I thought this was rather interesting... the new Hemi uses a speed density system. No MAF to be found and from numerous reports the actual combustion design of the engine was really difficult to get past emissions. Then one would ask, why go with speed density when MAF's are sometimes assumed to be more accurate for fuel delivery.
Just thought it was interesting and no, I'm not saying speed density is better than MAF. My point is that either system will work because they're doing the same thing. It depends on how they're handled and implemented more than anything else. As Grumpy has pointed out in the past, the early MAF setups had some really annoying "bugs" if you would, same with early SD code.
Back to the Hemi being SD. Does anybody know what kind of computers are being used? They've gotta be powerful and fast to be able to operate the engine, throttle by wire, displacement on demand, cruise, learning throttle control, and various other tid bits. I'm just guessing here but 32-bit MAP sensor resolution wouldn't suprise me. Are we going to see GM ditch the MAF for a cheaper pressure transducer (MAP)?

 

I don't know what processor is in the PCM, but Chrysler loves speed density. I really cant think of anything off the top of my head that they make that has a MAF sensor. Maybe the Viper?

Look at all the mustang guy's that swap cams and injectors without tuning the engine and still put out decent power. I think MAF allows for some pretty half assed tuning to get by so it will probably always be around.

I prefer speed density, and with a turbo setup its the way to go. Go read ls1tech.com for a while and see how many people over there are ditching the MAF and going speed density.

 

Interesting, we talked about the same thing this weekend in Iowa. Talked a little about the lack of MAF and the originations of their PCM.

But I forget all the good info like who makes the PCM's for chrysler.

If Bruce reads this and chimes in, I know he knows who makes them(tool long of a day for me lol)

later
Jeremy

 

The new Chrylers are supposedly a PITB. There's no real PCM, everything's on a mylar sheet, and when you crack the case, everything just flies apart.. According to *sources*.

Just another testimonial to sticking with GM ecms, IMO.

 

thats one of the nice things about speed density is that there are multiple axis that define the VE. Inherently adds resolution. could do the same thing with MAF and maybe use RPMs and airflow to define what pulsewidth is commanded, but if your doing that then you might as well just do away with the more expensive maf sensor and use MAP.

Interestingly, I actually took a close look at my '96 Taurus and was surprised by the lack of complexity. theres only a MAF for the fueling and the only real neat thing they do is have a backpressure sensor for the EGR. Probably to help improve corrections for when its active. Other then that it looks alot like the tpi cars. They just have a vac ref'd FPR, couple of sensors, and the MAF. Not even a knock sensor present. Cant say thats such a bad thing given all the mechanical noise that old engine makes, although it detonates like crazy in the heat.

 

As far as the sensors go, I dont think the MAF will dissapear any time soon. There are some plusses to it, like simplified tuning and less need for temp compensation. The only thing taht really bothers me about the MAF is that it can gag the motor when you stick the gas down. Its a real problem with the MPFI on the fnord. Blows black smoke in the morning if I stick the gas down too fast when I first start out.

 

So thats how those wacky germans are doing it these days?

 

To everybody that thinks the MAF will be around for much longer. Consider the speed and power of todays simulation machines (computers) and you might find no reason for it. Every minute of every day their simulations get more and more accurate. They're entering complete models now and simulating them so the only reason things wouldn't work out is if the 3D model is different than what is simulated. They're getting DARN close to not even "need" a dyno simulation. When I say 3D model I'm not talking just engine with intake and exhaust. I'm talking total powertrain simulation. They do testing of the chassis, enter it all in and could tell their simulations whatever they want. Humidity, temp, altitude, all that stuff is done now requiring very little dyno tune time. Just thought I'd through that out there. Now as far as the "performance" cars that are expected to be modified... I'd believe the MAF's would stay with them to allow for the aftermarket to do what they do best.
I just thought it was interesting .

