How do you determine value of the shunt resistor? For example looking at the values for the CW308T STM32 targets [1] I see values ranging from 10 Ohm to 33 Ohm. How do I chose the best value for my target MCU?
What about location of the shunt? Is it better to connect it between power supply and target VCC pin or between target GND pin and GND?
Where should the measurement be taken? The voltage drop over the shunt or between LO side of shunt and GND etc? The schematic of the CW308T-STM32F seems to indicate voltage drop over the shunt (SHUNTH/SHUNTL).
If I have a choice of powering the target device with VCC from 2.7-5.5v what is preferred for better power measurement?
With a shunt resistor, you’ve basically got a tradeoff between higher signal strength (larger value resistor) and target functionality (too much voltage drop causing issues with the chip). I’d use the 10-33 ohm as a rough ballpark, then refine from there. Often you don’t need to find an optimal resistor size, just something reasonable will do.
What about location of the shunt? Is it better to connect it between power supply and target VCC pin or between target GND pin and GND?
Most chips route different power pins to different parts of the chip, but ground is often shared, so putting a shunt on the VCC side can allow you to remove the power consumption of uninteresting parts of the chip, improving your attack.
Where should the measurement be taken? The voltage drop over the shunt or between LO side of shunt and GND etc? The schematic of the CW308T-STM32F seems to indicate voltage drop over the shunt (SHUNTH/SHUNTL).
For the normal chipwhisperer setup, the measurement essentially works as:
V_meas = V_h_ac - V_l_ac
V_meas = 0 - V_l_ac
Basically, it is the voltage drop we’re measuring, but the high side of the shunt is very stable, so we’re basically only getting the AC measurement of the low side of the shunt. This is all, of course, in reference to ground.
If I have a choice of powering the target device with VCC from 2.7-5.5v what is preferred for better power measurement?
I’m not sure, as I haven’t done any comparisons between leakage at different voltage levels. It might be something worth testing out.