Problem with OpenFlash Tablet please help *urgent*

[steve99]

Originally Posted by Target70

I thought he had both definitions(1.5 and 2.X) loaded in RR. I am almost completely unfamiliar with the software, but I thought he could choose which definition he was using out of the list of definitions that he had loaded in RR. He was just uncertain which one was selected:


quote: im not sure if it got crossed with the old one because when i use rom reader i left both of the definitions in the list and clicked on the 2.xx one before pressing apply


Or does even having more than one definition file loaded in RR cause the interference problem?

I only ever have 1 def in RR as its to easy to confuse oft 1.5 def oft 2 def and the original RR defs for a rom as the names are virtually identical. Always check some tables after you open a rom to see it looks ok.

I think RR picks the first def in the list unfortunatly It might be wrong one if you have multiple defs for same series rom

[Tromatic]

Word from OP? Do I smell BS?

[renfield90]

Originally Posted by Tromatic

Word from OP? Do I smell BS?

He bricked it and probably doesn't wanna talk about it.

[phrosty]

Originally Posted by renfield90

He bricked it and probably doesn't wanna talk about it.

More like he's trying to warranty it and is covering up that he's the one who messed it up

[Tromatic]

I'm curious if there is any data out there on "real" failed or bricked FR-S/BRZ ECU's. My first guess is the the dealer is not going to buy "Uh, it just happened on it's own" when even their ScanTool or Toyota tools do not connect.

[renfield90]

Originally Posted by Tromatic

I'm curious if there is any data out there on "real" failed or bricked FR-S/BRZ ECU's. My first guess is the the dealer is not going to buy "Uh, it just happened on it's own" when even their ScanTool or Toyota tools do not connect.

As an embedded systems engineer: not impossible, but really really really unlikely. We once received a batch of chips with a common manufacturing defect that could render the watchdog timer useless under certain conditions - THAT was fun to try and debug - so it's not inconceivable some sort of flaw in a chip could corrupt the non volatile memory area in an unrecoverable manner or that cosmic radiation comes through and damages the hardware somehow, flips a bit in memory at the wrong time, etc.

Only Toyota/Subaru would have that data, but the fact that I'm bringing up cosmic radiation should tell you how rare this kind of failure is.

IMO it's pretty lame there's no permanent bootloader on board. We consider that practically mandatory for a project that could require any kind of firmware update, precisely so that it's harder to brick our products.

[steve99]

Originally Posted by renfield90

As an embedded systems engineer: not impossible, but really really really unlikely. We once received a batch of chips with a common manufacturing defect that could render the watchdog timer useless under certain conditions - THAT was fun to try and debug - so it's not inconceivable some sort of flaw in a chip could corrupt the non volatile memory area in an unrecoverable manner or that cosmic radiation comes through and damages the hardware somehow, flips a bit in memory at the wrong time, etc.

Only Toyota/Subaru would have that data, but the fact that I'm bringing up cosmic radiation should tell you how rare this kind of failure is.

IMO it's pretty lame there's no permanent bootloader on board. We consider that practically mandatory for a project that could require any kind of firmware update, precisely so that it's harder to brick our products.

I believe it has a second bootloader area, the problem I believe is is you write rubbish code to the processor it goes off with fairies or loops, triggers watchdog which resets it runs same code again ect. Means you cannot talk to ecu via OBD port so your snookered unless you remove ecu and force it to boot mode via hardware strap, then rewrite user area

[renfield90]

Originally Posted by steve99

I believe it has a second bootloader area, the problem I believe is is you write rubbish code to the processor it goes off with fairies or loops, triggers watchdog which resets it runs same code again ect. Means you cannot talk to ecu via OBD port so your snookered unless you remove ecu and force it to boot mode via hardware strap, then rewrite user area

And the solution for that is simple, if you detect a watchdog reset in the bootloader you stay in the bootloader for a certain amount of time and broadcast an error message, so that the tool connected to the ECU can do something about it.

Ultimately it's a question of how much time/money do you spend developing the firmware vs. what actually adds value for the end user, so I understand them not building out an entire infrastructure around the fact that maybe a total of 10 people worldwide are going to mix up their definitions.

[steve99]

Originally Posted by renfield90

And the solution for that is simple, if you detect a watchdog reset in the bootloader you stay in the bootloader for a certain amount of time and broadcast an error message, so that the tool connected to the ECU can do something about it.

Ultimately it's a question of how much time/money do you spend developing the firmware vs. what actually adds value for the end user, so I understand them not building out an entire infrastructure around the fact that maybe a total of 10 people worldwide are going to mix up their definitions.

Agreed it would be better that way except that if the ecu encounted an issue in normal operation triggered watchdog you may not want the engine to "die" for a period of time just immediately reset and get going again But don't think toyota/subaru expect or want people to be flashing regularly with non standard devices using reverse engineered ecu tables. its a risk you must take if your going to flash/tune.

[Tromatic]

Did OP brick himself as well?

[Target70]

I don't know, but I was looking forward to a resolution. As OP has defaulted to morse code, I don't think we will find out. Just curious, does anyone know the price for a replacement ECU?

[Wayno]

3 kidneys

[Target70]

sorry, but my black market organ price conversion tables are out of date, could I get that in USD or at least crackerjack box tops?

[steve99]

Originally Posted by Target70

I don't know, but I was looking forward to a resolution. As OP has defaulted to morse code, I don't think we will find out. Just curious, does anyone know the price for a replacement ECU?

saw new euc for about $500 usa plus you need to get dealer to code it into dash/imobiliser ank keys probably another $100 dont know.

i was quoted $500 aussie from wreckers for secutity set from wrecked car ie dash/ecu/imobiliser and keys then fitt all those yourself.

his ecu is likely recoverable with low level boot mode but requires removal from car and connection directly to microcontroller in ecu via serial interface.

if i get a bricked one for cheap/free i will have a go at recovery, no one done one yet as far as i know