[Throtex]

Here's something I was wondering... I'm no gearhead but I'll hack at any digital circuit you like. Has anyone bothered putting out a chip for the TLS? If not, is there a good reason for this (i.e. no substantial gains)?

I'm saying this because I've been having some trouble finding a 'worthwhile' summer job since I can't really come back to work for any company (grad school thang) and I wouldn't mind commiting some time to hacking at an ECU if I could figure out exactly what the heck I'm tuning for. Anyone have any ideas on this?

If it seems like something that might be worthwhile, is there any way I can buy just an ECU for our cars somewhere? (for simulation purposes)

 

[rominl]

so far there is no hacked chip for the tls yet. i think a month or two ago, some shop here in the bay area posted on the a-tl forum saying that they are making a new ecu for our cars, how it's gonne be good, blah blah blah,,,,,,, and dyno is two days, blah blah blah........

but i don't think i have heard from them anymore...... the guy's name is leo that's all i could recall.

and from what i know, no one really knows what the heck is in the ecu and how it's programmed and how it works....

 

[Throtex]

so far there is no hacked chip for the tls yet. i think a month or two ago, some shop here in the bay area posted on the a-tl forum saying that they are making a new ecu for our cars, how it's gonne be good, blah blah blah,,,,,,, and dyno is two days, blah blah blah........

but i don't think i have heard from them anymore...... the guy's name is leo that's all i could recall.

and from what i know, no one really knows what the heck is in the ecu and how it's programmed and how it works....

Hmmm... so I guess if the chip itself is affordable I could play around with it Might help to have some pinouts of the sucker, but I imagine I'd have a tough time getting even that provided to me. Don't really wanna work by trial and error figuring out which pins are power and ground just to start it up

 

[rominl]

Hmmm... so I guess if the chip itself is affordable I could play around with it Might help to have some pinouts of the sucker, but I imagine I'd have a tough time getting even that provided to me. Don't really wanna work by trial and error figuring out which pins are power and ground just to start it up

yeah without the basic pin out list you are pretty much screwed. i think each ecu isn't cheap, like 200 or 300 bucks? if you don't know about the vdd and gnd it would be easy to fuck out the chip badly.... it'd be interesting to know who could hold those information though

 

[Throtex]

Actually, I just ran into this site:
http://www.hondata.com/products.html

Some pretty interesting stuff, haven't fished through it much, but if they cracked the system for the cars they've got on their site, I really don't think it's changed much since then.

Unfortunately, I'm thinking it probably wouldn't be all that worthwhile for our cars. Like I said, I'm no gearhead, but ECU tuning applications seem like they'd make the most sense in a boost control application when you want to avoid piggyback controllers... so until someone makes a turbocharger for our cars it might not even be worth the trouble

 

[rominl]

Actually, I just ran into this site:
http://www.hondata.com/products.html

Some pretty interesting stuff, haven't fished through it much, but if they cracked the system for the cars they've got on their site, I really don't think it's changed much since then.

Unfortunately, I'm thinking it probably wouldn't be all that worthwhile for our cars. Like I said, I'm no gearhead, but ECU tuning applications seem like they'd make the most sense in a boost control application when you want to avoid piggyback controllers... so until someone makes a turbocharger for our cars it might not even be worth the trouble

that's some intereesting read. i don't remember now but i think i heard that the ecu in our cars are better designed than that say on the audi a4, that's why it's so much harder for people to hack our chips. and above all, you may need some special programmer to re-program the chip as well...

 

[Throtex]

that's some intereesting read. i don't remember now but i think i heard that the ecu in our cars are better designed than that say on the audi a4, that's why it's so much harder for people to hack our chips. and above all, you may need some special programmer to re-program the chip as well...

I wasn't thinking re-programming so much as taking an FPGA and having it mimic the OEM ECU... then allow some external device (such as an in-car controller) tell the FPGA what to change... don't see why there can't be on-the-fly "reprogramming".

Taking the current ECU and reprogramming it would probably be a pain. And, I don't see why they would have even put it on reprogrammable silicon... it'd be uselessly expensive if they did. It's probably not erasable.

