DIY Alkalinity Monitor

Borat · Oct 5, 2023

And what about 12v power supply? Is it supplying up to 2A?

If you are using less than 2A power supply - check that you are releasing pump torque when pumping water in or pumping water out. Holding pump torque consumes a lot of current (almost as much as rotating the stepper motor) - so you may have insufficient current if you are holding the torque and simultatneusly pumping water in or out.

My PSU provides up to 6A - but it feeds Alk monitor + 2x 3-head dosers, but I ensure that only one device is consuming current at any singe time (by scheduling dosers to dose in non-overlapping schedules).

iLMaRiO · Oct 5, 2023

Power supply is 12v/2a.

There is a priming procedure, before the beaker cleaning, that dumps 1ml of acid, but after that, i'll dump 20ml water, stir and then clean everything, so the dumped acid should be removed with the water (or at least, very very diluted)

Borat · Oct 5, 2023

iLMaRiO said:
Power supply is 12v/2a.

There is a priming procedure, before the beaker cleaning, that dumps 1ml of acid, but after that, i'll dump 20ml water, stir and then clean everything, so the dumped acid should be removed with the water (or at least, very very diluted)

When I pump water out - I pump 20% more than I pumped in (e.g. you add 75ml sample, then about 10ml of acid, total 85ml - I would pump out 115ml to be sure there is nothing in the beaker).. This ensures that if one of the pumps' calibration drifted - there would still be no sample water left in all scenarios. Do you do that?

Borat · Oct 5, 2023

Just a small thing: I was looking at your earlier posted code... the statement:

this->reagentUsed = round(this->reagentUsed*100)/100;

is a bit dangerous as sometimes float can be implicitly cast to int, I would re-write it as

this->reagentUsed = round(this->reagentUsed*100.0)/100.0

Can you post your most recent code (possibly with void main() part)?

iLMaRiO · Oct 5, 2023

Borat said:
When I pump water out - I pump 20% more than I pumped in (e.g. you add 75ml sample, then about 10ml of acid, total 85ml - I would pump out 115ml to be sure there is nothing in the beaker).. This ensures that if one of the pumps' calibration drifted - there would still be no sample water left in all scenarios. Do you do that?

Yes.

iLMaRiO · Oct 5, 2023

Borat said:
Just a small thing: I was looking at your earlier posted code... the statement:

this->reagentUsed = round(this->reagentUsed*100)/100;

is a bit dangerous as sometimes float can be implicitly cast to int, I would re-write it as

this->reagentUsed = round(this->reagentUsed*100.0)/100.0

Can you post your most recent code (possibly with void main() part)?

I'll change that.
The code is unchanged, at least for the kh part because i'm almost sure the issue in on hardware side (the code does a lot of other things, it's not just a kh tester)

Borat · Oct 5, 2023

iLMaRiO said:
I'll change that.
The code is unchanged, at least for the kh part because i'm almost sure the issue in on hardware side (the code does a lot of other things, it's not just a kh tester)

..also as I understand - you are using ESP32 (is that correct)? When you power ESP32, do you use 3.3v pin or VIN pin for input voltage (I am guessing your voltage regulator converts 12v -> 3.3v)?

iLMaRiO · Oct 5, 2023

Borat said:
..also as I understand - you are using ESP32 (is that correct)? When you power ESP32, do you use 3.3v pin or VIN pin for input voltage (I am guessing your voltage regulator converts 12v -> 3.3v)?

I use an embedded ESP32, it's a PCB that we desined months ago. There is a regulator from 12 to 5 and a regulator from 5 to 3v3. ESP32 and pH i2c are powered with 3v3. Stirrers with 5v, steppers directly to 12v, dosing pumps with some mosfet 12v

Latest 2 test (honestly, it's just one, that triggered a re-test due to high difference from the nightly test):

Code:

FIRST
pH;HCl;dKh
7.81;0.00;0.00
5.31;8.00;5.97
5.09;8.50;6.35
4.79;9.00;6.72
4.66;9.10;6.79
4.58;9.20;6.87
4.51;9.25;6.91
4.51;9.30;6.94
4.43;9.35;6.98
4.35;9.40;7.02


RETRY
pH;HCl;dKh
7.85;0.00;0.00
5.32;8.00;5.97
5.01;8.50;6.35
4.51;9.00;6.72
4.41;9.05;6.76
4.43;9.10;6.79
4.33;9.15;6.83

"real" kh is 8.4 (hanna checker, tested at the same time of the test)

I would like to skip the 0.05 dosage and just use 0.10 as much as possible. Then when <= 4.5, i'll dose 0.05 (it should be a single dosage to finish the test). I think that these pump are not reliable at dosing drops (0.05ml) without a tip on the hose.

