So looking over the datasheet and the evaluation board instructions, I see how you can test a single CDC.
For my project, I would like to use eight AD7746s. At first, it would seem erroneous to use multiple EVAL-boards. But I also figured I would reach out to this community so I do not make any assumptions.
My first thought is connecting all the digital inputs (SDA, SCLK, bINT) together into one EVAL board. But looking at the functional block diagram of the EVAL board, this will not work since the board only has one set of excitation outputs, EXCA and EXCB.
Is there an alternative approach, or is using 8 EVAL boards the worst case scenario?
Would it be cheaper / better if I just do a PCB layout that expands the EVAL board to the number of AD7746s that I need?
Thank you very much.
Sorry for the delay in getting back to you. If you need multiple sensor inputs then have you considered maybe the AD7142 or AD7143 parts? These parts have multiple sensor inputs so they would reduce the need to use multiple AD7746 chips.
Another multi-input part worth considering is the AD7147. There is an evaluation board available for this part.
I understand your situation better now. But I see some potential issues with trying to use multiple AD7746's in parallel. Firstly the AD7746 part has a fixed factory set I2C address so you won't be able to address to multiple parts in parallel. A work around would be to use a multiplexer switch to switch the I2C lines from the processor to each part as you wanted to read/write to it. In addition, using multiple AD7746's will mean multiple 32kHz excitations frequencies none of which will be exactly the same (leading to beat frequency interference). If the sensors or senor connection traces are close together and active at the same time then you could see some coupling of the excitation frequency and this will affect the performance. By how much I don't know, it would depend on the coupling strength.
So in summary, you could use the evaluation board as a base and add more parts but you will have to consider the I2C addressing issue and the excitation signal cross coupling between the sensors that will cause interference.
Hope the above helps.
Thank you for taking the time to answer my question.
Actually, there was a point when the AD7147 was considered.
However, for our purposes, we need the best resolution, and the AD7746 provides 24-bit resolution down to 4aF.
Also, for all the components you have recommended, there is only one excitation source output (SRC for AD7142/AD7143 and AC_shield for AD7147).
Since our design involves having a capacitor matrix, there are multiple CIN+ inputs and SRC outputs that we would like to measure.
As of now, we will be using one EVAL board to test each CDC one at a time. But we would like to expand this design so that we can measure multiple outputs at a given time.
Though it would look silly to have 8 separate EVAL boards hooked up to 8 computers to do these measurements, this seems to be our worst case scenario. Is it possible to do this efficiently? Or any other suggestions?
My initial thought would be expanding the current EVAL design to have more outputs, but that is also another worst case option.
Thanks for the quick reply.
Your initial analysis for our potential design set-up is very helpful.
I guess my mentor and I will cross that bridge when we get there; but thanks to you, we can focus the modifications based on these specific considerations.
For now, this addresses my concern.
I'll open a new thread much later once we get to that design step.
Again, thank you very much!