Hi,
I often need to check sleep currents, and calculate energy totals, for devices that are experiencing an additional load from a connected debugger. A J-Link for instance will load down a Serial Wire Debug port at 1v8 by about 8.5 uA, the majority of which is a load on the Vtref pin. It would be quite handy if I could null out this extra drain in the Joulescope UI. Ideally with a text field + toggle button setting, so I can turn the compensation on and off without having to retype the value. But presumably you’d need to do the compensation in hardware to make the reported power come out right after downsampling?
Right now I just have to unplug the debugger cable every time I do a test - and I haven’t gathered the energy to make a 4-pole disconnect switch yet 
Hi @jamesl-dm - We do have a way to perform offset calibration today. In the Device Control widget, select Calibrate then click Start for Current offset. Although we say to disconnect all sensor terminals, this process will calibrate to any real current. This correction is applied before the autoranging computation, so significant variations will definitely alter autoranging behavior. This value can “argue” with the hardware overflow detection, which is unaffected by this offset change. 8.5 µA is 47% of the 18 µA range, which is enough to potentially cause strange autoranging behavior. Can you limit the minimum current range to 180 µA? 8.5 µA is only 5% of 180 µA, which is well within our design tolerances for this “disagreement”. The only downside is a slight increase in offset error, which is probably negligible given the J-Link variations:
The downside is that you have to remember to restore factory calibration, or the JS220 will have significant measurement error, as you intentionally calibrated to.
Instead of calibration, we could add a TARE that the instrument would apply after the autoranging. The instrument would still autoranging correctly, but just display “0” at the TARE value. We could make the TARE function automatic using the current measured value or allow you to enter a manual value.
The math does work out to do this entirely host side. While the math is not complicated, it does require an additional average voltage query to compute the power and energy waveform. It would be more consistent to apply the TARE on-instrument.
At this point, we don’t have plans to add more features to the JS220. I will keep in mind in case we do decide to add features, and I will definitely add it to our potential feature list for the third-generation JS320.
Does Calibrate work for you?
Hi Matt,
Thanks for the info - good to know that the calibrate procedure is an option. It’s a bit tricky to use though, since you’d have to arrange for the tare current to flow all on its own, and it’s generally accompanied by the load current you’re trying to measure. So a separate tare function would be welcome if time eventually permits.
Glad to hear there’s going to be a JS320 one day - looking forward to seeing what the next gen looks like.
Regards,
James
1 Like
