MyE28.com

Since I've had the MBC installed and gotten a taste of what 10 psi feels like I've been really itching to get it to run there safely. Last weekend I got a set of 24 lb injectors from TCD to see what I could do with them. I don't have a 535 AFM, so I had to modify the m20 AFM to work. After a few hours, I ended up with a big pile of aluminum shavings and a big hole drilled in the AFM flapper door. With the boost cranked up to 10 psi, it still detonates between 3k and 4k before it really comes on cam. I can get enough fuel in there to get a super rich high rpm cut out, but it still detonates down low. I guess the stock 325i chip doesn't allow enough fuel, or has too much ignition advance, to keep it from pinging until it hits 4k. I'm just about ready to blow my next load on a split second kit and kick the AFM to the curb, but if its the stock chip causing problems, that wouldn't quite do it. Any ideas?

I think I remember Todd having a similar issue when he orignally turboed his e30. I seem to remember the problem being too much advance in the mid-range with the stock chip. He eventually burned a custom chip to solve it I believe. Contact him, he'd probably be willing to sell you one.

Jeremy

PS - why are you drilling holes in the AFM flapper??

Dan,
there are a # of issues with the M1.1 and M1.3 ECUs. Your only scratching the surface. Your biggest issue will be long term fuel trims if you don't run a O2 clamp.

RussC

Well, at my altitude, with a 2.5 bar FPR and lots of spring tension, the 24lb injectors supply far too much fuel. It barely idles with no modification. Spring tension fixes the idle, but it dies when you give it any throttle. The extra spring tension in conjunction with the "bypass hole" makes it driveable in the low and mid range, but as soon as you put it under load and open the AFM flapper up more it chokes and stutters on the extra gas. So I needed to reshape the inner curve of the AFM. After all that it still slightly hiccups at about 5500 RPM, but it runs good overall. It runs great NA with stock 325i injectors at 2.5 bar up here at 5500 feet. Those are 14.75 lbs/hr at 3 bar. The 24lb/hr injectors are certainly much larger, and it really doesn't need any more fuel anywhere other than on cam under boost. It took drilling and a lot of grinding to take all the extra fuel out of where it didnt need to be. I think I'll try the AFM off of my old man's MAF converted s38 next before I spring for the split second kit. Perhaps before the compressor.

russc wrote:Dan,
there are a # of issues with the M1.1 and M1.3 ECUs. Your only scratching the surface. Your biggest issue will be long term fuel trims if you don't run a O2 clamp.

RussC

I actually dont have an oxygen sensor installed on this one. I find its easier to tune these running open loop. I'm really thinking that I might end up with a split second kit just for fuel. I think I can probably trick the stock ecu into working correctly with the signal calibration box. As long as the ignition advance curves are good to me I think it'll be alright.

Your car will never work right without running the ECU closed loop. The ECU needs the O2 feedback to properly calibrate for bigger injs. If it has no feedback, your stuck running stock injs, period. I wish you would have given this info before.

RussC

turbodan wrote:

russc wrote:Dan,
there are a # of issues with the M1.1 and M1.3 ECUs. Your only scratching the surface. Your biggest issue will be long term fuel trims if you don't run a O2 clamp.

RussC

I actually dont have an oxygen sensor installed on this one. I find its easier to tune these running open loop. I'm really thinking that I might end up with a split second kit just for fuel. I think I can probably trick the stock ecu into working correctly with the signal calibration box. As long as the ignition advance curves are good to me I think it'll be alright.

It wont be perfect, but it'll be close enough. Even with an O2 sensor the stock ecu wont do it all for you. The way I understand Motronic 1.1/3 is like this: The oxygen sensor control adjusts the basic pulse width. Thats the basic injection value at idle, under load and at WOT. So any adjustment the oxygen sensor makes affects the entire range. If its rich at idle, and the O2 leans the mixture out, this adjustment will also apply at WOT. Its an electronic way to slightly reduce the actual injector size, to compensate for minor things like altitude, engine wear and air leaks. These factors can be corrected for by adjusting the basic pulse width. It is not designed to fine tune the fuel maps for different VE curves or forced induction.

I dont believe it can adjust to an unlimited extent either. I can say for sure that it wouldn't be able to adjust out the extra 9lbs/hr of fuel these injectors deliver. It might be able to lean out 30% (a randomly chosen figure) of that, but it'll still be rich. And that 30% adjustment applies to the entire map. At idle, cruise, and WOT.

