All I hear is how the ecu in what is close to the cheapest car on the market is some advanced 64 bit computing engine that will shaft you on the power any mod you put on will produce.

Put me in my place, shove it in my face and rub it in, I want PROOF of this. Not just "A guy told me" or "A tech at the dealer told me" I want a toyota document, a picture of the processor with a silkscreen of it's origin and the manufacurers papers that state that model processor is 64 bit capable, anything to PROVE this.

I think it's BS, so prove me wrong, PLEASE.

BoogieQ, you always come up with some of the best things i read on SL. You don't just ask "ahhh what exhaust should I get?". You are determind to figure this little engine that hopefully can... and could, out aren't you. Good, we need more people out there who are willing to push the envolope to figure the truth out about them and what they can do. :clap:

But what make's you think that the ECU's are not what they say they are?
Obviously they are smart enough to learn, how we drive.

What do you think they are?
Cause i think we all have come to realize, that these ECU's are one of our biggest road blocks in overall modding for HP.

Im very intersted in hearing some of your thoughts on the why and whats of the ECU.

64-bit is hardly advanced. The first 64-bit CPU came out in 1991.

* 1991: MIPS Technologies produced the first 64-bit CPU, as the third revision of their MIPS RISC architecture, the R4000. The CPU was commercially available in 1991 and used in SGI graphics workstations starting with the Crimson, running the 64-bit version of the IRIX operating system.

* 1992: Digital Equipment Corporation introduced the DEC Alpha architecture which was born from the PRISM project.

* 1994: Intel announced plans for the 64-bit IA-64 architecture (jointly developed with HP) as a successor to its 32-bit IA-32 processors. A 1998-1999 launch date was targeted.

* 1995: Fujitsu-owned HAL Computer Systems launched workstations based on a 64-bit CPU, HAL's independently designed first generation SPARC64. IBM released 64-bit AS/400 systems, with the upgrade able to convert the operating system, database and applications.

* 1996: Sun and HP released their 64-bit processors, the UltraSPARC and the PA-8000. Sun Solaris, IRIX, and other variants of Unix continued to be common 64-bit operating systems.

* 1997: IBM released their RS64 full 64-bit PowerPC processors.

* 1998: IBM released their POWER3 full 64-bit PowerPC/POWER processors.

* 1999: Intel released the instruction set for the IA-64 architecture. First public disclosure of AMD's set of 64-bit extensions to IA-32 called x86-64.

* 2000: IBM shipped its first 64-bit mainframe, the zSeries z900, and its new z/OS operating system — culminating history's biggest 64-bit processor development investment and instantly wiping out 31-bit plug-compatible competitors Fujitsu/Amdahl and Hitachi. 64-bit Linux on zSeries followed almost immediately.

* 2001: Intel finally shipped its 64-bit processor line, now branded Itanium, targeting high-end servers. It fails to meet expectations due to the repeated delays getting IA-64 to market, and becomes a flop. Linux was the first operating system to run on the processor at its release.

* 2002: Intel introduced the Itanium 2 as a successor to the Itanium.

* 2003: AMD brought out its 64-bit Opteron and Athlon 64 processor lines. Apple also shipped 64-bit PowerPC chips courtesy of IBM, along with an update to its Mac OS X operating system. Several Linux distributions released with support for x86-64. Microsoft announced that it would create a version of its Windows operating system for the AMD chips. Intel maintained that its Itanium chips would remain its only 64-bit processors.

* 2004: Intel, reacting to the market success of AMD, admitted it had been developing a clone of the x86-64 extensions, which it calls EM64T. Updated versions of its Xeon and Pentium 4 processor families supporting the new instructions were shipped.

* 2005: In March, Intel announced that their first dual-core processors will ship in the second quarter 2005 with the release of the Pentium Extreme Edition 840 and the new Pentium D chips. Dual-core Itanium 2 processors will follow in the fourth quarter.

* 2005: On April 30, Microsoft publicly released Windows XP Professional x64 Edition for x86-64 processors.

* 2005: In May, AMD pre-released its dual-core desktop processor family called Athlon 64 X2. Athlon 64 X2 (Toledo) processors feature two cores with 1MB of L2 cache memory per core and consist of about 233.2 million transistors. They are 199 mm² large.

* 2005: In July, IBM announced its new dual-core 64-bit PowerPC 970MP (codenamed Antares).


So, if part of you argument is the fact that such a highly advanced CPU would not be fitted into such a conservative car, you're making an incorrect assumption.

Also, regarding the ECU, cant its programming be cracked? I know its been done with other car comps, why not this one?

Because it's a Toyota

Try adjusting your throttle cable till it's pulled 50% @ idle. I was trying to up my idle so it wouldn't stumble but after driving around for a day the idle went from 2500 back to 750 stumbling.
All I hear is how the ecu in what is close to the cheapest car on the market is some advanced 64 bit computing engine that will shaft you on the power any mod you put on will produce.

Put me in my place, shove it in my face and rub it in, I want PROOF of this. Not just "A guy told me" or "A tech at the dealer told me" I want a toyota document, a picture of the processor with a silkscreen of it's origin and the manufacurers papers that state that model processor is 64 bit capable, anything to PROVE this.

I think it's BS, so prove me wrong, PLEASE.

I can see some of boogie's argument, I mean the echo had a what? 8-bit cpu? he's not arguing they don't exist.. just why all that processing power.

however, its probably not that expensive given the advances in fabbing these days

I certainly hope that it's at least a 64bit processor on that bad boy.

With all the thousands, possibly millions of calculations it has to do with my crazy driving, I'd hate to think that I'm being backed up by a 32bit Playstation One.

OH Crap! Slick Roads, Must Stop! ...... loading......... Loading..... Please wait, Loading.....


:P

But what make's you think that the ECU's are not what they say they are?
Obviously they are smart enough to learn, how we drive.


