|
adam-
|
 |
« Reply #45 on: January 26, 2017, 02:06:54 AM »
|
|
|
Spiel.
Really helpful and pretty interesting. |
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
 Offline
Posts: 1056
|
|
« Reply #46 on: January 26, 2017, 06:57:36 AM »
|
|
|
At the moment, not too keen... have too much stuff to do already. Mid-february should be better.
It's also expensive $$$ wise. Because these custom code jobs take obscene amounts of time. I am not talking about the first time when it can take a week to do it (and it never pays off), but even the subsequent modifications where you have most of the code and have to slightly mod/recompile it for another binary.
Because I do this for work, and not as a hobby after work, for me time = money. It is very likely I can remap 2-3 different cars in the time it takes me to do those code mods for you...
With those prices the people who buy this stuff from me, are those who will tune a bunch of these cars in the future, not just their own car, as it's just not really economically feasible otherwise.
Btw as for the topic at hand - to sort the boost fueling you don't need to do anything special. Actually NA ME7 does not care about boost in any special way when it comes to basic timing and fueling.
The issues are with torque control and target load. Actual load it has no problem to calculate because it comes from MAF.
If you want to just drive in a straight line, flip some APPL bits. Make the throttle opening be fixed to the pedal based on RPM.
Kill all torque management. Basically make the ECU like a cable throttle ECU that can only react to what is happening, and not control the airpath.
This of course will not work with an automatic gearbox, but with manual gearbox it will work just fine. Of course you will lose all "smoothness" of the car and it will drive more like a traditional cable throttle car.
But if you only use the car for weekend fun car there is really nothing wrong with that. It will also allow you to bypass all of the load control shenanigans and focus on tuning fuel, spark and camshaft overlap.
From the torque failsafes keep only the ones that are pedal % vs throttle %. So if your throttle for some reason jams open the car will still cut out and not drive you through a wall.
Traction control obviously won't work either.
This all makes a lot of sense (obviously)... but one has to think that at this point if it were me I would make a PnP stand alone and fit that to the car and just keep the stock ECU to swap in for emissions testing (if that is needed where you are). As neutering the stock ECU like that, plus lack of boost control, protections, etc. to me it's not worth it at all. As you said, there are VERY few people that know what to do to get an NA ECU working right under boost.. I've driven and looked at some commercial tuner's BT R32 file because after driving it I couldn't believe how horrible the car drove. It was mind blowing that people were paying big money for these tunes and just dealing with that kinda stuff. |
|
|
|
|
Logged
|
|
|
|
|
prj
|
|
« Reply #47 on: January 27, 2017, 10:51:31 AM »
|
|
|
Not every standalone can drive the clocks and other canbus stuff in the car.
But yes, it is much easier to fit one and map that instead.
What I suggested is far from ideal, but for someone who is new to mapping in general, it seems better to spend more time actually tuning the engine and less time making the ECU do something it was not designed to.
|
|
|
|
|
Logged
|
|
|
|
NOTORIOUS VR
Administrator
Hero Member
Karma: +58/-7
Offline
Posts: 1056
|
|
« Reply #48 on: January 27, 2017, 11:32:20 AM »
|
|
|
Not every standalone can drive the clocks and other canbus stuff in the car.
But yes, it is much easier to fit one and map that instead.
What I suggested is far from ideal, but for someone who is new to mapping in general, it seems better to spend more time actually tuning the engine and less time making the ECU do something it was not designed to.
Unfortunately you're right... having a fully functioning car these days w/ CAN is a lot of work but as you say making an ECU do something it was never designed to do is what you're left with. Generally, one has to be realistic with their goals as well IMO.. These days with the information on the internet its too easy to get lost in numbers and what other people do. Plus rarely do you get the whole story just part of it and possibly the end result at the time shown to you and many forget that it probably took a lot of trial and error, sweat and tears and beer to get there. But really, starting with less then ideal circumstances like what's outlined in this thread means you need to really have a handle on what you're going to require/expect at the end and contract the jobs to the right people to make it happen (Motronic or Stand alone). |
|
|
|
|
Logged
|
|
|
|
c4andmore
Newbie
Karma: +1/-2
Offline
Posts: 11
|
|
« Reply #49 on: January 27, 2017, 10:12:43 PM »
|
|
|
Not every standalone can drive the clocks and other canbus stuff in the car.
