SSC40D - They Won't Give My Car Back

The EPROM is probably soldered right to the ECM circuit board. They don't use very many "chip in socket" designs in the auto industry. The vibration problem just doesn't allow it.

 

i don't think it would be too hard to engineer a "chip in socket" type with a clip to secure it in place. i do see your point though.

-drew

 

Socketed parts are not as reliable as soldered parts unless one uses a relatively expensive socket. And, if the EEPROM were socketed it would then be exposed to handling damage.

The in-situ update should work well, except it appears that Toyota's update documentation either does not require or does not emphasize the importance of putting a maintenance charger on the 12V battery to keep it from draining. It's easy to fix a little oversight like that.

 

The memory you are speaking of is located within the CPU itself. The technology being used is referred to as "flash ROM' in the automotive industry. There are no EPROMs to replace, as was done a decade ago. The CPU itself contains RAM, Flash ROM, and often EEPROM for diagnostic codes and other calibration information. To reprogram the CPU, the current program must be complete and match several checksum tests, validating its integrity. During reprogramming, the CPU runs a "download and execute" program, which unlocks the CPU flash ROM and provides instructions to reprogram itself. If the programming is not completed successfully and completely, when the CPU reboots a programming error is identified. This results because either an incomplete program is present (most likely), resulting in the inability to properly execute code that would allow for reprogramming again, or a situation where enough code is present to execute some program functions, however since the integrity of the program is not valid and one cannot be sure which part is valid, the program code that is able to run prohibits the reprogramming attempt because one cannot be sure the reprogramming would be done correctly. Reprogramming should never be interrupted for any reason, and a battery charger has been long proven to be a protective measure to ensure battery power is not interrupted during the reprogramming.

 

I'm surprised the flash-loader code isn't in ROM - other devices have a protected area which contains the minimum necessary to upload everthing into the flash. This minimal bootloader is very simple and well-tested, and in many cases is used to install a more complex loader which then does the real work of installing whatever update is required.

The important point being that in the event of a failure, the initial loader finds that there is not a complete image in the flash memory (because there's a location you program first to indicate "upload in progress" and then program again afterwards to show that the upload is complete), and it re-starts the upgrade cycle.

That's the correct way to implement a firmware bootloader/upgrader, and I'd be very surprised if Toyota couldn't get that right in 2003.

 

As an embedded systems engineer, I'm equally mystified. I'm willing to believe Toyota had their reasons (rather than simply having not thought about it!), but without hearing what those reasons were, I'd have to say it turned out to be a bad gamble...

 

I'd guess hubris. They assumed that the first shipping rev of the code was correct and would never need a field upgrade. Oopsie.

 

Actually, I'm betting that Japanese Technicians, read the instructions three times, and get it right the first time, knowing that they would bring eternal shame on their family for 10 generations if they botched it.

While American Technicians pay the same attention to the instructions that they do to the assembly manual that came with the kid's Christmas Bike, with approximately the same results

 

Hmm. Yeah, your theory does sound more likely...

 

Tinnnnnnnngg the sound the hammer makes when it hits the nail on the head.

 

The processors used in the automotive industry have migrated from EPROM to integral ROM due to program security concerns. There is much concern when aftermarket companies start to reprogram engine controllers, more concern when they could reprogram antilock brakes, stability controllers, and air bag controllers. There was a short period of time when the program was contained in ROM, with EEPROM being used for calibrations and diagnostics. This did not allow for cost effective updates if they were needed. When FLASH ROM was proven to survive the automotive requirements, this technology was used to address service as well as allow for common controllers to be loaded with different programs for different applications, thus reducing cost. The cost of the processors is directly related to its features. Having ROM, Flash ROM, RAM, and EEPROM costs more than leaving out the ROM. From a programming standpoint, the processor still needs to boot and execute code. Firmware is restricted to the micro's micro code. All other software is just that.

From an automotive perspective, there is not enough memory available in the processors to have completely fail safe, super robust code. Therefore there are routines and procedures to allow for reprogramming in the unlikely case it is required. To have ROM with program code sufficient to boot and run the reprogramming routine is not considered practical, cost effective, or necessary. This may be debatable, however I have participated in the evolution from simple processors and EPROMS to hard ROM to Flash ROM to dual redundant micros or PIC systems and the driving factors of choice are cost and security. Reprogramming is cheaper than new ECU's, however one has to wonder when reprogramming problems occur. I have seen cases where the new program was bad and the successful reprogramming caused the ECU to lock. In the automotive industry, we are always challenged by cost, and often the best alternative is not cost effective.

 

mdacmeis (mark): Thank you for the concise and informative techical explanations you provide. They're really appreciated!

It's a poor reward for all the work you put into writing up these postings, but... I'm giving you my points!

 

Mark I've got to hand it to you that was one of the best articles posted on this subject. And the last line said it all. Thanks

 

Here is part of the official procedure that Toyota sends the technicians for those of you wondering. It does state that they should hook up a charger to the aux battery in bright red.

Hank

 

Well, there you go: the procedure doesn't tell the tech to turn the charger *on*. :_>

 

Hedwig, How goes it is your car fixed? I just got a call from Kourtney this afternoon my silver #9 is in Helena! I will fly out on the 13th to get it and drive it home. I hope you have gotten this glitch worked out.