Thats why I dont use the O2. I'd rather make the adjustments myself, like jetting a carb, than count on the O2 sensor to make it work for me. I had nothing but trouble trying to tune my last car with the o2 sensor input. Its all about what the ecu knows. Without an oxygen sensor, it doesn't know it has 24lb injectors installed. All the ecu sees is RPM, a basic load signal from the AFM, coolant temp and throttle position. The idea is to modify the AFM load signal so that it runs appropriately at low loads as well as under boost. The stock ecu with closed loop lambda control will never do that, which is why I've abandoned the O2 sensor. I can make it work without that. That is unless the ignition advance curves are too aggressive for 10psi, which is something I can't tune with AFM modification.

I have an 061 M535i ecu that does not use an O2 sensor and an used M30 AFM if you need them.

Well it looks like it was the ignition timing. I was digging through the pile of parts I had hoarded for the turbo project late last year and I found a chip I got on ebay from some guy that was supposed to have retarded WOT timing maps. I have no way to verify this, but when installed I ran it up to 12 psi with no detonation. It did seem to go real rich at the top end, over 5500, as the car seemed to drop off more than it used to before 6k. I've got it set at 10 for now, a nice, smooth, detonation free 10 psi. I've still got to get the mixture worked out though. It feels like theres more in there.

Um,
You probably wont get the tune you want the way your running the ECU, Sorry. Especially if its a M1.3 ECU. They were never meant to run open loop. The maps are pretty good, but the ECU relys on fuel trims to work properly. You wont get the tune right.

Plus, if the ECU sees a O2 failure long enough(CE code), it may go into any number of limp modes. Some limp modes are non-intrusive, some are very malevolent.

Look forward to seeing a dyno of the system to see if you can get the AFRs right w/o O2 feedback.

Good Luck,
RussC

I'm not worrying about any kind of long term failure modes. Its only designed to replace the AFM signal and coolant temp signal in case of component failure. Everything else, as far as I know, is just as simple as it seems. The fuel maps might not be right on, especially for 24lb injectors and forced induction, but they dont change. I'm counting on doing some fine tuning. In this case I'd rather do that myself. Like jetting a carb. Thats why I like Motronic 1.1/3, its modern enough that it can do everything it needs to do, but not so much that it dicks itself up with self diagnostic crap like limp modes. I dont have a bulb in for the CEL, and thats about all I expect to do there.

Theres one shop in town with a dyno. I'm hoping to get in there early april. I will certainly report back as soon as it goes down.

turbodan wrote:...I think I'll try the AFM off of my old man's MAF converted s38 next before I spring for the split second kit. Perhaps before the compressor.

Your dad has a split second maf kit & you gave me all that grief for liking my modern performance maf & the split second stuff? What the hell?

Which split second device are you looking to get?

Martin in BellevueWA wrote:

turbodan wrote:...I think I'll try the AFM off of my old man's MAF converted s38 next before I spring for the split second kit. Perhaps before the compressor.

Your dad has a split second maf kit & you gave me all that grief for liking my modern performance maf & the split second stuff? What the hell?

Which split second device are you looking to get?

I only give you grief for the comparison you make between your MAF with tuned air fuel ratios to a stock AFM with stock air fuel ratios. You claim the gains are from removing the restriction the AFM presents, but in your conversion you also changed the air fuel ratios. You dont account for the different air fuel ratios when you claim the AFM is more restrictive than the MAF. Theres no doubt you gained power, but how much of that was from the different air fuel ratios and how much was from removing the alleged restriction the AFM allegedly represents? Its a bogus comparison with more than one variable. Seventh grade teachers string kids up for that sort of thing. And theres no good reason to go around stirring up hysteria that the AFM is corking up 25 hp. If yours was, it was probably broke.

Anyway, I'm not replacing the AFM to eliminate a restriction, I would be doing so to be able to adjust the mixture on the fly, as you can do with the ARC2. It really is a pain in the ass to have to remove the AFM and grind metal out to make adjustments for one single point in the operating range. I'd much rather turn knobs.

See,
You made my point. The ECU may be in some limp mode right now, which is screwing up your tuning. All M1.1/3 ECUs have limp modes, I know, Ive triggered them o mine. The limp modes associated with issues with the AFM are very intrusive.

Again, good luch. Keep us informed.

RussC

turbodan wrote:I'm not worrying about any kind of long term failure modes. Its only designed to replace the AFM signal and coolant temp signal in case of component failure. Everything else, as far as I know, is just as simple as it seems. The fuel maps might not be right on, especially for 24lb injectors and forced induction, but they dont change. I'm counting on doing some fine tuning. In this case I'd rather do that myself. Like jetting a carb. Thats why I like Motronic 1.1/3, its modern enough that it can do everything it needs to do, but not so much that it dicks itself up with self diagnostic crap like limp modes. I dont have a bulb in for the CEL, and thats about all I expect to do there.

