Pi-casso : an automatic alkalinity tester

OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
You will still have to compensate for starting pH otherwise you could easily find yourself with an additional swing of 0.3dkh depending on time of day your sample is taken.

Going to try a sort out temperature first and then look at pH.
I'm finding starting pH very confusing. The most surprising part is how long it takes to return to normal after the test, > 2 hours, not stirred.
So at 1/2 hour test intervals the starting pH is only about 7.1.
Try measuring the volume required to shift the sample from 7.8pH to 4.4pH

You need to ignore the first acid input until the sample is below 7.8pH
When I get a chance I will have a look at this.
 

ZaneTer

Well-Known Member
View Badges
Joined
Jan 14, 2018
Messages
920
Reaction score
877
Rating - 0%
0   0   0
Going to try a sort out temperature first and then look at pH.
I'm finding starting pH very confusing. The most surprising part is how long it takes to return to normal after the test, > 2 hours, not stirred.
So at 1/2 hour test intervals the starting pH is only about 7.1.

When I get a chance I will have a look at this.
Hi Chris that is an extremely long time before the reading normalises. I suspect you are going too low with your pH during testing. You only really need to reach 4.2

I think you are over saturating the probe with hydrogen ions and this is the cause of it taking so long before it can register a normal pH again.
 
OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
ATM it's titrating down to 3.7. Will need to take some time to fully understand the modified half gran calculation method that the calkulate software uses. But reading elsewhere seems to suggest titrating down to 3.7-3.5 gives a slightly more accurate result when using the gran plot method for calculating alk.
Been roughly following: Alkalinity and Acid Neutralizing Capacity
In the following image alk is top (yellow) and starting pH is bottom (green)
test3.png
Tests are every 30min and the actual test takes approx 14min.
Normal pH should fall between 7.9-8.1
From the left:
Test 3 had a 90min gap from previous. Note, starting pH is higher and alk is higher.
Test 14 and > is after the probe was cleaned in a 1M HCL solution for 5min. Note, the starting pH is higher but now alk is lower.
The probe was similarly cleaned about 2weeks ago.
Chris
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
ATM it's titrating down to 3.7. Will need to take some time to fully understand the modified half gran calculation method that the calkulate software uses. But reading elsewhere seems to suggest titrating down to 3.7-3.5 gives a slightly more accurate result when using the gran plot method for calculating alk.
Been roughly following: Alkalinity and Acid Neutralizing Capacity
In the following image alk is top (yellow) and starting pH is bottom (green)
test3.png
Tests are every 30min and the actual test takes approx 14min.
Normal pH should fall between 7.9-8.1
From the left:
Test 3 had a 90min gap from previous. Note, starting pH is higher and alk is higher.
Test 14 and > is after the probe was cleaned in a 1M HCL solution for 5min. Note, the starting pH is higher but now alk is lower.
The probe was similarly cleaned about 2weeks ago.
Chris
I have nothing to contribute at this point, but im following this with great curiosity :) . Whatever comes out from it will be super insightful
 
OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
I tried adding a pH/temperature correction script. Didn't have any real effect on the daily alk swing.
Tried inputting temperature directly into the atlas pH circuit software. No real diff.
So just wrote an offset according to the alkalinity and temperature reading. Not sure if this is going to be probe specific or not, but seems to work.
In the image, blue is temperature (not including any pH) adjusted alkalinity and red is raw alkalinity. Y axis scale is 0.05dKH.
alkTestSetup.png


I suspect the small increase over time is from evaporation.

Chris
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
Very cool. I think its ok if theres little anomalies, as long as the relative range is good and we can see daily fluctuations that should suffice bulk of home aquarium requirements (dosing, sps husbandry etc)
 
OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
The precision after temperature correction is not to bad. +- 0.025 dKH
Temp correction is only a real concern if room temps around the test unit swing more then say 4-5 degC.

The correction factor is simply:
Python:
((temperature - 25) * correction_factor) + alkalinity result

correction_factor = 0.0145

The unit has since been moved onto my frag tank.
The result are very close to what I get from a Salifert test. Everything seems to work fine except one thing.
The readings change a little after cleaning the pH probe. I suspect this is due to bacterial/organics build up on the pH probe tip, causing a reduced response time of the probe. The following image shows the slow downward trend of pH.. Every 9 days the probe was cleaned. After cleaning you can see pH increases. At that point alkalinity also drops approx 0.16 dKH from the previous test.
ph_alk.png


Just thinking what is the best way forward.
- Can probably write some correction factor based around a daily pH average.
- Not sure if using some other type of acid, like sulfuric, would change the rate of bacterial/organic accumulation. I'm guessing not.
- Maybe it's a bad probe. Again I don't so, but will get another one soon to see.
- Something I'm missing altogether.
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
The precision after temperature correction is not to bad. +- 0.025 dKH
Temp correction is only a real concern if room temps around the test unit swing more then say 4-5 degC.

