Arduino based auto alkalinity tester DIY - Page 3

JoesTank · 06/12/2015, 10:38 AM

If you have access to a 3d printer, take a look at this: http://www.thingiverse.com/thing:210756. You can get a stepper motor and a driver for about $15 from eBay, and a few more bucks for the rest of the parts from a hardware store. So total cost for the unit would be around $20 max.

If you don't have a 3D printer, you may want to consider getting one, it could be very handy

karimwassef · 06/12/2015, 11:11 AM

I'll post my latest drawings this weekend. I have the motor, so some fabrication can start.

karimwassef · 06/12/2015, 06:54 PM

Here's the current design. This is as simple and easy to build as I see myself making

It comes in three parts. The faceplate, the injector box, and the test box.

Here's a view without the faceplate

Here's the injector box without the walls

Here's the test box

It has the motor with a shaft and an off-center coupler to a plate with a hole in it. This rotation is only 0.5" in diameter to get a swirl going.

The plate is sandwiched between two other plates and the side walls have notches to allow the swirler plate to line up.

I drew the test tube in four different swirl orientations superpositioned.

Here it is without the walls

The green board behind the test tube is the color sensor

Here's the swirl mechanism alone

karimwassef · 06/12/2015, 06:58 PM

the test tube bottom is glued with a soft rubber and the bottom has a tube for export

I'll have to experiment to see if any liquid goes down the tube during test. If it does, I may need to add an actuated valve to keep it clear until the test is finished.

karimwassef · 06/12/2015, 07:00 PM

Here's the connected unit

bobbravo2 · 06/12/2015, 07:21 PM

Cant wait to see it materialize. Subscribed

karimwassef · 06/14/2015, 05:11 PM

Actuators stuck at the post office needing a signature. Lost the weekend.

I've got the acrylic, motor, Uno, relays, and syringes. I'll take pics once I have the actuators

I am struggling with the inputs and outputs on the Uno. Going to Mega would simplify things, but I already have the Uno and want a proof of concept first.

I think I need a startup sequence where I use a doser pump to fill a small container with fresh saltwater each time I run a new test. This is because I want to separate the main tank from the test setup... Still thinking this through.

I may have to jettison the RODI to flush and export and just use saltwater. Any input on this would be welcome...

karimwassef · 06/15/2015, 07:21 PM

photo 0D69124D-FB37-4722-A56A-0FBA7021C0E0_zpsef4d00aq.jpg

karimwassef · 06/15/2015, 09:29 PM

I can probably get this to work with an Uno, but it's too constricting.

I'm going to the Arduino Mega instead.

If anyone has experience with sensor shields to convert quadrature encoder output, please share it. I'm going to code it for now, but I would like to learn more about it.

karimwassef · 06/15/2015, 09:41 PM

Here's the map of channels ... this is why I had to go Mega

karimwassef · 06/15/2015, 11:29 PM

It's not complete, but it's a start

karimwassef · 06/15/2015, 11:30 PM

Since I'm using a relay board, I'm skipping the components to drive the relay and making the "connection" straight to the board.

Also not including cleanup capacitors to ground. That'll come after I get a clean concept up and running.

formula409 · 06/16/2015, 06:41 AM

Can you link up the actuators, Ill have a look

formula409 · 06/16/2015, 06:47 AM

Have a look at how this guy has made his microfluidic chip/LOC
Interesting approach to colorimetric tests

His detector system is nice too

https://hackaday.io/project/4693-medical-tricorder

karimwassef · 06/16/2015, 11:31 AM

http://www.naimotion.com/mm3mex.htm

Here's the datasheet

http://www.naimotion.com/PDF_folder/...ges/mm3mex.pdf

Mine is the 6V 0.5"

zachts · 06/16/2015, 11:40 AM

Quote:

Originally Posted by karimwassef

Actuators stuck at the post office needing a signature. Lost the weekend.

I've got the acrylic, motor, Uno, relays, and syringes. I'll take pics once I have the actuators

I am struggling with the inputs and outputs on the Uno. Going to Mega would simplify things, but I already have the Uno and want a proof of concept first.

I think I need a startup sequence where I use a doser pump to fill a small container with fresh saltwater each time I run a new test. This is because I want to separate the main tank from the test setup... Still thinking this through.

I may have to jettison the RODI to flush and export and just use saltwater. Any input on this would be welcome...

I think you'll need to flush the test tube with RO to get a clean test otherwise it would be like not rinsing your vials when testing the old fashioned manual way....and build up salt and test solution residue very quick.

karimwassef · 06/16/2015, 03:38 PM

I usually wash with my new saltwater before starting a new test. This way, I don't contaminate my tubes with tap water or even dilute with RO.

I would flush at the end of a test with the current batch of saltwater. I would also flush before the next test with new saltwater.

jeromeit · 06/16/2015, 03:51 PM

woah... riding along.. .this looks amazing!

jeromeit · 06/16/2015, 03:54 PM

Hmmmm... I wonder if the output info can somehow make another controller like the apex respond.. I guess it could via IFTTT

zachts · 06/16/2015, 05:22 PM

Quote:

Originally Posted by karimwassef

I usually wash with my new saltwater before starting a new test. This way, I don't contaminate my tubes with tap water or even dilute with RO.

I would flush at the end of a test with the current batch of saltwater. I would also flush before the next test with new saltwater.