But yes, it is much easier to fit one and map that instead.
What I suggested is far from ideal, but for someone who is new to mapping in general, it seems better to spend more time actually tuning the engine and less time making the ECU do something it was not designed to.
IME most of the stand alone systems are intended to provide quick and easy AFR/ Ign. control unlike ME7 which is designed to allow OEMs and others who are properly trained to write code that can control every aspect of engine operation including emissions, regardless of the SI engine design and operation short of DI engines which would use later Bosch MEx systems. The ME7 system was meant to be everything to everyone (at the time of it's introduction in the late 90's), wanting a sophisticate EFI engine management system. That is why it is so complex and comprehensive. The broad capability of the ME7 is a blessing and a curse. Without knowing all of the ME7 system architecture and the specific strategy employed by a mfg. using the ME7 or similar ECU, it's challenging to get everything right when making major changes in engine configuration or operation. |
|
|
|
|
Logged
|
|
|
|
turbojohan
Full Member
Karma: +5/-0
Offline
Posts: 185
|
|
« Reply #50 on: January 28, 2017, 06:42:34 AM »
|
|
|
IME most of the stand alone systems are intended to provide quick and easy AFR/ Ign. control unlike ME7 which is designed to allow OEMs and others who are properly trained to write code that can control every aspect of engine operation including emissions, regardless of the SI engine design and operation short of DI engines which would use later Bosch MEx systems. The ME7 system was meant to be everything to everyone (at the time of it's introduction in the late 90's), wanting a sophisticate EFI engine management system. That is why it is so complex and comprehensive. The broad capability of the ME7 is a blessing and a curse. Without knowing all of the ME7 system architecture and the specific strategy employed by a mfg. using the ME7 or similar ECU, it's challenging to get everything right when making major changes in engine configuration or operation.
ME7 is far easier for ARF than stand alone. Hard part is allready calibrated at factory, just program correct MAF and injectors and AFR is good at all rpm/load. Most standalone ecu's need to be calibrated at all rpm and load to get AFR right. |
|
|
|
|
Logged
|
|
|
|
|
IamwhoIam
|
|
« Reply #51 on: January 28, 2017, 07:37:47 AM »
|
|
|
You can get a plug and play Syvecs ECU for this type of car/engine, which will make everything inside the car work like stock. Then you only have to tune it, which is dead straightforward.
|
|
|
|
|
Logged
|
I have no logs because I have a boost gauge (makes things easier)
|
|
|
|
gt-innovation
|
|
« Reply #52 on: January 29, 2017, 02:26:40 PM »
|
|
|
You can get a plug and play Syvecs ECU for this type of car/engine, which will make everything inside the car work like stock. Then you only have to tune it, which is dead straightforward.
For a good 2000+ euro if you are not a dealer..That price i got for a 2008 subi which nothing was stock on the engine(mafless - different heads and much more). |
|
|
|
|
Logged
|
|
|
|
|
RBPE
|
|
« Reply #53 on: January 29, 2017, 07:07:32 PM »
|
|
|
The reason they fail is they are not really supposed to be pressurized to 40+ psi... That combined with seeing 1000C pretty often severely reduces the service life.
Also the temperatures in that map are irrelevant. The map is filled by running lambda 1 on engine dyno steady state. You can see top end is not fully calibrated.
In reality because there is at least some enrichment way before that, on a NA application the temps never reach that high...