Borat · Oct 5, 2023

iLMaRiO said:
I use an embedded ESP32, it's a PCB that we desined months ago. There is a regulator from 12 to 5 and a regulator from 5 to 3v3. ESP32 and pH i2c are powered with 3v3. Stirrers with 5v, steppers directly to 12v, dosing pumps with some mosfet 12v

Latest 2 test (honestly, it's just one, that triggered a re-test due to high difference from the nightly test):

Code:

FIRST pH;HCl;dKh 7.81;0.00;0.00 5.31;8.00;5.97 5.09;8.50;6.35 4.79;9.00;6.72 4.66;9.10;6.79 4.58;9.20;6.87 4.51;9.25;6.91 4.51;9.30;6.94 4.43;9.35;6.98 4.35;9.40;7.02 RETRY pH;HCl;dKh 7.85;0.00;0.00 5.32;8.00;5.97 5.01;8.50;6.35 4.51;9.00;6.72 4.41;9.05;6.76 4.43;9.10;6.79 4.33;9.15;6.83

"real" kh is 8.4 (hanna checker, tested at the same time of the test)

I would like to skip the 0.05 dosage and just use 0.10 as much as possible. Then when <= 4.5, i'll dose 0.05 (it should be a single dosage to finish the test). I think that these pump are not reliable at dosing drops (0.05ml) without a tip on the hose.

OK..

I would check the voltage on all devices as well... importantly that voltage does not fluctuate. I use 5V as power supply to Arduino and the actual voltage from the voltage regulator (if measured) is 4.95v which is pretty accurate (+/-- 1% tolerance).

iLMaRiO · Oct 5, 2023

Borat said:
OK..

I would check the voltage on all devices as well... importantly that voltage does not fluctuate. I use 5V as power supply to Arduino and the actual voltage from the voltage regulator (if measured) is 4.95v which is pretty accurate (+/-- 1% tolerance).

I'll check.
AMS1117 are very very stable, in my experience, but is better to check

Borat · Oct 5, 2023

iLMaRiO said:
I'll check.
AMS1117 are very very stable, in my experience, but is better to check

The reason I say you should check is that the Ezo board voltage range is 3.3v - 5v, so input voltage is near the lower bound of the acceptable voltage. Another reason is that you are using 3.3v logic on drivers that expected 5v signal. OK, 3.3v is probably still interpreted as "HIGH" - but you can never be 100% sure, especially if the actual voltage is, say 3v..

iLMaRiO · Oct 5, 2023

Borat said:
The reason I say you should check is that the Ezo board voltage range is 3.3v - 5v, so input voltage is near the lower bound of the acceptable voltage. Another reason is that you are using 3.3v logic on drivers that expected 5v signal. OK, 3.3v is probably still interpreted as "HIGH" - but you can never be 100% sure, especially if the actual voltage is, say 3v..

well, i had the exact same issue (but with a much more unstable ph) with the dfrobot analog board that is 3v3 only
that' why i'm thinking it's a bad reagent , probably i'm not diluiting to 0.02n ma to something higher (but counting that as 0.02)

BassamKassem · Oct 13, 2023

Hello @Borat . hope you are fine and doing great . i need your help understanding below variance in my new device that i made recent .. it is pretty much the same as the first one but the only difference is that am using DFRobot PH Board v2.0 while the old one using V1.1 both devices using same PH probe (Adwa) and they are pretty good ones

am facing variance in KH reading while no KH source being injected to the system at all for about 4 days now

below are the test results for you to advise

Attached is the last version of code am using

Test date / time	Start PH	KH Result	Acid used
10/13/2023 1:00	8.36	8.29 DKH	12.00 ml
10/13/2023 3:00	8.47	8.40 DKH	12.14 ml
10/13/2023 5:00	8.45	8.30 DKH	12.00 ml
10/13/2023 8:00	8.39	8.34 DKH	11.93 ml
10/13/2023 10:00	8.42	8.44 DKH	12.07 ml