 

It wouldn't surprise me if the lack of MAF sensor is simply due to cost savings. Crack open the hood of the nearest 2005 L6 equipped GM truck and show me the MAF...... However, perform the same procedure on a 2006 and the story changes. I'll apologize to any Chrysler afficianados for the following if taken the wrong way. But, if there's a company that always looks for the cheap way of doing things in the electronics/controls arena, it's Chrysler. The PCM construction sited above is a good example. Until recently, their traction control systems have been brake only without powertrain intervention. I'm sure there are other examples, too. If they don't *have* to spend the dollar on something, they don't.

I expect this will be standard equipment on vehicles within 10 years after all automotive powertrains going hydrogen-based...... Seriously, the OEMs are nowhere near that level of sophistication for on-line control for all powertrain configurations. However, there is a *LOT* of work in that area which has contributed to engine design work. And, hybrid powertrains will require more modeling. Unfortunately, the more complicated the algorithm, the more computer you need to run it. Hey, maybe we ought to talk with Mr. Lutz about this. He's always complained about needing to put cash in the trunk to sell cars. Instead, we'll put supercomputers in the trunk! Consider that compared to the cost of a MAF sensor.....

Part of the following expansion is the planned installation of 60-90 dyno cells.
http://www.autointell-news.com/News-...y-28-04-p3.htm

If only we could put a sensor on the vehicle to determine the difference..... (Sorry, couldn't resist that one.)

 

That last quote, sorry, what I ment to say was that the only time the simulations are off is if the REAL world 3D model is different than what is simulated.

As for cost being a major player, definatly. The thing that doesn't work out is generalizing cost saving measures. They're EVERYWHERE and EVERYBODY is doing them, not just Chrysler. If it's not there or the bean counters figured it would be cheaper in the long run to produce a car with expensive electronics vs more tuning then so be it. Turn around and beating the competition with your model is all about time and in the past, tuning has been somewhat of a slow process but now they don't need to ship the cars to remote locations. They've got dyno's that can simulate any conditions so it's all about saving money and cutting down development time.... not having a MAF is rather strange be it Chrysler or any other company. Maybe, just maybe, the MAP sensor they're using is reporting a faster signal... it is afterall between the air metering device and valves instead of up-stream! Lots of variables, who knows for sure. I bet I couldn't find 1 person in Chrysler that could answer the question... oh well.
So some think the MAF is here to stay, I think it's on it's way out, time will tell.

 

Mopar has been electrically challenged since the first pointless ignition systems. Just a glance at the lean burn set-up will document how little they knew about electronics. It was so bad that for several years they used the GM CCC carbs on their fleet cars.

Hydrogen has some really tough problems that they haven't come close to solving yet, for even a remotely practical personal transportation vehicle.

 

not according to my sources but i have yet to actually check this with my own eyes as for the pitb structure. the chrysler (starting in 2000 on some vehicles ) mcu is a mpc565 motorola power pc AKA ( spanish oak) there is absolutely no hope of disassembling this code in binary form. secondly it uses the newer CAN architecture for communications. also it has a very tight seed/key structure making it difficult to read let alone program.

as for why chrysler uses the MAP it has to do with cost. on good word i have it sevral years ago chrysler approached hitachi about usng a MAF system and was put off by the price.

as for the a/d bus havent physically checked it but it is most likely a 24/32 bit a/d. the new maps sensors do have a much higher volt division per vacum icrement then previous models.

also chrysler in there strategy uses the Drive by wire system to reduce the amount of Accel enrichment needed becuase it can aviod large trnasient changes with controlled throttle opening. steady state fueling with either system MAF or SD is almost identical. in terms of emissions output both are the same if you can control the throttle transients the holy grail of any emissions calibrator.

 

why debate?

just use both like the LS1 does.. lol

 

Debate spurs thought.

Actually, I've been planning to for years, but it's taken alot of time to figure out what an engine really wants, rather then just using what GM's done. If one subscribes to *Simplicity is the key to design*, it takes alot more effort then just throwing resources at a project, and understanding what amount of complexity, is the min amount of simplicity. BTW, the LS1 way is EPA/CAFE inspired whereas, I want to let the engine govern what makes it happiest, and cater to that without any other considerations.

