PH probe issues

[iamdan]

Thanks for that, have replied to GitHub comment as well I’m really over my head when it comes to programming so I’d rather leave that in your competent and capable hands

 

[iamdan]

I think for the “error correction” we need to include the NaN for those that do see that error and in my case and others I’ve seen having same issue where the reading drops to like in my case -0.15 being the reading I believe of the calibration of the probe at the time so to include anything under 1 or 2 to be ignored

While I appreciate my chat gpt code changes may not be ideal I don’t understand what you have asked me to change? Is it line 11 from = 0.0 to < 1.0?

 

[Sral]

Yes, pretty much.

For me line 13 in the file pH_fetch _data.cpp looks like this:

if (isnan(saveValue) || saveValue == 0.0){ return; }

And will be changed to:

if (isnan(saveValue) || saveValue <= 2.0){ return; }

Afterwards you would also have to compile the file in the command console:

g++ pH_fetch_data.cpp -o pH_fetch_data

 

[iamdan]

You mean run that g++ command on my reef-pi if I edit the file directly there?

 

[iamdan]

Otherwise if I make that change in GitHub do I have to do that compile thingy?

 

[Sral]

You mean run that g++ command on my reef-pi if I edit the file directly there?

Exactly. You have to find the file under /var/lib/reefPi/pH/pH_fetch_data.cpp
and make the appropriate changes.
Afterwards you have to run the g++ command to compile the code, otherwise the service that runs in the background will still run the old executable.

 

[iamdan]

ahh i get ya now

Ran it, got this - this ok?

robo-pi@robopi:/var/lib/reef-pi/pH $ g++ pH_fetch_data.cpp -o pH_fetch_data In file included from pH_fetch_data.cpp:9:
pH_sensor.cpp: In member function ‘float pHSensor::readMeasurement()’:
pH_sensor.cpp:43:38: warning: converting to non-pointer type ‘float’ from NULL [-Wconversion-null]
43 | if (openConnection() == 0) return __null;
| ^~~~~~
pH_sensor.cpp: In member function ‘float pHSensor::readFloatValue()’:
pH_sensor.cpp:101:12: warning: converting to non-pointer type ‘float’ from NULL [-Wconversion-null]
101 | return __null;

 

[Sral]

Oh yeah, I got those too, didn't seem to do any harm so I left it for now.

 

[iamdan]

Did all that stuff you advised last night and it still does funky stuff, so think we r close but it’s not there yet

 

[Sral]

Did all that stuff you advised last night and it still does funky stuff, so think we r close but it’s not there yet

Darnit, I‘ll see if I can find time today.

Another quick fix I thought of is to use the following transform formula:
v > 2 ? v : 8.2

It’s not perfect, but that would make every faulty value 8.2 instead of 0 or -0.15

 

[iamdan]

Yeah can try that, where n how do I put that exactly

 

[Sral]

Yeah can try that, where n how do I put that exactly

The pH Tab, your pH entry there has a line called "transform formula".

 

[iamdan]

Oh that part always wondered what the heck could go there lol, any luck looking at the code?

 

[Sral]

Oh that part always wondered what the heck could go there lol, any luck looking at the code?

Sadly not, real-life duties said no

 

[Sral]

Any luck so far with that transform formula ?

I have also updated the service code, which can be downloaded as always.

 

[Sral]

It now disregards all values below 2 and writes the communication buffer into a file. At least I hope it does, since I had no way to test this ^^

 

[iamdan]

It now disregards all values below 2 and writes the communication buffer into a file. At least I hope it does, since I had no way to test this ^^

Sorry to be the bearer of bad news.

This is all that happens after applying your latest build - Constant reading of the below.

The only file i see being generated in the pH directory (don't see any pH.dat we are supposed to be referring to)

 

[iamdan]

Contents of errordata.dat:

 

[Sral]

Hm, thanks a lot. I looked back to when I made the pH circuit work with the ESP32, and I did of course get the message structure wrong ... so the program always went into error mode. Sry, my bad for not checking properly enough.

Updated to version 1.2, deployment as always
Should not fall into error mode all the time and the errorData.dat is now more readable.

 

[iamdan]

Hi Sral

Been running your new build for many hours now, accuracy of ph readings is much better but still does the glitch to -0.15

Contents of errorData.dat