Theres one shop in town with a dyno. I'm hoping to get in there early april. I will certainly report back as soon as it goes down.

Dave Reid says he unplugs the ECU on his e34 once every week or so to clear the fault codes and cancel whatever "limps" it's thrown itself into. This is probably a good idea for you as well, Dan.

Jeremy

Jeremy wrote:Dave Reid says he unplugs the ECU on his e34 once every week or so to clear the fault codes and cancel whatever "limps" it's thrown itself into. This is probably a good idea for you as well, Dan.

Jeremy

No, I'm not superstitous. I guess thats right right word for it. To me wiping out your ECU every week just to do it is stupid. To assume its in some mysterious, very subtle limp mode for no reason at all doesnt make sense. Of all the literature BMW has regarding the various features of Motronic 1.1/3 I've yet to come across any information regarding any long term or delayed limp modes.

In my case, if you unplug the AFM, it goes into the AFM limp mode. I can only imagine thats because the AFM can only produce a signal from about .2 volts to about 5 volts. When unplugged, the ECU sees zero volts, which would be out of normal parameters for the AFM, so it goes into limp mode. With bigger injectors, this means the car doesnt run. It idles, but dies the instant you open the throttle. Thats the AFM limp mode, and it doesnt do that. Even my modified AFM's signal is still within parameters, so theres no reason for the ECU to decide its bogus and go into a limp mode.

The others are for the coolant temp sensor and intake air temp sensor. The replacement value is 170F for the coolant sensor and 86F for the IAT sensor. Both of those are simple thermosistors, so I guess the normal parameters might be anywhere between an open and a short circuit. Now, since the IATS and CTS are stock and working properly, I can only assume they're not in limp mode either.

Other than that, theres no information out there about any other emergency running fixed replacement values, or limp modes. Other things like a faulty oxygen sensor or purge valve will trigger a CEL, but there is no information about any limp modes (or emergency running fixed replacement values). I dont even know what kind of limp mode it could go into without the O2 sensor input. There really isnt any fixed value it could use, and there arent any other fuel maps in the ECU other than the standard idle, off idle and WOT maps. I don't consider open loop to be a limp mode, but thats about all it can do here. Or am I missing some information?

Not so much that it goes into a limp mode as such, but the 1.1/1.3 tries to to "learn" and changes different things to make things match what it thinks it should be seeing. A heavily modified (turbo) setup will have modifications that throw some inputs that look wacky to the ECU. As it tries to adjust, the car can develope some odd running conditions and problems.

That's why Dave resets the ECU, to bring it back to zero, not because of any kind "superstitions". Course, he's also running closed loop, so that may have something to do with his need to do it as well.

Jeremy

Thats one thing no one is really clear on. In general I think the adaptive capabilities of 1.1/3 are vastly overestimated. Even as I read through the technical bulletin from 1987 on the spiffs of the new Motronic 1.1, I'm not quite sure myself. The two specific points discussed are the Adaptive Idle Speed Control and Adaptive Lambda Control. I couldn't care less about idle speed control. It idles fine. The Adaptive Lambda Control is described as follows:

"Compared to its predecessor, the M1.1 Motronic has an adaptive system with extended regulating and operating range to adjust the air/fuel mixture according to the feedback of the oxygen sensor, which includes the operation and regulation of the purge valve from the evaporative system. Altitude influences as well as air leaks are also compensated. Fixed programmed values are compared with actual ones.

In case of deviation of the fixed values, the system automatically corrects and stores the new values in its memory. There is a constant renewal of the correction factor. The memory can be cleared for troubleshooting purposes by disconnecting the power supply. If the memory is interrupted, new adaptation takes place in a few minutes of engine running time.

Malfunctions of the Lambda Control System that could cause the engine to exceed exhaust emission standards are recognized and stored in the diagnosis memory. The "CHECK ENGINE" light then will be illuminated and the 2-flash code will be displayed with ignition key in position 2.

Important: If there is a malfunction indicated in the Lambda Control System, there can be different reasons that are not necessarily related to a defective oxygen sensor. Follow troubleshooting procedures thoroughly."