 

true... but its also flexible enough that it can give the engine EXACTLY what it wants...... or just what it NEEDS... depending on what you're tuning for... heh, or in my case, both.

between the lean cruise, DFCO, ect i give it just what it needs in the cruise areas... and its flexible enough to give the engine exactly what it wants at high RPM/airflow...

while you're right in how its "inspired", the made it pretty flexible. it has a speed density setup and a MAF setup that work better then anything GM has come up with so far... i just wish it was hacked as open as the earlier ECMs.. theres still little things i want to do... like a built in 2step for example..

 

It would be really cool to hac one of those. But... Little is known about the hardware, theres no info on how to DL the calibration, and youd be stuck in assembly still. The new encarnation of the moto assy. looks like a real pain. The manual alone for the inst. set for the 68000 spans around 700 pages. I personally dont think we'll see anything soon unless someone takes the initial step in determining what it takes to interface with it and determine what some of the hardware does. Once you have that, then everything else naturally follows.

 

Just use a hac'd ecm, and an unused input. Then creatively program in what you need (oh does that ever sound so easy, LOL).

 

well the interface to it is GMs PWM version of the ODBII standard... theres special instructions for high speed mode, but you dont NEED high speed mode.. its nice though... i can post up bins if you want to look at them.
they are somewhat hacked... but the hacked info isnt out in the open.. someone already did write a new "OS" for it too... a 2 bar SD setup for boosted cars... HPtuners sells it.... its a whole new OS, not just a remapped/calibrated stocker... its nice... they want 299 for it ontop of the program though (i already have the program/cable)


info is out there, just not public yet... part of the reason is the interface... its not standard serial traffic that you can just plug into the PC with a transistor..... its PWM... i have a whitepaper on it somewhere... you can get a neat little preprogammed/locked PIC chip that does it for you, but i dont have too much experiance with it.

 

.
Way ahead of ya! Working on the $0D. The trans half of things is slow going, though. No info at all on any of it other then whats in the calibration section, which isnt alot of help since a good deal of the trans tables are in with the trans code. Once Im done with that, Im going to chop out all of the trans stuff to make room for more interesting things. The nice thing is that this has TIC and TOCs available so driving stuff or reading in a freq. based MAF with 16 bits of precision becomes a possibility

Edit: lost the quote first time around...

 

Some more food for thought; Honda and Acura have been using c as the programming language for a couple years now. The processor is 32-bit! Maybe the newer pcm's have enough memory and power to be the same? If so, it would make them a lot easier to work with than doing low level assembly. I've got a buddy that works for Honda in their calibrations department that is installing 2 Acura 4-cylinder pcm's to control his poncho in his 68 bird . He figures since he has access to all of the comments and they're programmable through a computer interface taht it won't be long before it's up and running. The hardest part he said was getting a few sensors and wheels mounted to the engine.

 

Ill take a look at it. dimented24x7@aol.com

We'd still need someone to actually make the hardware side of things publicly available before a hac can happen. Wouldnt want a bunch of legal strings attached to any work being done.

 

Only problem is that where is teh decompiler going to come from If memory serves me correct those are illegal as hell to have. Not to mention that it would have to match the version of C used to program it.

if anyone has a decompiler though, ill certainly take it!

 

Jon,

BTW, how long are those new cals now? 100,000+ lines? More?

 

the hardware is the bottleneck IMO.

theres some LS1 Bins here:
thats for my 2002 M6 Fbody
http://66.83.134.202/travis/HpTuners/Travis/

the bins here are from my friends 99z.. and some are also illegal for you to use.. they have a hypertech tune on them. so dont download thoes.: (he can use them since he owns the hypertech)

http://66.83.134.202/travis/HpTuners/Jay/


if you go up a level, you'll see subdirectories with names:

Griff - 99 SS
Jay - 99Z... ve mods for cam, timing ,ect...
Marvin - 99z cammed, but only tuned for idle.
PJ - 99z stockish
Travis - 2002.. stock and modded all to hell.. lol

in the files from online folder, theres some bins i got off a public site.... this site: http://www.horist.com/hptuner/ you can look at and download BINs from there.. somewhat useful... i use it to compare stuff all the time.

if you're really into it, the guys at www.hptuners.com are somewhat helpful... but on some of the more sensitive stuff, they're tight lipped...