Yes, I am sure none of those things help - you're overlooking the bit where I said it was for knowing the info on the components, blasé statements that give no real information and some things for people to think about hence the laminar/turbulent, banks read and heat statements. You're dwelling on it in that it is fundamental to doing this tuning which I have already said it isn't. Actually, Bosch add "and/or through driving" as opposed to merely on a dyno - probably also a fundamental point in terms of pedal feel with the mods if people do compensate for KFVPDKSD via DK/LLR type tuning. (Throttle and idle control acronyms to the noobs). Anyhow, there are no ways to compensate in code for stuff like the lack of KFVPDKSD... This absolutely destroys part throttle, because the ECU is always fighting you.
Not a problem on a supercharged car, but a big problem on a turbocharged car. This is further compounded by the lack of pre-throttle pressure correction when calculating throttle control. So you can try to encode this information into the tables, but it sucks during transients.
Never mind the hardcoded load limits on some of the ECU's. Lack of boost control - which can be added, but it's all custom code. And so on.
Don't get me started on the DSG<->ECU interaction and maintaining proper idle control when doing it right. Oh and asm code mods anyone on tricore DSG controllers?
Well, no, although I agree in that emulating Bosch's FI vehicles as best possible would be the best route for some, it's not always a practical solution for many. As for the ecu fighting you statement - well, you can negate such things to a degree and have it bypass maps completely if it affects the tune. Not saying it is necessarily best to go down that route or that it even makes tuning easier, just an option people should be made aware of. My point is - none of the people on this forum who are hobbyists will EVER get this done right.
In fact, most of the "pro" tuners on this forum will never get this done right.
It's like the hardest job on the ME7 that you can do. It requires intimate knowledge of the ecu including coding knowledge.
I agree that it is the hardest thing to learn in this field but no, it's not exactly difficult is it? Quantum Mechanics is a hard field to learn, not this, this is just time consuming learning the system interconnectivity and the electronics side (if you are new to those things). This difficulty of which you speak though can be hugely negated too, people can learn anything, they just need a good teacher! I would prefer it if you helped them in that respect instead of showing general disdain for those that set out to do it as I have seen you do. Anyway - Nick, start at 1097C on your file and reference it with regards to either the file I posted up or this one, I forget what's in what. I'd suggest looking at the file marked "FI throttle" in DK LLR control which is (probably more aimed at noobs) - info I have taken from the English FS in terms of KFVPDKSD. There should be plenty for the noobs to get their heads around in that one for that. I've also added a bit of patent info on it as well as some of the other systems and I have added some general misc info and info on the electronics side too which should all help. https://app.box.com/s/2red8vinrhsuz54zt7sh88tf9x9q528760MB file so couldn't add it here. |
|
|
|
|
Logged
|
|
|
|
c4andmore
Newbie
Karma: +1/-2
Offline
Posts: 11
|
|
« Reply #54 on: January 29, 2017, 08:34:35 PM »
|
|
|
ME7 is far easier for ARF than stand alone. Hard part is allready calibrated at factory, just program correct MAF and injectors and AFR is good at all rpm/load.
Most standalone ecu's need to be calibrated at all rpm and load to get AFR right.
