uSupply implementation - problem with current sensing

[eneuro]

you could also create pseudo-differential input by measuring the virtual ground with one channel and the current sense value with another channel and subtracting them in the firmware.

The problem might be that there will be time difference between those two readings, while eg. @ 250kHz ADC clock and about 13 clock per ADC sampling about 52us delay. If we had to switch reference voltage that means as I well rember additional 25 ADC clock and be carefull and read if there is no bug hardware bug when switching reference voltage-as I remeber ATTiny85 had one-did'nt analysed it but read last time on AVRfreaks.
So, switching between voltage references and reading 2 ADC channels means some delays and discarding first ADC read for some AVRs -I wrote ADC routine for that for ATTiny AVRs and had to look into its datasheet while forgot about it since than, but question is if those delays will be acceptable in application and check AVRfreaks for any known MCU HARDWARE bugs which happends-that can save a lot of time, while MCU does not work as expected from datasheet sometimes 

[mat_fr]

Hey guys,

I wake this thread up just for a quick question on RevA/RevB project, for the sake of comprehension. I bought a MAX4080 to handle the current sensing (yeah, why not after all ?). I was just wondering what the RC filter (or isn't it one?) at the output of the MAX chip? I can understand it is to stabilize its output but the datasheet doesn't mention the need for any filtering.
So I wondered if there was any obscure reason to add this filtering.

[David Hess]

I wake this thread up just for a quick question on RevA/RevB project, for the sake of comprehension. I bought a MAX4080 to handle the current sensing (yeah, why not after all ?). I was just wondering what the RC filter (or isn't it one?) at the output of the MAX chip? I can understand it is to stabilize its output but the datasheet doesn't mention the need for any filtering.
So I wondered if there was any obscure reason to add this filtering.

You should start a new discussion thread.

I do not see anything in the MAX4080 datasheet about requiring an output filter.  It might have been added to reduce noise or to ignore short current spikes.

[mat_fr]

You're right of course, I should have started another topic, I just didn't wan't to multiply the subjects, this one still being about the same project.
Anyway, I breadboarded the circuit without the RC filter, with a MAX4080 and an LM358 to limit the Current and set the voltage, and it seems to (finally) work nicely.

Thanks again for all your advices, I'll try to make a real thing out of it in the next few weeks.