alex_pescaru wrote:

Disco_Mikey
A little bit off-topic...
The secondary BL, like the primary one, also protects some memory zones from tampering.
But it is a lot more permissible than the primary one. Question is: did you tried to modify it (the memory access table in SBL) in order to gain full access to all memory zones and in this way you can alter whatever information you want, making it possible to reuse an used one?

What?

alex_pescaru wrote:

Please read the below as it is and not between the lines:

1. You said that as soon as you reduced the voltage, even slightly, the same thing happened.
a) The error appear/is/reside in the CJB or in the Nanocom? Which malfunctioned because of the voltage drop? The CJB electronics or Nanocom electronics?
b) In this case, or in the end, is the CJB recoverable or not?

2. If the error is generated by a voltage drop, how can you fix the problem? Voltage drops could appear, thank God (or God forbid), quite a lot on the field. You said that you figured out a full solution. Should we understand that's a hardware solution? And if it's not hardware and it's a software one, how can a software solution could solve the problems caused by a voltage drop?

3. In the end, could you tell us where/which was the problem? Thanks!

Sorry for the above, but I was forced to make a lot of suppositions as all the explanations are a little vague...

Hi Alex,

As you are one of the people with more in depth knowledge of these systems I would really like to hear your feedback on this. P.s be gentle, i had a beer or two I have tried to answer fast and to the point i hope, before the long write up about this problem.


1a) I don't believe the error appears/is/resides in either the nanocom or the CJB, and none of the electronics malfunctioned on either system. the problem with a voltage drop is that the data itself, is susceptible to corruption during transmission when the voltage is not at the required level. I.e a block of data being transmitted can either be badly transmitted, or badly received due to the drop in power. On a system like a D3, with a power issue, this would of-course be intermittent based on the vehicles power fluctuations. So, I believe, when a programming session is open and data is being written, one missed bit of data and your in trouble.

1)aa **You can easily eliminate one factor:**
The nanocom is powered by the car via normal usage, although of very low power consumption, the case of the nanocom electronics failing due to a power drop can easily be eliminated by providing an additional 12V supply directly into the nanocom via its DC socket. But I think this is irrelevant compared to the power consumption of the cars ecus. And, although the bulletin states the recovery method is for CBJ's non responsive due to a power loss or communication break during flashing, the majority of the cases in my belief, and after lots of testing, are due to power loss, and to be specific a gradual power loss, not a sudden one, which will allow for data corruption.


1b) The CJB, if unresponsive due to a loss in power/comms when flashing, based on all our tests, should be recoverable every time. This also refers to your previous post regarding primary/secondary boot-loaders but too long to get into now. However I will say that the BCM seems like one of the most robust ECUs on these vehicles, but trying to manipulate and modify the PBL is probably the only sure way to brick it if you get it wrong and if indeed you can do it. Finally, there is of-course the scenario that the electronic components themselves have failed in which case its a completely different story.


2) We cant fix a voltage drop problem, i don't see how anybody can without a proper source. We can measure the voltage before flashing but if i drops halfway through the flash your in the same situation, it would be irrelevant. Basically don't flash anything in the field, or garage, unless you can provide a proper power supply to begin with. This applies to all ECUS, not just the BCM. When BBSSPY referred to a full solution, I believe he was referring to a non hardware solution and assuming the most basic off issues, like a good power supply, were already addressed an in place. I don't see how a software solution can be implemented for a power drop. You cant stop the flashing half way? And when the device running the software monitoring the power loses power also?

3) It never ends

Some answers may seem direct, please don't take them as such, I just thought i would give my thoughts.

cheers

Chris

BBS SPY wrote:

engebo

Have no worries or concerns, just plug in your new CJB and enjoy