The thing is, if I had the specs for the ECU (pinouts, expected data, replies... etc... just general protocol information) it'd be easy as hell to mimic the OEM chip on an entirely different chip. Blah, I don't think it'll happen though... like someone's just gonna lay all that info out on my lap

 

[Throtex]

On a side note, I could just piggy-back a logic analyzer on the OEM chip while it's running in the car and try to figure out what's going on from there, but damned if that wouldn't be a pain in the ass (interpreting the signals, I mean, not actually hooking up the analyzer).

I'd still need to know at least where Vcc and Gnd are though

 

[rominl]

On a side note, I could just piggy-back a logic analyzer on the OEM chip while it's running in the car and try to figure out what's going on from there, but damned if that wouldn't be a pain in the ass (interpreting the signals, I mean, not actually hooking up the analyzer).

I'd still need to know at least where Vcc and Gnd are though

hahah wait, didn't you said you got tons of time during summer? do you already know how many pins there are on the ecu? if you really go through the hard way i think it's not too hard to figure out the vdd and gnd (no real solution yet) but you are damn right that using analyzer to interpret the other signals is really tough...

 

[Throtex]

hahah wait, didn't you said you got tons of time during summer? do you already know how many pins there are on the ecu? if you really go through the hard way i think it's not too hard to figure out the vdd and gnd (no real solution yet) but you are damn right that using analyzer to interpret the other signals is really tough...

I've used an analyzer to debug a Pentium circuit... don't talk to me about a lot of pins And actually, if it's a large package, there's probably several Vdd's and Gnd's I need to hook up.... hrrrm.

 

[Throtex]

Actually, the more pins the better... less likely that a pin is being used for serial I/O (which would be annoying to try to figure out... there's tons of serial protocols, so unless they're using I2C, SCI, or SPI, I wouldn't be able to identify it off the bat). Anybody here ever work at Acura and would like to provide us with some documents on the ECU?

 

[rominl]

Actually, the more pins the better... less likely that a pin is being used for serial I/O (which would be annoying to try to figure out... there's tons of serial protocols, so unless they're using I2C, SCI, or SPI, I wouldn't be able to identify it off the bat). Anybody here ever work at Acura and would like to provide us with some documents on the ECU?

darn, now it's getting out of my league

 

[Throtex]

Don't mind me, I'm just thinking out loud

 

[baggsTLS]

I just read that link on atl.com. It was Weapon*R in the bay area. Leo was trying to accomplish the piggyback ECU mod back in January.....I guess there was no proof in his pudding.

 

[Bebber]

Practically speaking, I don't think you'll have enough info to recreate the whole ECU on another chip. That's a lot of work involved there.

Oh, and if you do go the FPGA approach. You might try Altera. There stuff works pretty well, even if they did rescind my signed job offer the day before my graduation due to "economic hardship". Who, me bitter?

I work for Arinc now.

 

[Throtex]

Practically speaking, I don't think you'll have enough info to recreate the whole ECU on another chip. That's a lot of work involved there.

Oh, and if you do go the FPGA approach. You might try Altera. There stuff works pretty well, even if they did rescind my signed job offer the day before my graduation due to "economic hardship". Who, me bitter?

I work for Arinc now.

I've worked with Xilinx FPGAs before... but I guess it's quite irrelevant if you can write VHDL generically enough

 

[rominl]

Practically speaking, I don't think you'll have enough info to recreate the whole ECU on another chip. That's a lot of work involved there.

Oh, and if you do go the FPGA approach. You might try Altera. There stuff works pretty well, even if they did rescind my signed job offer the day before my graduation due to "economic hardship". Who, me bitter?

I work for Arinc now.

hey congrats josh!!!!!! this is awesome! i was on the same boat. altera retracted my offer, though i didn't go for them anyway

 

[Bebber]

hey congrats josh!!!!!! this is awesome! i was on the same boat. altera retracted my offer, though i didn't go for them anyway

You too man! That sucks. Altera seemed like a really nice company to me and all was going so well until, well you know. I even turned down my Intel Microprocesor design offer for Altera, little did I know that Altera would rescind my offer only after Intel filled the position with another candidate.

I wonder what the future holds for Altera? I was the chair/president of IEEE at USC at the time, and I made it a point to publicize to my friends that certain companies can pull some very dirty tricks on us engineers. I have a feeling that no USC student will take Altera's word with much confidence in the near future.