Last for tonight :)

Here's a typical output when measuring a 10nF silver mica (almost no cap
variation with voltage) cap which reads as exactly 10.000nF on my SR720 LCR
meter. I am still working on improving accuracy and polishing the source
code. Note that there is no arbitrary calibration going on here - this
accuracy level comes from just getting the voltage measurement, current
control, and ADC sample timing correct. This output shows two filtering
schemes, a simple average of 50 iterations (which took about 10 seconds to
complete) and a Mean-Median filter (tosses out the 90th and 10th
percentiles before averaging). Next I am going to try a Savitsky-Golay
filter (moving least-squares fit which can give 1st derivative directly).
The value for the lowest voltage is always a bit further off right now
because there is a disturbance in the current regulation when the current
flow to the capacitor is turned on. I've improved that as much as I can in
hardware and I am working on further processing it in software by possibly
changing the sampling rate or ramping the current up from a lower value
during the beginning of the measurement.

Cap tables cleared.
Stage 1 readings=3D172 tdiff=3D1007 v=3D    1.2
Stage 2:Approximate capacitance: 10035.9pF
-- Simple Average --
Iterations: 50
Overall Average: 10.003nF
Voltage(V),cap(pF)
   5.0,9.8728nF    -1.30 %
  10.0,9.9727nF    -0.30 %
  15.0,9.9119nF    -0.91 %
  20.0,9.9588nF    -0.44 %
  25.0,10.034nF     0.31 %
  30.0,10.005nF     0.02 %
  35.0,10.015nF     0.12 %
  40.0,10.020nF     0.17 %
  45.0,9.9693nF    -0.34 %
  50.0,9.9809nF    -0.22 %
  55.0,10.005nF     0.02 %
  60.0,10.046nF     0.43 %
  65.0,10.023nF     0.20 %
  70.0,10.023nF     0.20 %
  75.0,10.024nF     0.21 %
  80.0,10.020nF     0.17 %
  85.0,10.003nF     0.00 %
  90.0,10.029nF     0.26 %
  95.0,10.016nF     0.13 %
 100.0,10.001nF    -0.02 %
-- Mean-Median Filter --
Overall Average: 9.9974nF
Voltage(V),cap(pF)
   5.0,9.8774nF    -1.20 %
  10.0,9.9564nF    -0.41 %
  15.0,9.9019nF    -0.96 %
  20.0,9.9479nF    -0.50 %
  25.0,10.006nF     0.08 %
  30.0,10.012nF     0.14 %
  35.0,10.004nF     0.07 %
  40.0,10.014nF     0.16 %
  45.0,9.9826nF    -0.15 %
  50.0,9.9730nF    -0.24 %
  55.0,9.9989nF     0.01 %
  60.0,10.026nF     0.29 %
  65.0,10.018nF     0.20 %
  70.0,10.036nF     0.39 %
  75.0,10.024nF     0.27 %
  80.0,10.019nF     0.21 %
  85.0,10.009nF     0.12 %
  90.0,10.027nF     0.29 %
  95.0,10.002nF     0.05 %
 100.0,9.9942nF    -0.03 %


As for a capacitor which actually DOES show voltage-dependent behavior,
this is a link to a image of a graph of the data measured by my device as
well as the nominal manufacturer's data for a 15uF, 100V TDK X7R cap. ARIMA
was the name of a motor drive I designed which used this cap:

https://s13.postimg.org/6y4pcol13/arima_cap.jpg


On Sat, Feb 24, 2018 at 9:39 PM, Sean Breheny <shb7@cornell.edu> wrote:

> Also perhaps of interest:
>
> Photo of both boards:
> https://nofile.io/f/mrmNZxw5ZtA/20180217_142115.jpg
>
> Bottom of main board:
> https://nofile.io/f/PpRWO6Qkwdt/20180218_223104.jpg
>
> Bottom of output board:
> https://nofile.io/f/TnmTsCPsJYh/output+pcba+bottom.jpg
>
>
>
> On Sat, Feb 24, 2018 at 9:27 PM, Sean Breheny <shb7@cornell.edu> wrote:
>
>> Block diagram finally complete:
>>
>> https://nofile.io/f/M9qcafdGqbR/capacitor+bias+meter+block+
>> diagram+feb+24+2018.pdf
>>
>>
>> On Tue, Feb 20, 2018 at 9:01 AM, Sean Breheny <shb7@cornell.edu> wrote:
>>
>>> Thanks for providing the smaller versions, Russell
>>>
>>> Are you referring to the 110v shunt regulator?
>>>
>>> I guess I should remind people that I designed this to use what I had
>>> lying around - it is definitely not optimized for mass production!
>>>
>>> I should probably have drawn a simple block diagram, too. Maybe tonight=
..
>>>
>>>
>>>
>>>
>>>
>>
>
--=20
http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.