Decoupling capacitor method


I’ve read Colin’s paper on using decoupling capacitors for current measurements: … penADC.pdf

I’ve been trying to reproduce it on my target and I’m observing no difference between having nothing connected to the measurement port and having my wire connected. LNA set to high with max gain. I’m wondering if wire length is an issue. I’m using a 36awg wire with 7 wraps. Then I made a twisted pair with a ground wire. The tail of the wire is about 20cm to reach the cwlite comfortably but I’m afraid it’s too long and the signal gets degraded. What is the optimal max length? Any tips to get the cw to reach?

Also the decoupling capacitor I chose is a 0402 1uF which seems to bypass 10-100MHz correctly. My device is running at 36MHz. However, I’m unsure of the material of the cap–should I replace it with something known? Would that make a great difference?

Also any advice to check that you’re getting a good signal? I’ve been collecting 100 traces and averaging it using cwanalysis and observing a straight line which means I’m just getting noise.


Hi, Yifan. I am now working on redoing your attack on PSV. while searching for any tips for hardware setup, I reached to your thread. I think you did have solve this problem in your research. Would you please release a little more detailed hardware setup of your attack ?
BTW, i went through your git repository, I did not find the eagle project of psvcw board which was mentioned in your DFA article.