Dropping ammeter readings during autorange changes?

I’m testing a small solar cell with a resistive load. I obtained the screen shot by blocking (lower voltage) and unblocking (higher voltage) most of the light from hitting the solar panel.

Notice that I can consistently get what appears to be dropped ammeter readings during (what I suspect are) autoranging changes to the internal shunt resistor.

From my test and measurement days, I recall that these sort of instruments have to perform a break-before-make connections to the DUT to avoid damage to the instrument. If the ADC readings were not momentarily suppressed during the break-before-make period, then indeed you would see invalid ADC readings.

The dropped readings aren’t really very problematic, because they are statistically infrequent, therefore not materially affecting the average, standard deviation, etc.

Where they are a problem, is with the min/max reading, thus completely rescaling the current amplitude with each dropped reading, and losing all intuition for the interaction between small-valued I, V, as I change the light, and the load resistor.

Is there a way to filter/average the current min/max readings? Or to disable ADC readings surrounding autorange changes?

I suspect omitted ADC readings would be better than misleading ADC readings.

BR,
Kevin

1 Like

OK, I just found how to disable autoranging on current. This cures the corrupted ADC readings, and completely cures the autoranging jumps. So, I have a satisfactory work-around.
Kevin

1 Like

Hi @kwible,

I suspect that you are seeing the range switch from the 18 µA range to the 180 µA range. During the range switch, Joulescope switches to a different shunt resistor. When this happens, Joulescope measures up to about 5 samples (2.5 µs) of data before settling to the correct value. The existing software includes some minor exclusion filtering of the samples, but not so heavy as to remove everything. Interestingly, this “corrupted” data does include useful information.
Adding improved algorithms to compensate for this effect is on the future feature list, but a little further out.

As you discovered, you can use a fixed range to remove this effect. Your measurement resolution will not be as good for smaller measurements that would normally switch to a finer range. For your application, the min/max lines are probably not useful, and you can turn them off. Click FilePreferencesWaveformshow_min_maxoff.

Excellent tip. Thanks for sharing.
Kevin

1 Like