Well yes, they 'learn' patterns and adjust tables to stay in optimum tune, however this is the first ECU I have come accross that is claimed to ruin power additions rather than optimize their programming to run well WITH the additions. It is possible that the ECU is running a very strict set of torque management tables that pull timing advance and retard the VVT-I to reduce the torque output.. but .. where's the proof? I'm hoping to find the answer once my obd-ii scanner comes in so I can log the ECU's functions in realtime when I add the boost.

64-bit is hardly advanced. The first 64-bit CPU came out in 1991.


Advanced for servers and home computing power, no. Advanced for a car management system???? You tell me, that's why I am asking. I'd like to see some proof that the processor(s)(yes, it can have multiples) are running 64 bit code.

Also, regarding the ECU, cant its programming be cracked? I know its been done with other car comps, why not this one?

Thats also what I am trying to find out. It SHOULD be able to be cracked. When you can access the program at that level, it opens unheard of potential to change EVERYTHING. I had this access on my GM cars thanks to guys that cracked it. I'm trying to get them to look at our market.

What do you think they are?
Cause i think we all have come to realize, that these ECU's are one of our biggest road blocks in overall modding for HP.

I would guess a network of 32bit processors at most, but hey, that's why I am looking for proof. Well, people 'blame' the ECU for robbing them of power but I have yet to see anyone post a diagnostic log from the ECU showing what is truely going on in it's 'brain'. I'm purchasing an OBD-ii scanner that's compatible with the Scion. See, it could be as simple as having the wrong plugs, or a restriction somewhere... a sensor that runs out of range and the ECU corrects in the wrong manner. Without watching the output logs of what the ECU sees after mods are installed, how can anyone blame it for the car running like crap?

Seems most think it isn't possible in the least for a mod to have no benefit at all, or in some cases, rob you of power. Yes, it IS possible for a mod to do nothing for you at all or hurt you in some cases... because of a bad design, or a small motor not needing a larger header, etc. Everyone is so quick to blame the ECU without looking at what the ECU is actually doing. Maybe it's trying to HELP but the mod installed is of no benefit at all.

However, I could be wrong and the ECU could have torque management so thick it will kill anything you try to do. It is VERY possible for the ECU to try and limit torque output by pulling ignition timing, etc. But without more details on this mythical creature, nobody knows for sure. (until they step up and give me some proof.. I'm waiting for someone to stick it to me! Come on!)

Hey I designed and built a 64 bit cpu back in college in 1978 - and I didn't even realize it until my Professor pointed it out to me.
From my 06xB driving I find that it likes slow throttle changes - nothing too fast - back off on the gas before you put the clutch in also helps for a much smoother shift. There are times when the engine is really responsive and times when it boggs down and I have to back off the gas. Usually if the revs are up over 2000 rpm and light throttle use it will take more and remain crisp. Some of the lag may be from the valve timing changes - can't be sure - maybe I will tap into the control wire with an LED run to the dash and see when it blinks to adjust the valve timing.

Try adjusting your throttle cable till it's pulled 50% @ idle. I was trying to up my idle so it wouldn't stumble but after driving around for a day the idle went from 2500 back to 750 stumbling.


Umm, if you held the throttle open anywhere near 50% you would be bouncing the rev limiter like a mad man in Neutral and barely be able to hold the car at a stop light with the breaks in an Auto. Not a very realistic comparison. Besides, for idle, ALL ecu's are programmed with a set Idle target. When the throttle plate is closed it uses an intake air controler to route the required air AROUND the throttle body to maintain the designated idle. If you prop'd your throttle plate open to give you 2500rpm at idle it's possible there is enough adjustment in the controller to close off the normal channel and atain the proper idle speed going off of your throttle plate opening.

Again, I'm not shooting down the idea of a 64bit ECU, it's very possible and probably within the realm of cost effective enough to use one.. just seems odd that this cheap little car would get such a boost in computing power from factory controled V6 supercharged engines still running 32bit wide programming and running just fine.

My guess is that it isn't 'smart' it's simply programmed for a whole hell of a lot of torque management. As an example, a 97'-03' Grand Prix GTP (3.8L Supercharged) has a torque managment program that pulls roughly 20% power on shifts and in other situations to prolong transmission life. When going to the 04+ models, same engine, new supercharger revision, same basic car, they had an ECU upgrade to a faster processor and issued some 45% reduction on shifts and in other areas to prolong transmission life. That doesn't mean the ECU is any smarter per se, but more strict in it's limitation of power through shifts, etc.

If this is the case in our ECU, we need to gain access to the programming to change it. This isn't possible at this time, obviously. Technology exists to perhaps make it possible, but I don't know if anyone is willing to try. Thus, we need to find out what the ECU can actually do to hinder us, and find a way around it, or at least know what our cars are doing FOR REAL instead of offhand speculation (which I am also guilty of as I have not logged realtime output of the ECU yet.)

I still find it funny that I have asked maybe 15 times on here of people that know a bunch and have turbo'd , etc , what the stock Timing values in the ECU are.. nobody knows... running boosted at an unknown timing level is just asking for hurt.

Ecu's also have tables in the managment section that if you start having detonation from too much boost, there is a time limit that the ECU has set in it as to how fast to start adding timing back until knock occurs again. This value can be in tenths of a second such as you hit a burst of knock, timing goes to 5, then up to 6 then to 7, etc as fast as the table allows. If it's set in longer incriments, you get the idea.

I think this year will mark some REAL GOOD progress for these cars, as well as the xA's. Hopefully we can get some good hard data into the community to benefit us all.

From my 06xB driving I find that it likes slow throttle changes - nothing too fast - back off on the gas before you put the clutch in also helps for a much smoother shift. There are times when the engine is really responsive and times when it boggs down and I have to back off the gas. Usually if the revs are up over 2000 rpm and light throttle use it will take more and remain crisp. Some of the lag may be from the valve timing changes - can't be sure - maybe I will tap into the control wire with an LED run to the dash and see when it blinks to adjust the valve timing.