 

Holy christ those bins are huge. 500K just for the binary.

 

It's not illegal when you work for them doing it day in and day out. Besides, his car is emissions exempt so he could hook up a damn Weber gas grill valve for fuel control and they couldn't say anything... so long as it's not belching flames .

 

I was talking about the decompiler. I guess theyre legal so long as teh software being reversed isnt copyrighted or protected.

 

i checked the .bins from end to end for copyrights. they should appear in the ascii text box and they dont. i do know that software copyrights are required to be readable in plian english in the software in ascii.

 

Well, if you say so.

 

i checked into this. and while it is a pitb it does not just fall apart.

 

Or maybe your way ahead of me, who knows

 

"It was so bad that for several years they used the GM CCC carbs on their fleet cars."

i have to disagree with you on this one Grumpy, sorry

Mopar never used the GM 3-C system on any of their cars. actually, never isn't quite the right word, the later 3-C Mopar systems were made by GM, but there is a twist,......
the first 3-C they used was pure Mopar, which was the same 3-C system that GM later bought from Mopar & has sense been the basis for the later GM ECMs.
Mopar had contracted with a company that was either partly or fully owned by GM to manufacture those systems & someone at GM got a look at them & liked them, with Mopar hurting for money, deals were made, which were kept on the QT. at the time Mopar didn't want people to know they were having financial
problems & GM sure didn't want anyone to know they used Mopar engineering.

i have worked on Mopars made in the 70s that had C-3 systems on them, its been a long time ago, but i know Mopar made some in 77 that were TBI, & they may have used them before then.
i don't know if i still have any of those old factory repair manuals now or not.
once, a very long time ago i had a few connections here & there.

besides lean burn, 3-C, & the system they finally went with, there were other systems Mopar was working on.
there was even talk about how to electronically control spark knock on their new electronic ignition system as early as 71, & possibly even before that, maybe thats part of where lean burn came from.

 

I was also swinging wrenchs back then, both in fleet maintance, and dealerships. The first CCCs I saw were from GM. If Mopar was the originator of the CCCs, then it even paints a worse picture for them, since they let that go, and continued with the odd stuff they've been known for.

The Lean Burn, as explained to me by someone familiar with the system, was just how disconnected Mopar was with engine management, who else would try to meet emission standards for fuel economy with just controlling the spark (well, in large part)?.

Not to mention that if the systems were worth a hoot, we'd be using them on *our* cars rather then the GM units. Before *we* did Programming 101, *we* looked at all the ecms available, and everything was open for discussion, the consensus was, that the GM units far out shined anything else around.

 

bwahahaha.,

if you think a 512 bin is huge, all of the LS2 and up motors have 1024......... not quite the same as the ECMs designed in the early 80s...

 

well chrysler and gm had different needs with there systems. chrysler had EFI back in the 50's with a vacum tube 300c hemi car that was photgrapphed at daytona the system composed of the whole trunk.. gm did buy the early ecm design from chrysler. chrysler code is ecletic but works as well as gm code if you understand the design behind it. i will agree that for a while (80's)the chrysler hardware was odd to say the least. yes chrysler originated the ccc carbs.as for currently. they all now the big 3 are using the same mcu the same compiler similar hardware designs and the code outside of ford is similarly designed.

 

512 K is like 8 of my OBD1 PCM bins combined. I think that one was also programmed in C. Either C got even sloppier then it already is, or theyre going extra super crazy with the stuff theyre putting in. Also a nother factor may be the 32 bit addressing adding extra fluff to the compiled bin.

