Additional tests for colorimeter (pH, alkalinity, calcium hardness, etc.)
OK I found the assay table files on my Mac, created a new one, 'pH' & transposed the average of the LaMotte & Hach 528nm average results (I have the standard issue 528nm green LED).
The data is pasted below but in the Colorimeter GUI it incorrectly gives me a pH of 9.5(ish) for a green LED absorbance of 0.555. I suspect this has something to do with the way the line of best fit is drawn... all the default assay tables' absorbances are linearly proportional to the ppm/uM level, but pH isn't (look at the graph). Also, I think it's struggling with the 0.00 absorbance value being pH 6.0, not 0.0... when i refit the data to a range of 0-2.2 pH (down from 6.0 to 8.2), the same test gives me a more sensible 1.5 value which is within the new/fake data set.
Any ideas how to get a good curve fit through the data set whose 0.0 absorbance is a pH value > 0?
Date: 2017-05-05 19:30:42 PDT
name: pH Test 07/12/12
- [6.0, 0.00]
- [6.2, 0.07]
- [6.4, 0.092]
- [6.6, 0.13]
- [6.8, 0.185]
- [7.0, 0.25]
- [7.2, 0.36]
- [7.4, 0.485]
- [7.6, 0.64]
- [7.8, 0.755]
- [8.0, 0.86]
- [8.2, 0.99]
We have instructions for importing and using any of our calibration data (including the pH data) here: https://sites.google.com/iorodeo.com/colorimeter/assays?authuser=0
Try going through these steps of downloading the data, loading into the plotting program and fitting the data before exporting.
By the way, this pH calibration data will only work for measuring pH between 6-0 and 8.2.
Let me know how it goes ...
Are you willing to share the API results (bromothymol blue)?
I have pH calibration for API pH 6.0-7.6 and API High Range pH 7.4 to 8.8. The curves are non-linear (polynomial), especially for the high range. I am new to IO Rodeo Forum and I am wondering if there are discussions regarding non-linear calibrations (I looked and didn't find any).
Here's our calibration:
K.Jin, A.Aufdenkampe, B.Fisher
API pH 6.0-7.6 bromothymol blue
pH | Absorbance
I imported your data into the colorimeter-plot program and tried looking at a couple a nonlinear calibrations. The 4-th order polynomial fit looks pretty reasonable.
Any error message? or are the radio buttons disabled?
There isn't an error message. When I select 4th order and hit the plot button, nothing happens. But I choose linear, 2nd order, or 3rd order it works fine. Same data I posted, same that you used. I'm using that Mac OS software version 4, which appeared to be the current version when I looked yesterday.
(Thanks so much for your responsiveness!)
Yes I think you are getting hit with a bug in the nonlinear fitting which I fixed on (2016-09-04). https://bitbucket.org/iorodeo/colorimeter/commits/all
The fix is in the source in the repository (https://bitbucket.org/iorodeo/colorimeter/src), but I don't think I have packaged up new stand alone versions for Mac or Windows since the fix.
I'll try to put together some new standalone versions this weekend.
Hi @Will-Dickson, I haven't noticed a bitbucket update in the downloads, and I'm afraid I don't have sufficient Python experience to update it myself. Keep me posted and I look forward using the updated version. Thanks!
Sorry about the delay. I've put new versions of colorimeter software (colorimeter-v05) for Windows and Mac OS here https://bitbucket.org/iorodeo/colorimeter/downloads/ Let me know if you have any issues.
@Will-Dickson Thank you! I very much appreciate your help!
Is there any changelog available ?
I don't have a change log, but this is probably something I should start doing. There are the repository commits here
https://bitbucket.org/iorodeo/colorimeter/commits/all The new release incorporates all the changes starting at 2016-09-04
@Will-Dickson Hi. We bought io rodeo colorimeter kit last week and we downloaded your new software which is the version 5. We seem to have a problem. When we click the start button of the measurement program, the program exits. But we don't find this issue when using the version 4 software. I am using a Windows 10 OS.
I can reproduce the issue you are seeing. I will look into it this week and see if I can fix the problem.
geekisquad last edited by
This post is deleted!