Not to mention the engine is a turd with little HP and Torque. Because of it's low output and N/A setup it is going to show drastic changes according to weather. if you get a real low pressure front run through you will lose a good amount of power output. Weather plays a big role in how a car performs. On a cold night it will feel like you gained HP, on a hot midday afternoon, it will feel weak... and even weaker in bumper to bumper traffic on a hot low pressure day. *shrug*

As for the shifting, you can free rev it forever after you push the clutch in, let the rpms fall and drop the clutch at the right RPM level per speed and gear you are using and you won't feel a thing. It's all timing of the release to rpm's to speed being travled at, not the ECU.

Due to pittly power output, there is little difference between 50% and 100% throttle in any gear below 3,000-3,500rpm... however this is mostly opinion as well.

BoogieQ has a point- I've seen no real evidence that the ECU actually "learns" from my driving. OTOH, it seems to me that it's just doing what it's supposed to do- using the feedback from the various sensors to very closely regulate injection timing and duration, ignition timing and cam timing. If it builds an internal table from feedback/ experience for that stuff, then it's just more able to do it's job. If you want to to things differently, then it'll need to be reprogrammed or have some device to alter the feedback signals installed- the Greddy emanage does just that...

And it's quite possible for it to control the acceleration rate, for example, although there's no real proof that it does. I work on light rail vehicles with truly ancient 16 bit processors that do exactly that, and also control slipslide in both acceleration and dynamic braking... and that's with an 8086 processor... lots of other stuff simultaneously, as well.

Referencing the engine as a "Turd" is uncalled for, however. I rather suspect that the reason mods don't necessarily do much is that it's pretty highly tuned in the first place, and that the only way you'll get much more out of it (without serious work) is to cause the emissions to go to hell, something the computer won't allow...

Show me another normally aspirated 1500cc emissions legal engine with significantly more power, and we can talk, OK? Yeh, sure, wilder camshafts will give more high end power, but not without paying a price in driveability and low end torque... not to mention fuel economy.

Referencing the engine as a "Turd" is uncalled for, however.

LOL, I drive the same car dude!

While it's peppy for a 1.5L engine, it is just rather slow unless you spin it to high RPM's all the time. I'm used to fast cars, so it's taking some getting used to. By calling it a turd I wasn't saying it's a horrible motor, it's just, weak in comparison.. but for what it is, it's doing a good job.

It will never be a giant N/A monster unless it is built up to spin to 10,000rpm's with wild cams.

Bewst is the key, and the ECU might be the keymaster.. but back on topic.. I wana see the proof of the claims! :)

I would like to see proof too. I cant believe its taking so long for somebody to really take you to task for asking this.

I certainly hope that it's at least a 64bit processor on that bad boy.

With all the thousands, possibly millions of calculations it has to do with my crazy driving, I'd hate to think that I'm being backed up by a 32bit Playstation One.

OH Crap! Slick Roads, Must Stop! ...... loading......... Loading..... Please wait, Loading.....


:P


:rofl: :rofl: :rofl: :rofl: :rofl: I'd like to think we have more then a PS one in our cars. . . I have a scan tool two of them, one is a snapon scanner, it's old but it'l do generic obd2 and I have a new OTC scanner, what should I look for to find out if it's a 64 bit computer? or am I just getting in over my head? let me know maybe we can put this thing to bed. :pray:

hey how do you think I feel, I went from driving a car that spins over 11,500rpms on the street till my sponsors told me to stop driving it on the street or else type thing. oh and it did crank out what my profile message says and some. will be racing in the pro modified fwd class once I can pass safty inspection!!! need 5k in safty equipment. any body want to sponsor my race car!!! :money:

oh I will try and find out if it is 64bit, but I would not doubt that it isn't because a lot of the newer cars are going to start running 64bit and higher, plus the car is what people are calling obd3 the vehicle has the same technology that is in the new engine control systems just not labled obd3 yet.

A scan tool will probably not tell you what kind of code and processor power the unit is running off of.

You have to understand, a PS1 ran MANY more computations per second than a car ECU has to. Heck, they only needed 2 or 3 P4 processors in the VW Toureg that won the Darpa challange and that car processed GPS waypoints, Laser scan readouts, visual camera inputs, AND had to drive the car.

Comparing a gaming system or a desktop computer to a cars control systems is a bit overkill. The data bus systems are beefy in a car to receive constant inputs from various locations, but the computations are not as extreme. All it sees is "IF THIS, ADJUST THAT" in basic terms.

The main point of this post was to try and get some hard information about the ECU it'self, not so much to discount it and diss it's abilities. I still hope some good info comes to light that can help us move forward but I'm not holding my breath :P

Not to mention the engine is a turd with little HP and Torque. Because of it's low output and N/A setup it is going to show drastic changes according to weather.
Due to pittly power output, there is little difference between 50% and 100% throttle in any gear below 3,000-3,500rpm... however this is mostly opinion as well.

Hey if feels pretty good to me coming from a 1 liter 3 cylinder Geo with half the HP and 1000lbs less weight. You have to expect some less power output and no throttle response at low rpm as it is not taking much air in and with the size of the throttle body just moving the throttle 1/4 inch drops the vacuum to almost atmospheric at those RPMs. The speed limit around town is 25mph so all I need is first gear anyway. LOL and I don't have enough time to focus on the tach before it redlines in first gear even at half throttle - maybe you didn't break in your engine properly but I have trouble keeping my wheels from spinning without even popping the clutch.

