Design of an optimum switching configuration?

Thread Starter

kaimdaim

Joined Jan 26, 2020
4
I have 9 driven electrodes. I want to read (I) and (V) measurements via these probes. Here is the pairing configuration(Current is measured via the first couple ,and the voltage is measured via the second couple):
1-2+3-4
1-2+4-5
1-2+5-6
1-2+6-7
1-2+7-8
1-2+8-9
2-3+4-5
2-3+5-6
2-3+6-7
2-3+7-8
2-3+8-9
3-4+5-6
3-4+6-7
3-4+7-8
3-4+8-9
4-5+6-7
4-5+7-8
4-5+8-9
5-6+7-8
5-6+8-9
6-7+8-9
As seen,the total number of pairings is 21.
So,what kind of switching configuration would be the best,in order to save time utmost for measurement process?I would like to get ideas to achieve an optimum design.
 

danadak

Joined Mar 10, 2018
3,873
What sampling rate, resolution, accuracy do you need ?

Assuming synchronicity needed in I and V measurement possibly dual 12 bit SAR approach -


1580123052231.png


This is a single chip solution, lots of other resources onchip available, see right
hand window resources used/left.

All channels configed as differential, if everything is ground referenced you could config
as single ended (saving pins).

The SARS are synced so that measurement of I and V is synchronous.

You would code to trigger the two SARS to make 2 of 4 measurements, then
go to next set of channels to get he other two measurements to do the math
operations. Or you could let the SARs free run and using DMA collect all the
samples, and then code to do the channel list configuration you stipulated.
This approach does add latency.

If channel sequence not the right order using sequencing SAR just use SAR ( non sequencing)
and add analog mux (also onchip) to front end driven by a LUT (onchip as well) to do the correct
channel addressing.

Many variants on this. If you need more resolution use Delsig (good to 20 bits) and
mux, again one chip.

IDE (PSOC Creator) and compiler free, board to use would be ~ $ 10.

https://www.cypress.com/documentation/development-kitsboards/cy8ckit-059-psoc-5lp-prototyping-kit-onboard-programmer-and


Regards, Dana.
 
Last edited:

Thread Starter

kaimdaim

Joined Jan 26, 2020
4
Thanks for this smart reply...
However it looks I have forgotten to mention about a detail above; that, during the measurement the electrodes are dipolar.For example,in 1-2+3-4 step,3 is negative pole and 4 is the positive pole,and then in the following step 1-2+4-5 the electrode 4 is the negative pole. That's why switching configuration has become a problem for me.
 

danadak

Joined Mar 10, 2018
3,873
Here is the basic approach using a LUT to invoke the channel sequencing.

Basically each time the SAR converts it clks the LUT to next state, which generates
the channel address for mux you want converted next. In code you read the
ChannelNumb status reg to get the channel being converted so you can do the
math (subtractions). The end of conversion would be used as an interrupt to
invoke SAR and ChannelNumb read, or do this with DMA.

You would dupe this so you have one working on V, the other on I, and synch
them as shown in the other post. Again this is all one chip.

This too is single chip solution.

1580127501280.png


The LUT config is just point and click binary state sequence you want -

1580127037539.png



Regards, Dana.
 

Attachments

Last edited:

danadak

Joined Mar 10, 2018
3,873
I just realized you had 21 states, just increase the LUT to 5 bits. Unused states
return the LUT address to beginning.


Regards, Dana.
 

danadak

Joined Mar 10, 2018
3,873
Thanks for this smart reply...
However it looks I have forgotten to mention about a detail above; that, during the measurement the electrodes are dipolar.For example,in 1-2+3-4 step,3 is negative pole and 4 is the positive pole,and then in the following step 1-2+4-5 the electrode 4 is the negative pole. That's why switching configuration has become a problem for me.
The poles switch sign in V or in code ? If former the differential input config for
the SAR takes care of that. The only challenge is handling the CM range of the inputs,
what is that ?


Regards, Dana.
 
Top