I was wondering what the main problem is with cracking the ECU. In no way, shape, or form, am I an expert on this. I just don't see why any of these expert electronics and tuners can not figure the ECU out. Then again, it doesn't seem like anybody that is working on the ECU that is having problems with it does not ever list the problems they are having.

Enough said about the problems.

What if there was a tool you guys could use to aquire data (all of the data) that the ECU is putting out.

Maybe I am just dreaming, but it seems to me if you have all the data that the ECU is receiving and sending that you would be able to reprogram the whole ECU, and it will do what you want it to do.

 

It's not even close to being that easy. First you have no idea what kinda language the ECU is using. Once ya figure that out you have no idea what all the data means. After you decipher what sensors are sending what values and what all they mean, you still have to get into the ECU. Even though Mazda can reflash it no problem, I can guarantee that the software is encrypted, and since 128bit encryption is not hard to do anymore, I am assuming thats what was used for security. Even 40 bit would take a long time to hack into.

It's been about a year that the car came out. If somoen strated working on it immediately, they might be done in another year or so. Maybe sooner, but I seriously doubt it.

 

On the other hand, all of the Japanese aftermarket tuners are just reflashing the ECU....I'm still wondering why we can't directly modify ours

 

Thanks for the info, I was just wondering what was taking so long, as I am such a novice in this area.

Like I said I am no expert at all when it comes to the ECU.

I do know of a guy who works for NASA that has an 8, and he has developed a data retriever for the 8. I let him hook it up to my car and you could read just about everything you wanted to.
Such as, fuel maps, exhaust gas temperatures, down to mpg, etc. I believe it is very similar to some sort of OBDII scanner. The only drawback is that you can not use his tool to reprogram, only to retrieve data.

Does anybody think this would help?
Or do you guys already have such tools?

Once again, So I don't get flamed, I was just wondering, and trying to see if this guys product would be helpful.

 

128bt encyption is a waste on a car ECU. It'll cost too much processor and memory to run it and the makers can ALOT about costs...

I think I like this tool. It only needs to be able to write as well and we're off!

 

hey, japan8, since you're in japan, do you know how the aftertuners there have managed to crack the code? seems to me that it'd be the same code, if so, why is the US so far behind in terms of decryption?

 

You've got me.

IS it really encrypted or is that just what people think?

Has anyone tried to crack the ECU system?

Why can RB acccess it? A partnership with Mazda? If so, why no warranty coverage for their parts like Mazdaspeed?

This is definitely something that I'd like to know more about...

 

Encryption [if it exists] is only part of the problem. The exact com protocols, and syntax are problems as well. What registers mean what. What are min/max values for each register, and on and on.

When someone "cracks the code" it does not mean they are inside the mainframe freely modifying all performance peramerters. Cracking any possible encryption will simply allow communication. After that, the obstacle of figuring out what to do will take just as long.

Support from the manufacturer or OEM's is the only realistic way to reflash successfully. IMHO. Not impossible, but a very intimidating challenge without the inside knowledge.

Good luck to anyone genuinly "hacking" at it.

 

Since it is a purpose built system I'm sure it’s written in micro code specific to what ever chips are in the ECU. In order to hack it you'd have to know the architecture of the chips and that’s proprietary. In time someone will figure it out I’m sure but don’t count on Mazda releasing something they’ve spent millions on to develop. Until then it’s replacement ECUs or piggies.

 

Just a note.....

One of my business vendors were trying to modify a Sony camera - a $150,000 camera. The camera [like Japan8 suggessts] has no encryption because it takes too much CPU time to go thru ENC/DEC cycles. Sony simply made it difficult to figure out what is going on.

They were communicating via RS-232, easy. No encryption to "crack". But they changed some registers outside of the allowable range and COMPLETELY disabled the camera. They killed a very expensive camera - it took about $15k to replace the card that handles system logic.

The point simply emphasizes the previous post.

 

When talking to Jim Mederer at Racing Beat, he showed his test/development bench that he'd built specifically to work with the ECU of the RX8. It is NOT a simple setup, nor an easy task to decipher the programming. Yes RB works with Mazda so they do have the codes, but Mazda did NOT tell RB what the codes necessarily mean or how they developed the maps. For obvious reasons this is a jealously guarded secret; no company wants to spend a fortune developing a system, then let their competitors just "borrow" it quickly and easily. Research is a very expensive and valuable commodity.

Jim said he's systematically testing each parameter to see what the codes do and what the reaction is on the car, and how it all interacts. He stated that when he understands what's going on and why, then he'll be able to begin making adjustments intelligently. He's quite a stickler for quality and professionalism. He said it's going to take awhile, and I believe him. He said he's already discovered there are several different maps used; one for idle, one for acceleration up to a certain rpm, another for acceleration above that rpm, another for deceleration, etc. He wants to make sure any changes he makes are truly beneficial without harming another parameter.

I have complete confidence that when he's finally ready to release an ECU mod, it will be a good one and worth it. I saw the data stream on his computer screen as he was showing the data coming off the ECU; not something I'd want to have to try and figure out.

So to answer the original questions, no it's not easy, and yes it's being done...but it's going to take time and you WANT the people making the changes to take the time to do it right. The ECU controls your car; your life and safety depend on it. You want any mods to be tested and reliable.

 

Thanks Ole Spiff,

For those who can't wait, there is always the fast but UGLY EVO's or WRX that come with factory FI.

I see the RX-8 as the cheap [1/3 cost] version of an NSX, which does everything well. Not mind blowing accelleration, but EXTREMELY well balanced in every aspect of the driving experience.

The EVO/WRX are maxxxxed out econo-boxes. The RX-8 is just scratching the surface of it's capabilitites.

I am VERY curious to see if RB will develop of participate in the development of a forced induction system that can run solely from the factory ECU. Like Ole Spiff pointed out, knowledge is pricey. RB won't give up any secret info unless there is a business oppurtunity to do so. There are MANY people that want a faster RX-8, but how many want to pay $$?

 

i know it's taking time here in the US....what i'm wondering is why it's taking MORE time here than in Japan

 

For the 2.0L Ford Escape, I have access to :
- the full strategy book which explains in pseudo-C language how the software in the ECU works,
- a full calibration 'doc' file containing all the programmable parameters (scalars, vectors and tables), their values and their memory adress,
and still I can't do anything with this info without an RCON or DCON (a Ford dedicated PC hardware) that enables communication with the ECU.

 

I have TWO flux capacitors running at 1.21GW each,and still can't get the ECU to do what I want.

That's why I have a piggy back. Not the best ultimate solution, but definitely the best available solution. Godd luck to those doing it the "right" way.

 

So does anybody think that this tool would help?

 

Which one? The TWO flux capacitors running at 1.21GW each???