I am looking through all my wiring diagrams and documentation, and have no proof either way. The ECU/ECM is only shown as a blue box with wiring going into it. There are no schematics of the processor(s) I am completely not surprised at this, but wanted to put this out there. I have every wiring diagram I could get from Toyota.

some one needs to take one apart to see what they find out

Although it is possible the computer is 64 bit, it seems unlikely. The real advantage of having a 64 bit processor is memory addressing not speed. I can't imagine that our car's computer would need 4GB+ of RAM to handle the engine management.

some one needs to take one apart to see what they find out

I love taking things apart to see how they work but I think I'll pass on this one. :nails:

i was thinking more in the lines of one from a junked scion

i was thinking more in the lines of one from a junked scion

He he he. Guess I jumped the gun on that one. :doh: :lalala:

Although it is possible the computer is 64 bit, it seems unlikely. The real advantage of having a 64 bit processor is memory addressing not speed. I can't imagine that our car's computer would need 4GB+ of RAM to handle the engine management.

I am glad someone else noticed this :)

Ahaaa but you forget about the stability control and traction control and ABS brakes - big instructions mean quick execution times and shorter programming routines. Doesn't mean that it HAS to address a lot of memory but the lookup tables for the fuel air mixtures are probably all multi-dimensional lookup tables and probably more than 8 bits in value. Would you like a slow computer running your air bag deployment?

If any computer is controling the airbag deployment, all it has to do is compare the value received from the accelerometers to determin if the value it has received is above the limit, in which case, execute deployment.

Even if it only runs at 10Mhz, that's 10,000,000 cyles per second it can go through. Some do more instructions per cycle than 1, some do 1 instruction per 4 cycles.

We have to remember, the code isn't all that big. The code that runs the ECU in a Pontiac Grand Prix with a supercharger is only 512K of code. Some of the pictures posted here contain more information than that. :P

Oh year and your timing for ignition and plug firing as well as valve timing. Consider this - the door clock control module is something like $485 so how much do you think is in the main CPU? It can be a RISC Reduced Instruction Set Chip and still be 64 bits for a few powerful instructions. Most of what it has to do is simple but they all add up and for resolution to a fraction of second for fuel injection duration it needs to be running fast enough to turn the injectors on and off in a fraction of a second. 6000rpm is 100 per second and it takes more than an off and on per revolution to operate the injectors and plugs - read all the sensors and control everything that needs to be controlled. If you have ever done any real time programming you would know that too slow a CPU doesn't cut it with time critical applications.

Our learning ecu can be proven with boosted guys, n/a guys, or anyone with mods. It doesn't hold back the motor from producing power, however with boost especially, it does hold its own problems. You can fully tune your piggy back, but the ecu will adjust itself within a few days and cause the motor to run too lean or rich. If you don't believe me, ask some of the boosted xb/xa guys here and find out for yourself. Be sure to ask someone thats actually gone through the tuning stages, most haven't gone to a dyno since most kits out there come with a pre-tuned management system, but yeah.

Does it cause power mods to not produce power? No. Is it really a 64 bit ecu? From what i understand it is, but i haven't seen it up close and person in terms of design. Is it a learning ecu that can cause issues in reguards of tuning and boost? Absolutely.

do not forget nintendo 64! just because somthing is 64 bit does not mean it is advance at all.... sheesh you guys are funny.

do not forget nintendo 64! just because somthing is 64 bit does not mean it is advance at all.... sheesh you guys are funny.

64bit means nothing in terms of advancement.

64 bit is the register size which has nothing to do with speed. It makes the processor capable of performing calculations on larger numbers and addressing more memory. Addressing more memory doesn't mean using more memory to process a particular thing it means literally storing an address to a particular piece of data.

Several of the comments made lead me to believe comparisons are being made to a PC's architecture. The car's computer is highly customized for it unlike the generic architecture of a PC. There are likely several different independent processing paths for different systems so that one doesn't interfere or impede the other.

the door clock control module is something like $485 so how much do you think is in the main CPU?

It's $485 because you will pay that much for a new one.

Our learning ecu can be proven with boosted guys, n/a guys, or anyone with mods. It doesn't hold back the motor from producing power, however with boost especially, it does hold its own problems. You can fully tune your piggy back, but the ecu will adjust itself within a few days and cause the motor to run too lean or rich. If you don't believe me, ask some of the boosted xb/xa guys here and find out for yourself. Be sure to ask someone thats actually gone through the tuning stages, most haven't gone to a dyno since most kits out there come with a pre-tuned management system, but yeah.

Does it cause power mods to not produce power? No. Is it really a 64 bit ecu? From what i understand it is, but i haven't seen it up close and person in terms of design. Is it a learning ecu that can cause issues in reguards of tuning and boost? Absolutely.

Sorry bro, but that doesn't 'prove' diddly squat. A couple guys ran boost and had less than ideal results.. why? Show me a log of real time ECU data to indicate IT is the problem and not a tuning mistake somewhere else in the line.

I'm not saying it ISN'T the culprit either.. but nobody has any evidence that it ACTUALLY is.

Oh year and your timing for ignition and plug firing as well as valve timing. Consider this - the door clock control module is something like $485 so how much do you think is in the main CPU? It can be a RISC Reduced Instruction Set Chip and still be 64 bits for a few powerful instructions. Most of what it has to do is simple but they all add up and for resolution to a fraction of second for fuel injection duration it needs to be running fast enough to turn the injectors on and off in a fraction of a second. 6000rpm is 100 per second and it takes more than an off and on per revolution to operate the injectors and plugs - read all the sensors and control everything that needs to be controlled. If you have ever done any real time programming you would know that too slow a CPU doesn't cut it with time critical applications.

So our car is the first car ever to require an advanced ECU to control functions that EFI cars have been doing for years, even on old OBD1 systems? Come on. 100 corrections per second even per injector is nothing for a dedicated processor. Again, this does not prove our ECU is a crazy machine set on world domination and claiming to be the ruler of all things power reduction.

I don't care if it's a 400000 bit processor, the real point I am trying to make with this thread is that many things around here are FACT yet nobody has ever shown a shred of proof that what they keep repeating IS fact.

All I am asking for is ONE little itty bitty shred of proof that this ECU is 64bit. Everyone claims it is.. so prove it to me! This isn't even about how fast it is, or how it limits power, it's about proof of claims that are just spewing out all over this place.

