it was unused.

2. You have to change the scaling of the measured-voltage-to-temperature table for the fuel temp sensor #2. Basically what B0212-B0213 does in an 01-05 LB7/LLY tune file. Note I said 01-05 LLY. You wont find B0212-0213 in an LML tune because EFILive does not have it mapped.

3. You have to disable range-extension of the fuel temp sensor. The ECM has a forward-biased pullup resistor inside it with a little FET in there that can change the value of the pullup resistor on the ADC input of the fuel temp sensor circuits. Obviously all ADC inputs to ECU's have to have either a somewhat high-impedance pull up resistor to +5volt reference, or a pulldown resistor to low-reference. Otherwise the ADC input would float erratically and not give accurate readings. Its also buffered with an op-amp in voltage-follower configuration...but I digress.

Anyway, the IAT and ECT sensor inputs have this feature too. Its to give the sensor more accuracy throughout its range of -40* to 300* or whatever. Think of it like high range vs low range in a transfer case. Around 130* or so, the ECM triggers the FET to change the pullup resistor value so now instead of the thermistor referencing a 10k ohm bias to +5v, its using say 5.6k pullup value.

If you dont disable the range switching, when you change tunes to the point of the ECM reading around 130*, it'll switch that internal FET and the readings from your DSP switch will then get all screwed up and it'll jump from like 130* to 200* or something. Then your B0457 table becomes all wrong again.

Or I guess you could also just carefully tweak sensor scaling range 1 table and sensor scaling range 2 table independently so even when it switches ranges, the measured temp value is overlaid/corresponds properly despite the shift in pullup bias....but thats probably more of a pain.


The rest of the stuff you need is right in EFILive. B0457, and the axis modifier tables....because B0457 only goes from like 220-260* or whatever, you need to broaden the range to give a greater difference in the fuel limiting factor value.

2015-2016, there are a couple extra things you have to do....you cant directly copy-pasta all changes from 2011-2014 OS to 2015-2016.

Its not rocket surgery, I messed around with this "DSP" method like a year and a half ago and it worked alright...but I didnt know as much about working in the LML binaries back then because I didnt own an LML until october 2015, so Id never worked on them before....so I wasnt able to figure out how to disable the sensor range-extension feature....so I had to make a complex DSP switch with some extra logic in there to compensate and change its outputs at the same time the ECM changed the pullup value. If I had all of the LML documentation and A2L, it would have been easy-peasy. The concept is simple as hell.

But I abandoned the EGR temp/fuel temp sensor based "DSP" because I never liked the idea of sensor-fooling just because its not true mode switching, I think its kinda ghetto/hack...but thats just my own stupid opinion. So I pursued my CAN-based method that actually commands the ECM to change modes and put all my eggs in one basket with that. Yeah, it was probably dumb of me to do that, should have kept the EGR/fuel temp based "DSP" and just said "screw it, good enough, uses existing DSP switches, inexpensive, works alright"....but whatever, thats because Im a true out-of-the-box tinkerer/engineering-design nerd and NOT a business man. :)

Ben

golzee
February 22nd, 2020, 12:18 PM
any way to do the fuel temp or egr temp method with out going into cax and all that. id like to do it on my 2012 but after playing around with winols for awhile I give up on that lol.