There are different tools for different people and their individual objectives. ME7 can be run as a load based, a speed density or Alpha-N system depending on people's desires. In fact Bosch sells motorsport ECUs built off of the basic ME7 architecture. These systems eliminate a lot of the maps offered in ME7 as they are unneeded for motorsports use. There are numerous aftermarket stand alone ECU systems and some offer pre-populated maps for a given application that typically need minor adjustments to obtain a satisfactory performing engine. Regardless of the system chosen the ECUs are tools to obtain an acceptable EFI/Ign. management system for a given application. The ME7 just happens to offer everything to everyone. With the ME7 however there is a much steeper learning curve - especially when only the pros who do engine calibration for a living typically have access to all of the ME7 documentation. With thousands of maps available in the ME7 ECU and the numerous subsystems that enter into the final calibration value, the ME7 can be a wonderful engine management system or a real challenge to get correct - especially when making significant alterations to an existing engine configuration. If you are just looking to make minor tweaks on an existing engine that uses ME7, then it may be very practical to stick with the ME7 system. If you're making major engine alterations that require big changes to the ME7 then an aftermarket stand alone system may be more appealing to some folks who don't want to deal with the complexity of the ME7. Engines have run perfectly fine on speed-density and Alpha-N systems for decades. |
|
|
|
|
Logged
|
|
|
|
|
Jim_Coupe
|
|
« Reply #55 on: January 30, 2017, 02:23:40 AM »
|
|
|
What makes a standalone ECU great is that it is straigt foward.. Fuel+Ign+Boost control.. I have experimented on this on my TT 3.2T E85 for a while now.. My route is Disable everything that can be disabled. I have disabled every damn thing i could find.. gaaaaarrhhggghghg... trail and error rules.. Just give med the damn fuel and ign and keep me at desired Lambda and nit to high EGTs is my philosophy. I dont care about stupid limiters  I dont bother to read for years to understand safety functions and limiters you dont have them in a standalone anyway "sort of". Well i keep BTS maps stock with raised TAGBTS limit.. its a nice little function to have.. I dont see whats the problem? But i shall admit there´s some strange things going on in this ECU if you dont bother to study it in the correct way. If I was selling tunes i would NOT go down this route though.... I think Nicks question is easier to answer than he thinks But dont trust me im an idiot that just boosts and see what happens. Just look at my posts vs Karma <---- hahah |
|
|
|
« Last Edit: January 30, 2017, 02:29:02 AM by Jim_Coupe »
|
Logged
|
E85oholic
|
|
|
Nick_T
Jr. Member
Karma: +3/-1
Offline
Posts: 42
|
|
« Reply #56 on: January 30, 2017, 10:02:28 AM »
|
|
|
Thanks for all the help and info guys! Especially RBPE!! Its next level help lol  So far im flipping happy with the car! Pulling like a monster!!! And part throttle is not bad at all! Maybe its just because of my massive exhaust housing lol  I cant push for anymore power as my clutch has said its had enough, so now im backing the boost down a little and start working on the finer tuning! Trying to source a place to cut me a SMF at the moment, and then find a sexy clutch kit! So far my current tuning issues: I couldnt get a nice ride with the 100mm maf housing, I had it adding fuel at idle and part throttle, and removing fuel at WOT. Was a nightmare. So as the stock maf is still perfect for my current power, i went back to it and did a few tweaks. Hitting 110 load with no torque intervention. Firstly, im still struggling on getting my proper requested AFR in my BTS fueling as i still cant find my KFFLDBTS map. I see in other maps its usually just before the KFLBTS map, but now i have KFLBTS1 and KFLBTS2. On the same note, if i come to a stop pushing my clutch in quick after a pull the car dies, but if i coast to a stop its perfect, rpm doesnt drop at all.  KFLBTS by Nicholas Tolken, on Flickr 4th gear pull, requested and actual AFR. You can see the flipping AFR climbing lol Hoping that its just the factors in the KFFLDBTS map that im missing! I havent done any KFKHFM corrections yet!  IMG_20170130_183522 by Nicholas Tolken, on Flickr The next issue i have is the timing advance, it likes to give a random spike in the high rpm! 4th and 5th gear pull. BTDC and LOAD.  IMG_20170130_183556 by Nicholas Tolken, on Flickr Now from that I have another question, it cant be normal for the load to drop so much right? From my understanding, I have the BOOST constant, BTDC is fairly constant, AFR is not running rich, so i can only think that its the actual flow thats becoming a restriction now? Or am i mistaken? My torque curve follows that pattern, and my torque should not be dropping at all! My next issue, I still cant get my RPM limit increased. I've done a search through my map looking for any 6700 values, yet all the values that I found are in my definitions already and changed? What am i missing -.- Its irritating because on the quarter mile i have to change to 5th gear lol I'm hoping that upping the limit will help prevent me from having to change to 5th gear before the line! Jim how does your cars ride feel with disabling everything? 'just boost and see what happens' LOOOOL pretty much me! |
|
|
|
|
Logged
|
|
|
|
|
Jim_Coupe
|
|
« Reply #57 on: January 30, 2017, 02:00:18 PM »
|
|
|
It feels good, more can be done... its a bit more rough and aggressive than stock. It jerks and jumps a bit lol.. It feels more like it was running on a standalone. Im logging with VCDS and analyzing my logs with LogView.. I had to buy that program.. But the resolution in logs SUCK BIG TIME.. im open for any tips to speed it up.. My fuelling is almost where i want it now. I have been fine tuning for a long time now. Im sure there is lots of stuff to do there still... I get O2 sensor B1 S1 code for lean condition when driving it at partial throttle when i drive it calm on the highway.. I have to look into that.. The AFR looks ok but i might have missed something.. hmm
Next step for me now is to squeeze out everything that possible out of that god forsaken DSG... A new clutch would be nice but im broke as a joke now.
