Help with Clock Glitching on Husky - Continuous "Trigger still high!" Error

Hello,

I’m trying to perform clock glitching experiments using the ChipWhisperer Husky kit with the following setup:

  • ChipWhisperer version: 5.7.0
  • Target platform: CW308_SAM4S
  • Starting from: SOLN_Fault 1_1 - Introduction to Clock Glitching notebook

The only modifications I made to the original notebook are:

  1. Setting PLATFORM = ‘CW308_SAM4S’
  2. Changing clock source to PLL with scope.glitch.clk_src = “pll”

However, I’m encountering a continuous “Trigger still high!” error. The target appears to be getting stuck in a state where the trigger never goes low, with the error repeating indefinitely:

Trigger still high!
Trigger still high!
Trigger still high!
Trigger still high!
Trigger still high!

What’s particularly puzzling is that I’ve successfully performed these same experiments on a ChipWhisperer Lite with an XMEGA target without any issues.

I’ve made sure to enable the glitch module specifically for Husky as shown in the documentation:

if scope._is_husky:
    scope.glitch.enabled = True

I suspect I’m missing some specific Husky setup step, but I can’t figure out what it might be. Any insights on what additional configuration might be needed for clock glitching on the Husky platform would be greatly appreciated.

Thank you for your help!
Pierrick

Hi Pierrick,

Are you seeing this message, or do you get some iterations where this doesn’t happen. You usually see it when the previous glitch attempt caused the target to crash, but if you’re seeing it every time, your target probably isn’t running at all. Did the programming step work complete without errors?

Hi Alex,

I initially misunderstood the problem—I thought the high trigger was forcing the reset. Since I hadn’t seen any successes before, I assumed the experiment wasn’t working. After running it again, I managed to get 6 successes out of 428 resets. Does that seem consistent to you with the solution notebook, the SAM4S, and the Husky?

Thanks!

Yeah, that seems reasonable. Not sure of the exact numbers I was getting last time I ran the lab, but I think it was around a success rate of 1%.