By the way, this is meant to spark interest and some investigation around here. "what muffler should I run" threads make me bored. :blah:

I'm trying to spark some conversation and maybe discover some truth for a change. :P

I certainly do not know everything nor will I ever claim to, but I CAN tell you I know enough to see a ton of BS and nobody seems to be able to smell it. :nails:

I just wanted to throw in what I learned from school. The only reason they told us that the ecu is so powerful is for safety. There are millions of lines of code in that little box. Basicly abs, airbags, traction control (for those that have it). abs can pump brakes up to 60 times a second. It takes a lot of movements, and at the right times to do these things Just lots of computing power for all the safety. Every ECU is also different. GM, FORD, DODGE all seam to use very simple ECU's. Nothing as complicated as the imports. Not near as powerful processors.

About learning how you drive??? I believe that is probably going to be mostly in an automatic. The ECU cotrolls the shifts in that. It remembers little things. Say all your driving around town never exceeds 35. But when you get on the highway you punch the throttle up to 60! The computer will remember that if you push the car past 35 it might hold the gear longer to see if you are "getting on the highway". Thats also what school said. As for robbing power??? I believe the ability is there. Why you ask. WARRENTY!! They don't want anything happening to there cars while under warrenty. The ECU will do everthing in its power to use the parimiters it was programmed with. Keeping the car EMISSION LEGAL! That is the power robber. They have to keep it emission legal also. SO.... you add lots of goodies. Takes it out of the emission legal stage. The computer will do its absolute best to keep it emission legal at all cost.



Go here to find out why there is such a large processor: http://auto.howstuffworks.com/car-computer.htm/printable

The "Door Lock Control Relay" is certainly more than jus a simple relay, and is obviously fairly complex. Here is the main circuit board (top and bottom)...

http://tijil.org/DLC_top.jpg

http://tijil.org/DLC_bottom.jpg

(That module does NOT include the remote entry receiver, which is a separate module.)

If something as "simple" and "inconsequential" as the "door lock control relay" looks like that, having a fairly impressive ECU "brain" wouldn't surprise me.

I have nothing that I can lay my hands on right now - I'll keep looking - but there was a Toyota "white paper" type of document discussing the design of the control systems for the latest generation tiny engine cars that discussed the multiple look-up tables that the ECU tries to match to driving style for both the engine and the auto transmission.

I doubt it is a 64 bit CPU, but it wouldn't surprise me if it was a lot more elaborate than we would think initially.

I haven't found the stuff I was looking for yet, but I did find info (from Green Hill Software) on the ECU in the Toyota Avalon, which has a 32 bit NEC CPU running a POSIX based OS and apps compiled in C/C++... :) From another Toyota paper it appears that at least since they started design of the ECU for the RAV4 in November, 1997 the 'standard' CPU for their ECUs has been a 32 bit processor. Still nothing specifically on the Scion engines...

I'll keep digging... I know I saw some of this stuff somewhere. :(

ahhh.. and the great resource finder known as Tomas has joined in the hunt.... he'll crack the bugger :P

That door lock relay is still not $485 complex... you have to admit that...

Sounds like your on your way to digging up the evidence that will either confirm this or deny it.

BTW, those that claim the ECU is killing your mod potential, are you all in CA with carb cars? The emissions comment got me thinking.

By the way, trying to say how many computations it takes for ABS and traction control modules is not going to amount to anything. 60 times a second to a half decent processing system is a fart in the wind.. multiply that by multiple systems and it's still not going to kill it. And again, cars from the early 90's had ABS systems, multiple airbags, traction control systems, remote keyless entry systems, digital climate controls, heads up displays... shall I go on? That was 15 years ago. The Scion ECU is not the first to be able to do all this stuff. While it's processing power is exponentially better that doesn't mean it's the first to be able to monitor and control all of these systems.

Anyway, keep up the hunt! :)

(1) No one suggested the "Door Lock Control Relay" was a $485 unit - it was just being used as an example of how much more complex than expected some of the little 'black boxes' are. (Actually it's a blue box and $425 retail... :) )

(2) All Scions are "50 state."

(3) The most recent info I have found on Toyota ECU CPUs seems to indicate that they are 32 bit, 200Mhz units... (That is still not Scion specific, but the hardware is likely the same, even if the software is model specific.)

Still looking, but not that hard... :)

__________________

Things like the Door Lock Control Relay are insanely expensive for several reasons: The cost of development/design/manufacture is spread over a relatively small number of units, the warehousing of a few million spare parts for hundreds of car models a decade at a time is, uh, "complex and expensive," and finally, because it is a closed market - they can get away with charging that much - you can't run down to RadioShack and get a cheap replacement - you MUST use their unit, or an even more expensive custom-made unit. :D

Even cell phones are more complex, but cheaper. This is because the actual transceiver is a commodity circuit pack incorporated into many millions of phones of different models, and even the non-common circuits and cases are often done up in millions of units. Spreading a couple million dollars on development and another million dollars on production tooling/setup over millions of units affects costs less than trying to pay for that over a hundred thousand units and spares.

Also, with most consumer electronics, like cellphones, there is no massive backstock of spare parts required by law (for models sold in the US, the manufacturer must maintain adequate spares to support any given model for a minimum of seven years).

Anyway, there are many legit reasons why the actual cost of things like car black boxes appears higher than some "equivalent" other consumer electronics, plus the eventual retail price of those automotive units can be marked up even more just "because they can." :D

This is a good thread, but comparing costs of electronics seems to be getting us off topic. Even a 16 bit ECU could do most of the things we're discussing.

What I want to know is how we can assume that this ECU can limit torque to a certain level (regardless of the # of BITS that it's capable of running.)

I mean, I guess the ECU could always run the injectors at a given duty cycle and always pull timing at a certain RPM, but suppose you change mechanical things that aren't monitored or controlled by the computer? Would you say that there are accelerometers and timers inside that are saying "Whoa...that 1/4 miles was just WAY too fast...I'm going to cut the power back a little."

