Vida CEM swapping

Post by **RickHaleParker** » 25 Sep 2021, 15:28

Could the shuffle order be found by trying different shuffle orders, one byte at a time. Then selecting the one that produces the highest latency for that byte.

In other words to get the first byte of the shuffle order use the mask B0 XX XX XX XX XX where B0 = 00 - 99 and the first byte of the shuffle ( S0 ) is 0 -5. Let B0 = the highest highest latency of the 600 measurements. Let S0 = the value that produced the highest latency.

B0 and S0 are now constants.

Then B0 B1 XX XX XX XX where B1 = 00 - 99 and S1 of the shuffle is 0 -5.
Then B0 B1 B2 XX XX XX where B2 = 00 - 99 and S2 of the shuffle is 0 -5.

Might keep on going, never know what you might discover.

Post by **vtl** » 26 Sep 2021, 08:15

In some cases - yes, can be found. However, there are cases where pin cracking is unreliable even with a known shuffle order. So finding out the order would be even less reliable.

Post by **RickHaleParker** » 26 Sep 2021, 10:57

I wonder how much of the unreliability is due to cheap transceivers and other hardware construction issues.
I got a big jump in reliability when I switched to Bosch CF160 transceivers and soldered breadboard.

Post by **vtl** » 27 Sep 2021, 09:02

RickHaleParker wrote: 26 Sep 2021, 10:57 I wonder how much of the unreliability is due to cheap transceivers and other hardware construction issues.
I got a big jump in reliability when I switched to Bosch CF160 transceivers and soldered breadboard.

Interesting. How did you power it? It's a 5V transceiver, you say it works fine with 3.3v?

Post by **RickHaleParker** » 27 Sep 2021, 11:12

vtl wrote: 27 Sep 2021, 09:02 Interesting. How did you power it? It's a 5V transceiver, you say it works fine with 3.3v?

I am powering it off all off V-USB. Which I tap at pin #26 ( Vin ). Vin is electrically the same point as V-USB with the V-USB to Vin trace on the underside not cut. Pin #24 ( 3.3V ) has no connection.

Post by **vtl** » 27 Sep 2021, 11:14

RickHaleParker wrote: 27 Sep 2021, 11:12
vtl wrote: 27 Sep 2021, 09:02 Interesting. How did you power it? It's a 5V transceiver, you say it works fine with 3.3v?
I am powering it off all off V-USB. Which I tap at pin #26 ( Vin ). Vin is electrically the same point as V-USB with the V-USB to Vin trace on the underside not cut. Pin #24 ( 3.3V ) has no connection.

Does it mean CAN RX to Teensy is 5v?

Post by **RickHaleParker** » 27 Sep 2021, 11:29

vtl wrote: 27 Sep 2021, 11:14 Does it mean CAN RX to Teensy is 5v?

RXO High level output voltage = .09Va min to Va Max where Va is the supply voltage. I have not measured the signal level.

Bosch CF160 Datasheet.

Post by **RickHaleParker** » 27 Sep 2021, 11:51

In a previous build I ran TJA1051 transceivers the same way without blowing the ports on the Teensy.

I suspect the Teensy 4.0 5V pin limit is 5V RMS not 5V peak. That is: It can take the 5V peak but not the heat produced by 5V RMS. Would not be the first time I encountered a datasheet that was lacking precision communication due to incomplete terminology . The V RMS of a data signal is going to be less then V peak.

As somebody that did Electronics Component Engineering for 15 year, I am not thrilled with PJRC's documentation. It reads like it was written by a MBA not a EE or a EET.

Post by **vtl** » 27 Sep 2021, 15:11

5v going along with better signal makes sense. I think T2Luke said he stays on 5v-powered MCP2515, because the crack success rate is higher than with 3.3v transceivers.

I may buy these Bosch, maybe early P2 will finally give up

Post by **RickHaleParker** » 27 Sep 2021, 15:57

vtl wrote: 27 Sep 2021, 15:11 I may buy these Bosch, maybe early P2 will finally give up

I mounted the Bosch on a breakout board. The chip is all you need on the board. The whole circuit is self contained.

I got chips, breakout boards and two already assembled in DIP format.

I could send you the two DIPs or two chips and breakout boards.