As found in BMW's TRI B 12 01 86. If your lucky, you can get your BMW parts guy to print a copy for you. Thats where I got mine. Anyway, to me it looks like this adaptive capability is only really effective as far as minor air leaks, wear compensation and altitude. To what extent it can adapt and how the adaptations affect engine performance are not clear. In my own experience, I can say that the adaptations seem to affect the entire range. From idle to redline. The basic pulse width, which is used to calculate the effective pulse width after engine speed, load and temperature are factored in. Unless youre running lean or rich all over, this adaptive capability isn't going to help. Especially not if youre running rich down low and lean on top. It'll lean it out so it runs correctly down low and even leaner on top, as was my experience with my white car. Once I disconnected the O2 input on that vehicle it ran predictably, and every change I made was left alone. The idea was to be able to force it to run a little richer on top while running correctly down low. Without adjusting the fuel maps, this isnt possible. I never had any problems with any odd running conditions or strange adaptations. It has no feedback signal, so it does nothing. The only way to be sure of this would be to examine a schematic representation of the 1.1/3 ecu. There used to be something like that on FR Wilk's old porsche 944 website, but that was gone last I checked.

So as far as I know, and guessing by the many months I;ve run 1.1/3 in open loop, the only difference is that it doesn't attempt to correct the baseline values. It doesn't get confused and do weird shit. It doesn't limp around. It just runs off of the baseline values. It says at "zero". If Dave likes the way his car runs immediately after a reset, why doesnt he try open loop? It'll stay like that.

turbodan wrote:...I only give you grief for the comparison you make between your MAF with tuned air fuel ratios to a stock AFM with stock air fuel ratios. You claim the gains are from removing the restriction the AFM presents, but in your conversion you also changed the air fuel ratios. You dont account for the different air fuel ratios when you claim the AFM is more restrictive than the MAF...

Again, I was making a practical comparison, with my car. The car ran better with a maf (and now map) than it did with an AFM. I don't know how to effectively tune the AFM for ideal afr's; apparently, neither do you.
I claim(ed) the gains are from replacing the afm with a piggyback kit & some tuning.

and solve your problems, I've been running 10 lbs boost trouble free for awhile you can tune it exactly how you want it. you can also see exactly what its doing to make the corrections

There is a large difference between M1.1 and M1.3. 1.1 only has 4 error codes, 1.3 ~20. Also, 1.3 has alot more than 3-4 maps. I believe there are 8. If your using 1.1, thats easier as it is less complex than 1.3.

Also, M1.3 never runs completely open loop anywhere, even WOT, like all modern BMW ECUs. There are trims for all operating areas. This came right from JimC.

I have the schematic for the 179 ECU(M1.3). The CPU(8051 derivative) samples the O2 sensor voltage through the on-board A/D. That way it can make decisions on fuel trims through out its range.

RussC

Martin in BellevueWA wrote:I don't know how to effectively tune the AFM for ideal afr's; apparently, neither do you.

Such drama. . To tune the AFM is one thing. To tune it for 62% larger injectors and a rising rate FPR is very different. Its not infinitely flexible. I'm sure your car ran better with the MAF. That has never been in question. The debate is how bad the AFM really is. Unless you have tested the MAF at the same air/fuel ratios as the old AFM, you can't say anything. You have to isolate the variable you're testing. I bet the gains from the free flow MAF hardware VS the AFM flapper are on the order of five horsepower on an m30b35 with the same air/fuel ratios. Thats about 2.5%. I wish you would prove me wrong.

russc wrote:There is a large difference between M1.1 and M1.3. 1.1 only has 4 error codes, 1.3 ~20. Also, 1.3 has alot more than 3-4 maps. I believe there are 8. If your using 1.1, thats easier as it is less complex than 1.3.

Also, M1.3 never runs completely open loop anywhere, even WOT, like all modern BMW ECUs. There are trims for all operating areas. This came right from JimC.

I am aware of the increased capabilities of 1.3. The four digit flash codes VS the one digit codes from 1.1. I dont think 1.3 has a limp mode for each fault code, however. If they're so subtle I can't tell if they are or aren't activated, I dont care anyway.

The fuel trim was a problem for me in the white car because it would adjust to the range that the car was in 95% of the time. Below about 3k rpm. When I got out of that, it wasn't able to react quickly enough to keep the mixture correct. So I got the detonation. If I ran the car between 4k and 5k all the time, which would have been tough, it would have been fine. It just doesn't work instantly, which is too slow for any kind of realtime correction. Its designed to compensate for air leaks or elevation changes. Neither of which cause massive, almost instant instant changes in the mixture.

cwl4m5 wrote:(Do a megasquirt) and solve your problems, I've been running 10 lbs boost trouble free for awhile you can tune it exactly how you want it. you can also see exactly what its doing to make the corrections

That would be a very last resort. I don't have a laptop or any experience with DIY FI. Since I found that chip I had laying around that had the timing retarded, I just need to get the fuel under control. If a larger AFM doesn't cut it, I might go with the split second kit. I don't see myself doing MS. Maybe I'll feel differently if it cointinues to frustrate me for another few months.

turbodan wrote:..The debate is how bad the AFM really is. Unless you have tested the MAF at the same air/fuel ratios as the old AFM, you can't say anything..

