<div dir="auto"><div>The filter doesn't make the signal constant between samples. It doesn't even make its slope constant between samples.</div><div dir="auto"><br></div><div dir="auto">What it does do is to average the fastest wiggles in a signal in a way similar to what you are asking for. :-)</div><div dir="auto"><br></div><div dir="auto">/mr</div><div dir="auto"><br><br><div class="gmail_quote" dir="auto"><div dir="ltr" class="gmail_attr">Den tors 11 feb. 2021 13:41cheater cheater <<a href="mailto:cheater00social@gmail.com">cheater00social@gmail.com</a>> skrev:<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">This is true in that the anti-aliasing filter limits the bandwidth of<br>
the signal, making it constant between sampling periods before it gets<br>
to the ADC. However, that's not the same as sampling the whole charge<br>
that was transported in the time between the sampling intervals. In<br>
specific, the anti-aliasing filter will shunt a portion of that charge<br>
to ground, bypassing the ADC.<br>
<br>
On Thu, Feb 11, 2021 at 1:32 PM Mattias Rickardsson <<a href="mailto:mr@analogue.org" target="_blank" rel="noreferrer">mr@analogue.org</a>> wrote:<br>
><br>
> Den tors 11 feb. 2021 08:48cheater cheater <<a href="mailto:cheater00social@gmail.com" target="_blank" rel="noreferrer">cheater00social@gmail.com</a>> skrev:<br>
>><br>
>> Here's one thing I've been wondering about. Is there a form of ADC<br>
>> where every sample represents the total charge that passed through a<br>
>> conductor since the last sample? Normally, an ADC will only sample<br>
>> voltage (or current) at the time of sampling, but we don't really know<br>
>> anything about what happened between sampling intervals. If we assume<br>
>> that the current was roughly constant during the sampling, that's<br>
>> fine. But if it changed during that time, then we have to do some sort<br>
>> of interpolation. Even if we do a linear interpolation, maybe the<br>
>> current going through the conductor was concave or convex. Being able<br>
>> to sample the total charge going through a cap would be much easier,<br>
>> since I'd be converting the input data into a charge amount anyways<br>
>> (to be added to the virtual capacitor).<br>
><br>
><br>
> This is essentially what an anti-aliasing filter does for you.<br>
><br>
> /mr<br>
><br>
</blockquote></div></div></div>