I will resume this Project maybe i in spring late April.. Right now im focusing on learning tuning my Me7.5 1.8.T "China GT28 snowracer". Its a chapter that i have missed. I can learn stuff from there to use on the 3.2. (I did everything in reverse lol).. So i hope to hear what you can push this 3.2T to.. There is always something to learn..
Do you log EGT´s if so what temps? And what fuel are you using?
|
|
|
|
« Last Edit: January 30, 2017, 02:09:50 PM by Jim_Coupe »
|
Logged
|
E85oholic
|
|
|
Nick_T
Jr. Member
Karma: +3/-1
Offline
Posts: 42
|
|
« Reply #58 on: January 31, 2017, 12:02:20 PM »
|
|
|
"im broke as a joke now" lol lol story of my life! Boost is too addictive! Thats awsome though, well done! Im borrowing a friends vcds15.15 cable. Has a decent sample rate when i use the turbo function! Thats a tricky one with only one sensor giving that error, did you check your fuel trims? And do you have a external afr gauge too? Like 3rd lambda in the DP? Flipping cool with your 1.8t! I had the agu which i did a me7 conversion on, and used eurodyne software! Made a very smooth 260kw @1.5 bar (t3t4 50 trim, .48ar). Was fun! Anyways, back to some 3.2T. Ive decided that im changing my manifold, found a flipping easy design online, So i will do that within the next two weeks! something like one of these, atleast then I know my lambdas will be perfect, and i can actually make use of my twin scroll then! Also need to get a new intake mani built up, lost 30 g/s from the thing i made haha  2012-08-08 21.17.10 by Nicholas Tolken, on Flickr  IMG-20120826-00112 by Nicholas Tolken, on Flickr I haven't made a logger for egts yet, for my external one. But it was reaching 920 degrees with a 5th gear pull, so thats fine. That was with afr in the high 12's. Running in the 11's will be better, but no need to really run such low afr right now as i have no knock anyway and egts are okayish for now. Im using our crappy 93 octane that we get here. Any chance you can tell me how you found your NMAX location? and KFFLDBTS lol Link to my file and some definitions http://nefariousmotorsports.com/forum/index.php?topic=12071.0title= |
|
|
|
« Last Edit: January 31, 2017, 12:04:36 PM by Nick_T »
|
Logged
|
|
|
|
|
prj
|
|
« Reply #59 on: February 02, 2017, 03:32:20 PM »
|
|
|
I agree that it is the hardest thing to learn in this field but no, it's not exactly difficult is it?
Not everyone's mind is built to reverse engineer assembly code. Some things you need natural intelligence to become good at, an ability to see patterns, an analytical mind. To tune this correctly you need to rewrite portions of the ECU code. If you are a computer programmer doing cars this is not difficult. If you are a car guy trying to do computer programming then it is extremely difficult. Bosch had a whole team of developers write this ECU code. You're not even writing code. You are reverse engineering it. Alone. |
|
|
|
« Last Edit: February 02, 2017, 03:34:30 PM by prj »
|
Logged
|
|
|
|
|