I never stated that the maf was better than the AFM only because it was less restrictive. You misquoted or misunderstood me, and you are continuing that misunderstanding here.

I've spent too much tine tuning the piggybacks to disregard the benefits of being able to set the maf/map up for more ideal afr's. That's always been a benefit with these kits. I've been to the dyno a couple times to tune the maf, because it gives me that ability. I've also got a wideband o2 sensor, because it helps me to better tune with a piggyback kit.

I've asked in another thread, how you & Jeremy propose to tune an AFM for ideal afr's. Jeremy said it was a super easy thing to do, after popping off the plastic cover, & you are drilling holes in the flapper door.

Martin in BellevueWA wrote:I've asked in another thread, how you & Jeremy propose to tune an AFM for ideal afr's. Jeremy said it was a super easy thing to do, after popping off the plastic cover, & you are drilling holes in the flapper door.

Back the bus up. You have a stock M30B35. Stock injectors, cam and intake. Tuning this engine is clearly different than tuning a 2.7l M20 using a 2.5l ECU, 62% larger injectors than stock and a rising rate FPR. 62% more fuel is a lot more than you can account for with spring tension, hence the hole in the flapper and AFM porting. Even then, you dont want to adjust all of it out, as that would defeat the entire purpose of installing them in the first place. The idea is to add fuel at only a few specific point in the range. Only where you need it.

Especially with a WBO2, it would be easy to adjust the spring tension and slightly reshape the inner curve of the AFM to get what you're looking for. Thats if you wanted to modify the AFM just to slightly change your AFR on a stock engine. That seems silly to me. Especially since a healthy AFM with a basic performance chip does the same job.

turbodan wrote:.. You have a stock M30B35. Stock injectors, cam and intake..

I did the maf and then the map because I like tuning my 535 with a 284 cam, 21lb injectors, & at one time, RD long tube headers.
Are we going to hear any more horsecrap about your confusion on why I removed my afm?

Martin in BellevueWA wrote:

turbodan wrote:.. You have a stock M30B35. Stock injectors, cam and intake..

I did the maf and then the map because I like tuning my 535 with a 284 cam, 21lb injectors, & at one time, RD long tube headers.
Are we going to hear any more horsecrap about your confusion on why I removed my afm?

Well I guess we're about done here. I still dont think a NA 3.5l with 15% larger injectors, some headers and a mild cam is quite as tough to tune as a FI 2.7l with 62% more injector (or more depending on fuel pressure) and 10 psi boost. I guess we'll just have to agree to disagree. Lets remember that the point of this thread was to discuss ignition timing curves on the 325i chip. You joined in to start that shit about the AFM's again, and to ridicule me for my "confusion" about tuning my engine. Thanks for your insight. I'll try to let your snippy remarks go next time so we can save this bandwidth.

Martin in BellevueWA wrote:

turbodan wrote:...I think I'll try the AFM off of my old man's MAF converted s38 next before I spring for the split second kit. Perhaps before the compressor.

Your dad has a split second maf kit & you gave me all that grief for liking my modern performance maf & the split second stuff? What the hell?

Which split second device are you looking to get?

Holy cow, people.

Watching someone trying to get Motronic to work well with boost is like having a friend that dates ugly, bland women and refuses to give them up. It's painful to observe.

Megasquirt is cheap. It might not be perfect, but it's good enough for most of us and beats the hell out of Motronic any day.

turbodan wrote: You joined in to start that shit about the AFM's again, and to ridicule me for my "confusion" about tuning my engine. Thanks for your insight.

Your confusion is with my wanting to ditch the afm & realizing a better running car. Thanks for getting it.

Martin in BellevueWA wrote:

turbodan wrote: You joined in to start that shit about the AFM's again, and to ridicule me for my "confusion" about tuning my engine. Thanks for your insight.

Your confusion is with my wanting to ditch the afm & realizing a better running car. Thanks for getting it.

Theres no confusion there. Read it again. I think your old AFM was dicked up if it was holding so much back. Either that or you were finally able to tune out the extra fuel from the (likely unnecessary) injector upsizing. And when you spout off saying "I gained 20 hp with my MAF" without mentioning the 284 cam, headers and injector sizing, its very misleading. Thats all my beef is, encouraging others to spring for a MAF and more for tons of $$$ without considering that they likely wont realize the same gains. The same gains that modified cars get from a custom chip for much less $$$.

The posted dyno sheet wasn't with the 284 cam, injectors, or headers; they came later.