Like what if you increase fuel pressure somehow or change injector size? Then the injector is still sending more fuel, despite the limit on pulse width. Then the 02 sensor sees a richer mixture and modifies the pulse width some more. How much can it really compensate, though, if you get more air AND more fuel in so the O2 still sees the ratio it's expecting? Remember that an O2 sensor only reads the mixture, not the volume of gas.

What if you change air delivery somehow (larger TB, bigger/more/shorter/longer intake runners/FI etc)? Sure, the MAF can read that and say "you're getting too much air...I better do something." But really, how much can it do? It can't mechanically close the throttle (I don't think scions will ever get "drive by wire")...it's just referring to different maps. What if you translated the MAF signals into what the ECU is expecting even if you're passing more air? The turbo Buick and other 'muscle car' guys do this all the time.

Even if this thing 'learns' how to kill power, there has to be ways around it by making systems more efficient and 'lying' to other systems in the ECU.

Also, don't forget that things like excessive noise from a header could be inaccurately translated as "knock" and a signal would be sent to pull timing. Things like this could result in power loss from a mod.

Keep the info coming...I've got a hundred more questions!

This is a good thread, but comparing costs of electronics seems to be getting us off topic.

Except for the small factoid that the originator of this topic is the one who brought up price, and that makes it 'on topic.' :D

^^^ Thanks for pointing out that technicality! :bow:

I suggest this, if you BoogieQ don't believe the ecu has its issues in reguarding boost, you prove to everyone else it isn't. Yes, i'm calling you out. Learning ecu's have huge limitations compared to other ecu's and programs out there, you started this thread asking people to prove it is the problem. Now i start the question, how isn't it the problem? And if you don't believe it is the problem, what problems do you see causing past issues with our 1nz? Yes, for a while there people either had bad luck with tuning, or believed what xxx said about tuning, and ended up being a very poor, costly tune. Thats a unavoidable issue being that it happens everywhere, not just scions/toyota, everyone. N/a too.

I'm not saying boosted 1nz's can't make power, any boosted motor can. But the limit that we are able to succeed without issues caused by the ecu, as well as a few others, are strict.

If the ecu uses table driven code to control the injection then the table can limit the amount of fuel that it can inject at a certain rpm because the table is a realistic size and would not have entries for more fuel injection at levels that it should not normally operate at i.e. 20lbs boost so yes the ECU could limit power output. If you required more fuel than the table allows then when that limit was reached the mixture might lean out then the oxygen sensor might trip from too lean a mixture and all sorts of things could / might happen. I was told by someone that knows a lot about these engines that when they get a little too warm the fuel mixture is enriched to cool down the engine. Imagine that!

As far as ABS brakes go on the xB at least it is controlled more by the computer than in prior older cars because the stability control operates the brakes to each wheel independant of the brake pedal when it needs to - a lot more complicated than OLD cars ABS systems. I am not trying to prove that a 64 bit is being used just that it could be very likely. Now show me a photo of the board that it is on and I could probably figure out what it was and I am pretty sure that it would NOT be some really rare chip but likely a readily available chip that has a low cost and is easy to program with lots of high level language software support from the industry.

Jan06xB's post makes a good point about the tables, but it goes back to my question: If there is a realistic limit to the # of tables or correction multiplyers, then a separate mechanical means of getting more fuel to the engine (ie - additional injectors with separate controller and maybe even pump/regulator) is possible while still showing the O2 sensors an acceptable ratio.

And I think we'd all agree that nearly all EFI NA cars are limited somewhat by the ECU when adding forced induction. Weren't some of the boosted cars that had problems on this board running stand-alone efi, anyway?

Is the point of this thread only to find out if the computer is 64 bit or is it to find out what limitations it can cause???

Regarding modern ABS, most newer cars use a separate ABS computer. Are we even sure the Scion ABS is controlled by the engine ECU?

Maybe Toyota is just THAT nice! :rofl:

Actually the Engine Control Unit seems to be directly involved with managing the ABS/VSC/TRAC systems, but gets some help from an aux controller for the VSC. Here are the basic schematics:
Engine Control PDF (http://tijil.org/holdfile/xB%20Repair%20Manuals/04%20xB%20Wiring%20Manual/H_systemcircuits/EngineControl.pdf)

About the forced induction cars being limited by the lookup tables (maps) available to their ECUs, the fresh-from-the-factory ECU in the tC does NOT have tables that allow it to handle forced induction situations without being 'reflashed' to include a new set of tables that give extended instructions. Here's a discussion from the tC forums here...

http://www.scionlife.com/forums/viewtopic.php?t=98810
and
http://www.scionlife.com/forums/viewtopic.php?t=118496

Yeah, I know, that's about the tC, but it is still interesting info...

Well sombody asked for these,so i took it upon myself to tear the ECU out, i don't know if they will help. Sorry the pics aren't that great, junky digital camera and definatley not Tomas quality pics. But i tried :doh:

http://i24.photobucket.com/albums/c8/proshopxb/103_0733.jpg
http://i24.photobucket.com/albums/c8/proshopxb/103_0764.jpg
http://i24.photobucket.com/albums/c8/proshopxb/103_0737.jpg
http://i24.photobucket.com/albums/c8/proshopxb/103_0751.jpg
http://i24.photobucket.com/albums/c8/proshopxb/103_0752.jpg

I'm interestd in this, too.

I can verify the new Yaris uses a 32-bit ECU, but can't find anything for the xB. Shouldn't this be in the service manual?