The correction factor is simply:
Python:
((temperature - 25) * correction_factor) + alkalinity result

correction_factor = 0.0145

The unit has since been moved onto my frag tank.
The result are very close to what I get from a Salifert test. Everything seems to work fine except one thing.
The readings change a little after cleaning the pH probe. I suspect this is due to bacterial/organics build up on the pH probe tip, causing a reduced response time of the probe. The following image shows the slow downward trend of pH.. Every 9 days the probe was cleaned. After cleaning you can see pH increases. At that point alkalinity also drops approx 0.16 dKH from the previous test.
ph_alk.png


Just thinking what is the best way forward.
- Can probably write some correction factor based around a daily pH average.
- Not sure if using some other type of acid, like sulfuric, would change the rate of bacterial/organic accumulation. I'm guessing not.
- Maybe it's a bad probe. Again I don't so, but will get another one soon to see.
- Something I'm missing altogether.
Looking good.
I would not worry a lot about it, since we can at least see the trend now. With that in place we can one by one figure out what’s the source , by replacing probe or cleaning regiment or titration acid. If anything makes it better (the slow decline ) then it’s awesome , but even if it’s not we can always come up with fitting calibration techniques as long as the decline is somewhat consistent and we know what’s causing it .

This is awesome work. Thank you for the amazing insights . I would love to build this and lend you a hand. I’m working on the stepper driver right now, which will hopefully allow me automate titration like fine fluid controls
 
OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
I’m working on the stepper driver right now, which will hopefully allow me automate titration like fine fluid controls
Sweet. What drivers are you planning on using? I'm going to change mine out for the tmc5160. Currently using tmc2100, which are comparable to a4988 and drv8825.
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
OP
OP
S

SDchris

Active Member
View Badges
Joined
Mar 3, 2015
Messages
123
Reaction score
164
Location
Sydney
Rating - 0%
0   0   0
Changed the testing interval from 1 hour to 2 hours. After 9 days removed the probe and cleaned it. This time there was very little change in the alk reading after cleaning. A little surprising! Will go through the process again to confirm.
The first two graphs are variations from cleaning the probe during ONE hour testing intervals after nine days. The third graph is cleaning the probe during TWO hour testing intervals after nine days.
Pink circle = cleaning point.

alk1ha.png


alk1hb.png


alk2h.png
 

Ranjib

7500 Club Member
View Badges
Joined
Apr 16, 2016
Messages
9,825
Reaction score
17,041
Location
Pleasant Hill, Concord
Rating - 0%
0   0   0
Changed the testing interval from 1 hour to 2 hours. After 9 days removed the probe and cleaned it. This time there was very little change in the alk reading after cleaning. A little surprising! Will go through the process again to confirm.
The first two graphs are variations from cleaning the probe during ONE hour testing intervals after nine days. The third graph is cleaning the probe during TWO hour testing intervals after nine days.
Pink circle = cleaning point.

alk1ha.png


alk1hb.png


alk2h.png
That's awesome. I am ok even with one reading every six hours :) .
 

LC8Sumi

Well-Known Member
View Badges
Joined
Dec 12, 2017
Messages
604
Reaction score
521
Location
Europe
Rating - 0%
0   0   0
I also have atlas pH probes (9 of them). They’re not bad for the price, but they’re not good either (+-0.2 pH reading error is not unexpected from them). Even at calibration you can wittness how it never really stabilises. I also have a GHL probe which is much more reliable, and is a lot more closer to Hach-Lange grade pH probes. Maybe the Atlas probe exchanges ions with the calib sol/sample/HCl & that’s the reason for the deviting readings. You could try and flush /store the probe with KCl after the measurements.
 

Mastering the art of locking and unlocking water pathways: What type of valves do you have on your aquarium plumbing?

  • Ball valves.

    Votes: 38 47.5%
  • Gate valves.

    Votes: 42 52.5%
  • Check valves.

    Votes: 16 20.0%
  • None.

    Votes: 21 26.3%
  • Other.

    Votes: 8 10.0%
Back
Top