First work out
- How fast does ph change 10 times a second, 1 time a second 1 time an hour
- How fast cant the electrovan operate and then operate again
- What range of ph values are allowable
- How precisely can I measure ph (to how many decimal places)
- Do I have the spec of voltage outputs for the ph Probe
Currently I see a lot of wich ful thinking, as you think the ph level will be one constant level, it will be changing and have noise on it, so values will twitter all over the place. Your scheme seems to assume ph values are 0 to 7, 8 , 9 and above, values from ADC and probe wil be something [
like- 0 to 7.9999999999
- 8.000000000 (or possibly not all depending on resolution steps)
- 8.000000001 and above
You need to think through the mechanical response times and measurement time.
I have no idea if a gain of 3.5 will do as I know nothing about the ph probe (there are many types out there). The ADC seems overkill as I doubt you will be taking readings more than once a second and that ADC can take 1 million samples a second, even PCF8591 or MCP3008 range could easily do the job and probably much cheaper.
You need to look at hysterisis, dead bands, and your basic algorithm as if you get a ph reading of exactly 8 you algorithm does nothing but your software could put electrovan in wrong position.
Just another techie on the net - For GPIO boards see http:///www.facebook.com/pcservicesreading
or http://www.pcserviceselectronics.co.uk/pi/