OK, that gives us a reasonable look at the contents of that 'black box.' :)

I see it *IS* one of the units made by Fujitsu-Ten, so I'll toss out some info I've been going through from them. The reason I didn't toss this in the pot before, is it is not specifically the units used with our engine, but with other Toyota engines from the same design era. Since they are now confirmed to be all Fujitsu-Ten units, this should help, even though they are not the identical module.

http://tijil.org/16-4.pdf

http://tijil.org/26-1.pdf

Lastly, here's a single slide from a PowerPoint presentation given by some engineers from Toyota at the American Control Conference, 09JUN2005. It's a 'generic' Toyota ECU...

http://tijil.org/Generic_Toyota_ECU.gif

Enjoy, folks! :D

from Geek.com
n 1981 NASA sent up the first space shuttle, which used Intel 8086 processors for a host of diagnostic equipment. More than 20 years on these chips are still being used to make sure the shuttle's twin booster rockets are safe for blast-off,

from the Wiki on the 8086


<i>
The 8086 is a 16-bit microprocessor chip designed by Intel in 1978, which gave rise to the x86 architecture. The processor runs at clock speeds between 4.77 (as with the 8088 version used in the original IBM PC) and 10 MHz.
[/i]

So the shuttle is still using a 16 bit processor, that runs at 10 mhz. Realize that the average desktop packs almost 100 times that.

I I think 16 or 32 is all our little box really needs

I forgot to mention the ECU in the pic's is from an 06, supposedly the pin confirguration from 03-05 has changed in the 06's, not sure if the ECU has changed it self inside. Just thought i would throw that out.

so it seems like ECU is actually 32-bit?

you know I've been wondering for quite some time now... where exactly did anyone get the 64-bit ECU from anyway?

was it just something that was said in some post somewhere?

Jan06xB's post makes a good point about the tables, but it goes back to my question: If there is a realistic limit to the # of tables or correction multiplyers, then a separate mechanical means of getting more fuel to the engine (ie - additional injectors with separate controller and maybe even pump/regulator) is possible while still showing the O2 sensors an acceptable ratio.

Your idea to boost the fuel pressure or bigger injectors would work within limits of the table adjustments that the ecu could make to lean out the air fuel mixture at low power levels to make it correct with the increased fuel per unit of injector on time. If it could not compensate enough to get the mixture correct at the O2 sensor then it could throw an error code.

There are people out there that can hack into the code but it will be expensive.

It's interesting that you can "chip" some cars. My '97 Mustang even has a port on it's ecu that you can plug right into.

But I've never seen anything for a Toyota.

so it seems like ECU is actually 32-bit?

you know I've been wondering for quite some time now... where exactly did anyone get the 64-bit ECU from anyway?

was it just something that was said in some post somewhere?

Sorry I haven't been around. All the info posted after my last post has been awesome! I like the research and fact posting!

To the quote above, i would guess someone said it at one point and the community just took it as gospel. As much as we all like, not everything posted on a board is factual or accurate. I am sure in some of my ramblings I said something I thought to be true that wasn't. Heck, even typo's can cause it to happen.

Props to you all for digging into this.. I would be more likely to say 32 at this point. The processing prowess of this ECU isn't what I really was after.. everyone here (at least that was vocal about it) was blasting "64bit ecu" in everyones face and convinced they were right but nobody knew why or where that information came from.

Sorry to bring this back to the top after so long, just wanted to add my thoughts.. I think this is 99% case closed.. the 1% from something not stating XB and 32 bit in the same diagram :P

Continue on...

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

It means the creater of this topic finally said what many of us were thinking in the fact that these motors can do much more than we think. It's time for everyone to stop being so negative around here and realise you can do great things given enough time.

From everything I've see on current Toyota ECUs, ours should be a 32 bit running at about 200 MHz.

The only way I'd change my opinion on that is if someone comes up with specific information from Toyota, Scion, or Fujutsu-Ten about OUR ECU that contradicts that.

Closed. :D

Id still like to know why it cant be chipped like a mustang or something. Is the programming too complex or what?

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

It means the creater of this topic finally said what many of us were thinking in the fact that these motors can do much more than we think. It's time for everyone to stop being so negative around here and realise you can do great things given enough time.yeah years of time and thousands upon thousands of dollars. not worth it when many other motors 32bit equipped or not can be modded MUCH easier to a higher extent.....

Id still like to know why it cant be chipped like a mustang or something. Is the programming too complex or what?

Because for warranty and federal law compliance, Toyota does not WANT us to be able to mess with the 'puter. :D

There is absolutely no advantage to the manufacturer in allowing that on an econo-car, and tremendous advantage (fewer problems) in not allowing it.

Tom

I was told that the reason the 06 has different pin configurations is because it uses a different computer. Also that there are some more advancments. everything came from a technician from toyota that has been with toyota for over 20 years. He stated that the ECU being used in many of the early 06 models is going to be the next OBD going to be OBD III. They are still working on it but the ecu was released as experimental and when you bring your car in they can hook up to the datalogging of the ECU and download the data. He said when he ran mine that he new I poped a CEL for MAF and everything even with me resetting the computer.

So if anyone knows more about this please shared!

The part of the whole discussion that I have a problem with is the contention that the ecu can learn to compensate for piggyback controllers. I see that as impossible. The ecu only knows what the sensors tell it, and what's built in to its internal mapping. If the piggyback alters the sensor information on the way to the ecu, then the ecu has no way of determining that, unless the information falls outside the mapped parameters, in which case it'll throw an error code... interpreting the information as a sensor problem...

And there's really no point to the ecu limiting power, only emissions. If mods actually increase the efficiency of the system, then the wot max performance mapping in the ecu will give more power, with the feedback from the sensors not being used to alter the settings... If the intake and exhaust systems actually allow better airflow thru the engine, then the ecu will compensate by adding more fuel to achieve the proper mixture. If that doesn't work, then it'll back off on the timing to prevent knock...

The relatively high output of our little motors indicates that the whole deal is very much tuned for performance, anyway, which seems to me to be the reason it's tough to get much more...

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

It means the creater of this topic finally said what many of us were thinking in the fact that these motors can do much more than we think. It's time for everyone to stop being so negative around here and realise you can do great things given enough time.yeah years of time and thousands upon thousands of dollars. not worth it when many other motors 32bit equipped or not can be modded MUCH easier to a higher extent.....