Borat · Oct 13, 2023

BassamKassem said:
Hello @Borat . hope you are fine and doing great . i need your help understanding below variance in my new device that i made recent .. it is pretty much the same as the first one but the only difference is that am using DFRobot PH Board v2.0 while the old one using V1.1 both devices using same PH probe (Adwa) and they are pretty good ones

am facing variance in KH reading while no KH source being injected to the system at all for about 4 days now

below are the test results for you to advise

Attached is the last version of code am using

Test date / time Start PH KH Result Acid used
10/13/2023 1:00 8.36 8.29 DKH 12.00 ml
10/13/2023 3:00 8.47 8.40 DKH 12.14 ml
10/13/2023 5:00 8.45 8.30 DKH 12.00 ml
10/13/2023 8:00 8.39 8.34 DKH 11.93 ml
10/13/2023 10:00 8.42 8.44 DKH 12.07 ml

These are pretty good readings, what's bothering you?

BassamKassem · Oct 13, 2023

Borat said:
These are pretty good readings, what's bothering you?

Am not bothered my friend , I don't know how the KH increases without being injected to the system

or this considered normal variance

Test date / time	Start PH	KH Result	Acid used	KH Variance
10/13/2023 1:00	8.36	8.29 DKH	12.00 ml	+ 0.1
10/13/2023 3:00	8.47	8.40 DKH	12.14 ml	+ 0.11
10/13/2023 5:00	8.45	8.30 DKH	12.00 ml	- 0.1
10/13/2023 8:00	8.39	8.34 DKH	11.93 ml	+ 0.04
10/13/2023 10:00	8.42	8.44 DKH	12.07 ml	+ 0.1

Borat · Oct 13, 2023

BassamKassem said:
Am not bothered my friend , I don't know how the KH increases without being injected to the system or this considered normal variance

Test date / time Start PH KH Result Acid used KH Variance
10/13/2023 1:00 8.36 8.29 DKH 12.00 ml + 0.1
10/13/2023 3:00 8.47 8.40 DKH 12.14 ml + 0.11
10/13/2023 5:00 8.45 8.30 DKH 12.00 ml - 0.1
10/13/2023 8:00 8.39 8.34 DKH 11.93 ml + 0.04
10/13/2023 10:00 8.42 8.44 DKH 12.07 ml + 0.1

There is always variance.. nothing to worry about. I think it's mostly related to dissolved CO2 quantity (starting PH). I get variations larger than that - during night hours when the PH changes.

BassamKassem · Oct 13, 2023

Borat said:
There is always variance.. nothing to worry about. I think it's mostly related to dissolved CO2 quantity (starting PH). I get variations larger than that - during night hours when the PH changes.

Thank you so much

Bazzo · Oct 13, 2023

I hope they are well.
@Borat , so we can say that there are no variations in the case of @BassamKassem , since all market tests Read with 1 decimal place?

It's more for understanding, as I'm building mine, and I'll be testing it until the weekend

Hugs

iLMaRiO · Oct 13, 2023

Borat said:
The reason I say you should check is that the Ezo board voltage range is 3.3v - 5v, so input voltage is near the lower bound of the acceptable voltage. Another reason is that you are using 3.3v logic on drivers that expected 5v signal. OK, 3.3v is probably still interpreted as "HIGH" - but you can never be 100% sure, especially if the actual voltage is, say 3v..

checked right now and voltage is stable near 3v3
(also, I had the same exact issues with the dfrobot analog board, not only with the Atlas)

Bazzo · Oct 14, 2023

Hello @Borat , I hope you are well.
can you help me?
I set up your project today, but I can't get any results.

I did the following steps:
I calibrated the dosers
I calibrated the pH probe
dilute the acid 1:3

I have the following results:
AlkX: 2023-10-13 / 22:51 DKH: 0.00
Step# : ACID VOL : PH0 : 0.00 : 0.89
1 : 4.00 : 13.08
2:8.00:13.07
3:12.00:13.07
4:16.00:13.07
5:20.00:13.07
6:24.00:10.30
7 : ovf : 0.00
8 : -1.96 : ovf
9 : ovf : -1.95
10 : ovf : 130120.12
48 : -0.00 : 0.00
49 : -0.00 : 0.03

I don't know what could be wrong.

if you can help me..