PIC mcu memory question

[Shadow1976]

Hi all,
Love this site.
I got a question about a pic 18fxxx. I know it's pretty difficult to read a protected chips code. That I am not interested in. I want to know if you can read the read/write memory of the chip and if so, what's the easiest way.
Thanks for info you may provide.

 

[BobK]

You mean the RAM? The RAM is in random state whenever the chip powers up, so there is nothing there to read unless the chip is running.

Bob

 

[Shadow1976]

I am assuming the data memory?

 

[BobK]

Why not explain what you are trying to do. Data memory is constantly changing during the running of a program, and does not contain anything when the chip is powered off.

Bob

 

[Shadow1976]

I am trying to find out what data you can read from a protected chip. Experiment.

 

[BobK]

I give up.

Bob

 

[(*steve*)]

I am trying to find out what data you can read from a protected chip.

Unless the top surface of the chip has been ablated, sanded, or etched, you can normally quite easily read the part number.

 

[24Volts]

I think the OP is asking if we can read the program memory
(the program running in the chip) while the chip is protected!!!

I guess!

 

[KrisBlueNZ]

I guess Steve knew that. I guess he was just being a smart-arse.

I think you can assume that no one here knows of any way to read the RAM in a running PIC.

 

[(*steve*)]

Google will be your friend.

There are a number of attack vectors, which one (if any) will be effective depend on the chip, how it is protected, possibly how many samples you have, and what your funding is like.

One method is to carefully monitor the current draw as the device operates and from that determine what it is doing.

Another method is to dissolve the chip's epoxy covering and use a scanning electron microscope to look carefully at the die. Apparantly with some types of electron microscope you can "see" charge.

Other ways involve doing stuff to the chip to corrupt the contents of that part of the chip holding the "protection" or making it spit out code or values.

This was once a far easier thing to do than it now is as chip manufacturers are learning from this stuff and improving their products.

Here is some very old information. And here too. Apparantly if you can afford $US9k you can get someone to do it for you.

 

[24Volts]

There are a number of attack vectors, which one (if any) will be effective depend on the chip, how it is protected, possibly how many samples you have, and what your funding is like.

One method is to carefully monitor the current draw as the device operates and from that determine what it is doing.

Another method is to dissolve the chip's epoxy covering and use a scanning electron microscope to look carefully at the die. Apparantly with some types of electron microscope you can "see" charge.

Other ways involve doing stuff to the chip to corrupt the contents of that part of the chip holding the "protection" or making it spit out code or values.

This was once a far easier thing to do than it now is as chip manufacturers are learning from this stuff and improving their products.

Here is some very old information. And here too. Apparantly if you can afford $US9k you can get someone to do it for you.

wow ... and this makes it another thing to worry about !!! LOL
what a world!!

 

[BobK]

I think the OP is asking if we can read the program memory
(the program running in the chip) while the chip is protected!!!

I guess!

Uh, no, the OP explicitly said he was wanting to read the RAM, not the program memory.

Hi all,
I know it's pretty difficult to read a protected chips code. That I am not interested in. I want to know if you can read the read/write memory of the chip and if so, what's the easiest way.

Bob