Just know there will ALWAYS be the bigger more powerful motor that's easier to make power with. Some people like to modify things simply for the accomplishment and not always to be the end all be all to street rods. Please stop crapping in our butter.

I think ZPI hitting 180whp on a turbo without an emanage and letting the ECU control the motor defunks this 'power robbing super neural net super computer ' ecu myth. You might say tons have failed before and ran into issues... perhaps ZPI knows what they are doing :P

Id still like to know why it cant be chipped like a mustang or something. Is the programming too complex or what?

It can be chipped. It's commonly referred to as ROM tuning.

Technosqure/Techtom equipment can be had to re-flash the ROM.

Trouble is finding someone with the hardward and software to do so. It's expensive to own. Upwards of $5000-10,000 for the license and hardware. Most people pay from $500-1000 to have their ECU flashed.

Best Regards,

I was told that the reason the 06 has different pin configurations is because it uses a different computer. Also that there are some more advancments. everything came from a technician from toyota that has been with toyota for over 20 years. He stated that the ECU being used in many of the early 06 models is going to be the next OBD going to be OBD III. They are still working on it but the ecu was released as experimental and when you bring your car in they can hook up to the datalogging of the ECU and download the data. He said when he ran mine that he new I poped a CEL for MAF and everything even with me resetting the computer.

So if anyone knows more about this please shared!obd II data can be looked up and is stored just like you said that is nothing new at all.

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

It means the creater of this topic finally said what many of us were thinking in the fact that these motors can do much more than we think. It's time for everyone to stop being so negative around here and realise you can do great things given enough time.yeah years of time and thousands upon thousands of dollars. not worth it when many other motors 32bit equipped or not can be modded MUCH easier to a higher extent.....


Just know there will ALWAYS be the bigger more powerful motor that's easier to make power with. Some people like to modify things simply for the accomplishment and not always to be the end all be all to street rods. Please stop crapping in our butter.

I think ZPI hitting 180whp on a turbo without an emanage and letting the ECU control the motor defunks this 'power robbing super neural net super computer ' ecu myth. You might say tons have failed before and ran into issues... perhaps ZPI knows what they are doing :P

and a lot of people don't agree with or believe zpi and or trust their results.

it data logs all of the fuel, and what throttle % the a/f ratio is, kind of like a AEM with datalogging, how you can replay everything. that is how he explained it to me, it is so they can make the tuning on the motor better, and possibly make more power and MPG.

not so sure about all that he's saying but car's have had "black boxes" since about 2000 and can only be used legally in drunk driving/serious at fault accidents

And there's really no point to the ecu limiting power, only emissions.


+1 and correct. This is the ONLY thing the ECU cares about. Emissions!

This is NOT the first car to have been discussed concerning "learning ECU". For several years we "battled" this on the 350Z.

In the end. No problem. The ECU strictly cares about Emission guidlines. Not power. A few hours on the dyno and with Techtom reflash, instant power. Yet, keeping emissions happy.

not so sure about all that he's saying but car's have had "black boxes" since about 2000 and can only be used legally in drunk driving/serious at fault accidents

Prove it please.

So assuming th ECU is actualy 32 bit, where does that get us? Regardless, doesn't the ECU still compinsate for mods or is that a farse now?

It means the creater of this topic finally said what many of us were thinking in the fact that these motors can do much more than we think. It's time for everyone to stop being so negative around here and realise you can do great things given enough time.yeah years of time and thousands upon thousands of dollars. not worth it when many other motors 32bit equipped or not can be modded MUCH easier to a higher extent.....


Just know there will ALWAYS be the bigger more powerful motor that's easier to make power with. Some people like to modify things simply for the accomplishment and not always to be the end all be all to street rods. Please stop crapping in our butter.

I think ZPI hitting 180whp on a turbo without an emanage and letting the ECU control the motor defunks this 'power robbing super neural net super computer ' ecu myth. You might say tons have failed before and ran into issues... perhaps ZPI knows what they are doing :P

and a lot of people don't agree with or believe zpi and or trust their results.

Why? Please don't tell me it's because at 162.78WHP the #3 rod should break at the wrist pin.

why don't you ever prove your points? you make me find things eh ?

i'll try and dig it all up to make a stupid point ON THE INTERNET!

not so sure about all that he's saying but car's have had "black boxes" since about 2000 and can only be used legally in drunk driving/serious at fault accidents

Prove it please.


These "black boxes" are nothing but your OEM ECU recording the last few seconds before catestrophic failure. Such as downshifting from 5th to 2nd. Causing generous RPM. ala 2000 celica GTS , 2zz. You can see the jump in RPM before destruction if you go back and look at the ECU. Not really going to do much more.


Certainly if you are going 150mph right before an accident.... they might question the recorded speed. :rofl:

I know the ECU does do data logging. it also stores powertrain codes, etc for later retrieval. This isn't anything new.. but here, again, we have a statement about 'black boxes' as if they record audio in the cabin and GPS waypoints you pass so they can spy on you.

*sigh*

but here, again, we have a statement about 'black boxes' as if they record audio in the cabin and GPS waypoints you pass so they can spy on you.

*sigh*

:rofl:

*insert spy hunter theme*

:lalala:

Speaking of in cabin spying... That is exactly why I will never own a vehicle with Onstar or any other similar system. The feds were already called on the carpet once by the supreme court for using the Onstar system to tap into vehicles and listen in. The only reason they got their wrist slapped is because when tapping into the system it would disable the use for the customer. If I remember correctly they were told as long as they could tap in without interrupting service what they were doing was ok.

I am not paranoid nor do I partake in illegal activities. I just don't like the idea of big brother having the ability to track my every movement and word when I am in my car. Not that I think they would ever have any reason to do so.

the ecu black box doesnt record audio it doest have gps points but it does record vehicle speed , rpm's , fuel maps , yaw , roll , abs inputs. just about all info from all sensors.