 

theres more of everything.

more tables.
larger numbers
everything.

and yea, its probably programmed in C... but its still running much faster then it has to...lol


as far as mapping out what/where everything is, its been done.. i mean i could take my HPtuners, change something, and figure it out quickly by comparing before and after bins... almost like cheating... only thing id have to figure out are the check-sums. (still not hard when using editing software.. lol)

 

one thing about C is that it definatly add some needless slop. Although given how powerful it is Im sure it doesnt matter. Of course, all that power promotes lazyness...

As far as using the tuner to find the addresses, thats good for something that youll only use yourself. But, for public stuff it definatly has to come independantly. Dont want any legal issues causing the plug to be pulled on a project. Im sure everyone remembers the 'Mechanic' thing. IMO, also, finding out where all the tables are is only about 15% of whats needed to reverse a computer. Even with a simple table the code still can do weird stuff, not to mention you still need to find all hardware and memory addresses. *** only knows what a mess that is in the new PCMs.

 

Really?, can you explain the differences in their needs?

Have you ever seen a preproduction CCC GM car?. When it was all done with relays, and basically R/C timers?. The one I worked on was a mule that failed in the SoCal heat, years before anyone had seen an ecm. It was a Station Wagon, that the under hood, and rear area, was jam packed with wiring. The guy driving it was shall we say well versed in what was he was driving, and exactly what I needed to do to fix his car (he was stranded, completely and totally). I'd almost venture a guess, GM learned a little about mule cars that day.

Like I said, even if GM bought any of it from Mopar, it just further demonstrates, that Mopar, didn't know what it had.

And FWIW, I was a Big Block Mopar guy, right up until they stopped manufacturing the Big Blocks, and Hemis.

 

nope cant say ive ever set eyes on a preproduction gm ccc car. not sure if id want to. not much to learn there.Mopar knew what it had then it still does now they sold it due to finacial problems. as for differing needs. chrysler had one box tin 2 seperate housings. they did this to reduce noise and heat in the logic side of the game. look at the control strategys implement by all of the big 3 over the last 20 yrs and become apparent they've all evolved in the same direction tain different paths.

i was and still am a big block mopar guy. te blocks suck so i build stroked and poked small blocks now becuae the blocks are so much stronger plus they are lighter.

actually the new gm code ford code and chrysler crp code is al writtn in c. when it comes to the mpc565 motorola mcu they are all using the same DIAB compiler.

 

Ya know, your level of arrogance is just overwhelming. When you say something like that just shows how truely close minded you are. It's amazing how you can work from such hindsight.

BTW, wasn't the EFI you credit Mopar with, actually a Bendix?.

Yep, mopar has had financial problems, and used the American Tax payer to get bailed out, and then sold out to a foreign company. They have continually shown that they don't know what to do or how to do it, since the mid 70s.

Given a common set of design goal parameters (and research funding for the code/ processing end), there are only so many choices for code, and processing. Not to mention the EPA mandates on some of the protocols. It's not cost effective to redesign the wheel an infinite number of times.

 

I"ve seen these setups (Renix) on 86-90 Jeep cherokees. I haven't seen anything else that used it, but they could be out there. Renault owned part of AMC and from what I've heard it was a joint venture between Renault and Bendix (hence the name renix). I guess Chrysler got the credit when they bought AMC.

 

The American Tax Payer only guaranteed the bank loans. If Chrysler had defaulted, the government would have had to pay their debt. As it was, Chrysler paid the loans back ahead of schedule, without drawing on any public funds.

Chrysler still sucks though.

