Confusion about CW312 and target board

Embedded Security ChipWhisperer Hardware
1

I have chipwhisperer husky and the CW308-Stm32F target board 。so I want to use adapter(cw308 to cw312 ) to connect the target board.

I assembled it using as shown and linked it to the CW312 but I noticed that LED 1 2 3 are off (not when I link the SAM4S), is it wrong for the way I am linking it? Or do I need to solder the target board and adapter?

adapter and target board:

4c2240c1991974a0f243f5572bc13e2
4c2240c1991974a0f243f5572bc13e21920×2560 219 KB

adapter and target board LED is off

473b55ae177313f4eb6b8f4cae4ab9c
473b55ae177313f4eb6b8f4cae4ab9c1920×2560 394 KB

SAM4S LED is on

8fca71a628cb17691e700d575880ccb
8fca71a628cb17691e700d575880ccb1920×2560 361 KB

The biggest problem is that I don’t know where it is, I hope someone can help me, thanks a lot!

2

Yeah, you need to solder the two boards together, simply connecting them won’t work.

3

Hi Alex

I’ve soldered it and now I’m able to connect to the target board.
But unexpectedly, my experiments with CW312 adapter +CW308 Stm32F4 target board are much worse compared to SAM4S.
For the same code, I get good results with both SAM4S and CWNANO (classification accuracy can be 99%+) , but 50% with CW312 adapter + CW308 Stm32F4 target board (dichotomous classification, so almost no differentiation at all).
I would like to ask what could be the possible reasons for this? Could it be a problem with my soldering?

I attached a picture after soldering:

50030b36b2080201f901ce59b561c0e

4

Can you post a picture of your power traces?

5

Sure

The following three diagrams are from CWANO, husky+SAM4S and husky+ STM32F4:

CWNANO:

ea508ab4d93568ee1a42af0d926539b
ea508ab4d93568ee1a42af0d926539b967×730 113 KB

husky+SAM4S

8f993bbe58457c9773152ab3436bd6c
8f993bbe58457c9773152ab3436bd6c1695×805 200 KB

husky+ STM32F4

7bcbfe14350e3a8f89a93554e47256c
7bcbfe14350e3a8f89a93554e47256c1632×882 149 KB

I think it’s pretty obvious how much noise, or jitter, can be felt.

I read this one recently and I checked my husky and there are no capacitors on the C49, indicating that I am a V3 husky, and I think it may have something to do with that.
Running SOLN_LAB_3_3 with a Husky doesn’t yield correct results

So I contacted someone who knows more about hardware (because I don’t) to come and solder the capacitors for me, hopefully this will help. Do you have any other suggestions?

6

The F4 has an internal regulator which we try to bypass by feeding in 1.2V into the CAP pins. The CW313 doesn’t have an adjustable regulator included like the CW308, but you can try feeding in a slightly higher voltage (I’d recommend somewhere between 1.2V and 1.3V) by disconnecting the jumper on JP1 and connecting your power supply to pin 1 of JP1 (the one closer to the 5V test point) to see if that improves things.

7

By the way, is there an outlet to purchase additional SAM4S target boards now? (Aside from buying a full set of husky) as I need it badly.

8

Your quickest bet is to email sales@newae.com, but I can also ask internally, if you’d like.

9

Thank you for your reply, I have sent the email.
I would also like to ask if I can avoid similar problems if I switch to the STM32F3 target board (since it also has a Cortex-M4 core).