Some update om this thread.
I have received both cables I ordered and I have used them both for some time now. Or at least used the v2 cable a bit to test it out.
The first cable I got was the one based on the Atmega162 chip. This required a loader (a new exe file to start) which was marked as malicious at virus total. This cable was marked as a v2 cable but was in fact a hex-can cable (the older version)
Cable price was $29
Here are some pics of the cable.
So as you can see this is just a hex-can compatible cable.
I already have a genuine hex-can cable so this was of no interest to me.
The loader for this cable is also marked and will most likely be removed by your av.
I opened up my genuine hex-can to have a look and I guess the reason for using the Atmega162 is that it is used by Ross-Tech in their hex-can cable
I would not recommend buying this type of cable that is based on the Atmega162 if you are in search for a v2 cable. I would assume that every fake clones based on this chip is not a v2 cable.
Also because of the loader which is marked as malicious you would probably need to only run this is a virtual machine.
I will probably give this cable away to someone that is ok with it been a older version.
So the next cable I got is based on the Stm32F405 chip which I think is more or less the same as used in the genuine v2 cable. I do believe I found a picture of a genuine v2 showing it is using the F4 chip.
Cable price was $49
Here is some pics of this fake v2 cable
The strange thing about this cable is that it did not require a loader or a mod of the original vcds exe file.
In the installation manual it said that you should disconnect internet when using but mine worked just fine with internet intake.
I have tested this cable for a month now and it works just fine. It is much faster logging with a higher sample rate compared to my genuine hex-can cable and all access to the car seems to be faster.
My curiosity for this cable and how it worked become to big so I removed the stm32f405 chip from it and inserted it into my reader. Unfortunately the chip was locked with RDP level 2 so no access to the debug interface was possible.
Fortunately this is no match for the ChipWhisperer so removing rdp2 was no big deal. (In fact I was a bit surprised on how easy it was)
Here is my setup when attacking the chip
(A bit messy but it works)
And here is the result :)
https://youtu.be/4JFo23tYOq0
Now I have debug access and can dump ram but rdp1 is still active.
But when talking to the chips bootloader I can send the read mem cmd (0x11, 0xee) and the glitch the rdp1 check. This works ok but from time to time I hit a chip erase (testing on a similar chip) which makes this method a bit unstable. My guess is that since it used the internal rc oscillator and calculates the baud rate on the fly then this is the reason for hitting the erase.
I will move over to use CAN since CAN require using the HSE clock which will make this more stable.
Also since I will provide the clock (be the masters of the universe) I can move over to clock glitching instead of the more unstable vcc glitch.
Anyway. Hop you found this interesting (or at least some of you that don’t hate the Chinese that much :)) and of course I would recommend anyone to buy the genuine version of it.
But if you are on a low budget and just need something that works then this could be it.
Also I don’t think that this v2 cable will work on any updated version of vcds that require a cable firmware update.
You will probably only be able to use the 20.4.1 version that came with the cable.
(Yes I checked and this is the unmodified 20.4.1 from Ross-Tech site)
Sent from my iPhone
Bookmarks