 

i agree, Mopar should have hung on to the 3-C setup. fortunately they didn't.
i also fully agree the GM ECM was the way to go when ya'll decided on which system to run with.
with a Mopar system it would have been too much extra work just to get to the chip, then it has to be unsoldered, plus having 2 modules is a pain.
ford EEC 3 was a very bad joke, EEC 4 was better but just plain stupid with no data stream until around 89 & even then it was limited to very few cars, there was no reason for them to wait that long, (by 85 & no later than 86 every car & light truck made should have had a data stream), although i have always liked the
cylinder balance test being built into EEC 4 port fuel systems, you don't know how much i hate saying that

sometimes odd isn't a bad thing, atleast Mopar jumped in with both feet with their first V-8 engines. they also played around with turbines until around 83.

from what i remember about Lean Burn, in the beginning emissions was part of it, but that it was also for fuel economy, under cruise 52 plus degrees of timing along with an 18 to 1 AFR toasted a few engines.

i would have loved to have seen that car Grumpy, i bet it was a sight looking at all the components stuffed in it

funstick, both GM & Mopar had the same needs as did ford, emissions & fuel economy, none of them could do much about performance at the time.
GM came out on top on followed by Mopar & then a distant last was ford, IMO.

who ever it was at Mopar that came up with the bright ideal of putting a temp sensor in the power module to control the altenator,... a long hard beating isn't enough.

the journey is just as important as the destination

Apeiron, Mopar does suck, but they are still better than ford

*edit*, i'll look around & see if we have a OBD2 Mopar PCM laying around, if we do, i'll crack it open & take a pic of it.

 

That's like saying chlamydia is better than gonhorrea.

 

there is something to be learned from 30 yr old technology with an entirely different design and control startegy that applys to today ? thats like saying 8080 code would help a linux programer with a virtual windows application. there is nothing to be learned from that car. nothing of value. except for now not to do the job.

first of all someone above posted mopar payed back its loans. this is true.

secondly the reason the goverment bailed them out was due to the fact that at the time they were building the ahbrams tank.

third there were in fact differing needs. and although the end gola is to control spark and fuel the way in which they were controlled in both terms or cost and application dictated differing routes. mopar had large injector port fuel injection and ran batch fire before buck had the GN sequtnial stuff.

fourth over the time line chrysler has actually beaten everybody at the electronics game. the first fuzzy logic transsmission controller 1990. the first EFI system in the mid 50's co developed with bendix. the first electronic ignition standard on a vehicle. the first computer controlled carberator. the first practically appliable efi system. the list goes on. in fact they were the first domestic supplier to meet the can protocol which in reality makes reflashing simpler down the road.

they make fine equipment in terms of design. i will agree there durability track record is spotty but there control concepts and strategy are excelent. if they werent such difficult boxs to hack you would be whining about the superioirty of gm ecm;s. hell durability wise ford makes the Best ECM bar none. the coders are smoking crack but they have an excelent harware department. really a shame considering the fiasco TFI was.

in the end though almost any efi system can be made to perform flawlessly if you have the general control conecpts down enough to figure ou the algorythms.

 

LOL, you don't even bother trying to grasp what your reading do you?. It's a riot listening to you go on and on. Try rereading what I said, and think about it.

Yes, the government bailed them out, but the point was they generated their own woes do to executives making dumb decisions.

BTW, how's that from scratch code project you were working on going?, considering you'd claimed it would be ready for Beta Testing last Feb. you've been real quiet about it.

 

Okay

Mod intervention time.

Everyone zip it and play nice.

Lots of fascinating insite here on old and new systems so I WOULD LIKE TO KEEP THIS POST OPEN but I wont if things degrade.

thank you

 

I am curious too. I heard some "good" engineers were going to re-write the assembly in C in a few months. I think it would be interesting to read how things were interpreted.
So, is it ready yet??

I don't know much about the history of the ECM but this thread has a lot of info that might be useful in the "history of prom...." thread.

 

Then why not ask people to ask guestions about things they don't know much about rather then encouraging them to argue topics they're poorly informed of/on?.

If he'd been wanting info rather then wanting to argue, he'd have asked a question like, *when was it the mule car broke down?*. If he had, then I'd have told him in 1973. Which was long before the CCC systems came on line. At the time it was ground breaking technology. But, rather then ask a guestion, he tried taking it out of context to try and make some sort of point about his position.