I agree flushing with new saltwater to be tested is also advisable, just like manual testing, though rinsing after the test with RO or better still actual distilled water would be even better to keep any salt from drying out and causing a build up over time. Depends on how easy and often you can inspect and replace the test vial though as it probably adds an extra step or two of complexity to the operation. but if running unattended for a long time I'd do it.

zachts · 06/16/2015, 05:24 PM

Quote:

Originally Posted by jeromeit

Hmmmm... I wonder if the output info can somehow make another controller like the apex respond.. I guess it could via IFTTT

I'd think it would be easy enough to generate an alert or remotely viewable results one way or another given the wireless and endless other communication options from arduino. I'd never trust it to make any adjustments to the tank chemistry remotely though without first verifying first hand.

karimwassef · 06/16/2015, 05:33 PM

So I was planning on using an output to close a relay to act as an input short to the Apex.

The idea being to eventually be able to control dose kalk based on a calculated need from the results.

So, let's say I want an Alk of 9 and I'm at 7, I know the amount of kalk needed to get there. I also should know the rate of my dosing pump.

Based on that, I can determine the length of time I need to dose. Break that into hourly events - say 2mins every hour and create triggers for the Apex.

pH probes would act as an alarm to override in case of a problem.

It may even be possible to add part A and B independently to dial in both Alk and Ca.

I'll need a lot more confidence in the accuracy and repeatability over time first.

zachts · 06/16/2015, 06:49 PM

Quote:

Originally Posted by karimwassef

It may even be possible to add part A and B independently to dial in both Alk and Ca.

I'll need a lot more confidence in the accuracy and repeatability over time first.

That's the part that worries me, if any one of the sensors involved failed it could be very bad while your away for long periods of time.

slow downward shifts in CA, ALK, and PH are much less of an issue on a tank your size than what might happen if things failed and your system though it needs to in correctly raise levels. You could program in some safeties to advert fast shifts upwards or put a cap on boost doses, but if it kept going long term that way the result could be a very dead tank after a few weeks......

karimwassef · 06/16/2015, 07:51 PM

Agreed. The good thing about having the arduino is that I can create more sophisticated alerts than I can with the Apex.

For example, capping a max daily dose on Apex is impossible to program. It has no way to sum up on time based on a measured variable. All I could do was limit the output per hour and then check hourly. This means that if I need a little more late in the day, I can't- even if I haven't dosed at all during the day.

I can add cumulative time on the Arduino and trigger an alarm and stop any more dosing until I override.

I also tried to program 'sudden change' alarms in Apex... Impossible. But I can program nearly anything for the arduino.

I also have two pH probes on the Apex that act as an alarm with a backup.

For an overdose to happen, this tester would have to fail, not catch the sudden change, not limit the daily dose and both pH probes would need to fail...

Not impossible, but very low probability. If all that does happen... Someone has it in for me!

zachts · 06/16/2015, 09:09 PM

Quote:

Originally Posted by karimwassef

Agreed. The good thing about having the arduino is that I can create more sophisticated alerts than I can with the Apex.

For example, capping a max daily dose on Apex is impossible to program. It has no way to sum up on time based on a measured variable. All I could do was limit the output per hour and then check hourly. This means that if I need a little more late in the day, I can't- even if I haven't dosed at all during the day.

I can add cumulative time on the Arduino and trigger an alarm and stop any more dosing until I override.

I also tried to program 'sudden change' alarms in Apex... Impossible. But I can program nearly anything for the arduino.

I also have two pH probes on the Apex that act as an alarm with a backup.

For an overdose to happen, this tester would have to fail, not catch the sudden change, not limit the daily dose and both pH probes would need to fail...

Not impossible, but very low probability. If all that does happen... Someone has it in for me!

knock on wood.......

06/12/2015, 07:21 PM	#56
bobbravo2 Reef Addict Join Date: Sep 2003 Location: Orlando, FL Posts: 967	Cant wait to see it materialize. Subscribed __________________ Bob G. - O.R.C.A Member Current Tank Info: 65 Gallon \| Radion Pro G3 \| MP40 \| MP10 \| Apex \| Jaebo DC9000 \| KNOP Calcium Reactor \| Mated Picasso Clowns \| Mix of LPS, SPS and Softies

06/12/2015, 10:38 AM	#51
JoesTank Registered Member Join Date: May 2015 Posts: 3	If you have access to a 3d printer, take a look at this: http://www.thingiverse.com/thing:210756. You can get a stepper motor and a driver for about $15 from eBay, and a few more bucks for the rest of the parts from a hardware store. So total cost for the unit would be around $20 max. If you don't have a 3D printer, you may want to consider getting one, it could be very handy

06/12/2015, 11:11 AM	#52
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	I'll post my latest drawings this weekend. I have the motor, so some fabrication can start.

06/12/2015, 06:54 PM	#53
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Here's the current design. This is as simple and easy to build as I see myself making It comes in three parts. The faceplate, the injector box, and the test box. Here's a view without the faceplate Here's the injector box without the walls Here's the test box It has the motor with a shaft and an off-center coupler to a plate with a hole in it. This rotation is only 0.5" in diameter to get a swirl going. The plate is sandwiched between two other plates and the side walls have notches to allow the swirler plate to line up. I drew the test tube in four different swirl orientations superpositioned. Here it is without the walls The green board behind the test tube is the color sensor Here's the swirl mechanism alone

06/12/2015, 06:58 PM	#54
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	the test tube bottom is glued with a soft rubber and the bottom has a tube for export I'll have to experiment to see if any liquid goes down the tube during test. If it does, I may need to add an actuated valve to keep it clear until the test is finished.

06/12/2015, 07:00 PM	#55
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Here's the connected unit

06/14/2015, 05:11 PM	#57
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Actuators stuck at the post office needing a signature. Lost the weekend. I've got the acrylic, motor, Uno, relays, and syringes. I'll take pics once I have the actuators I am struggling with the inputs and outputs on the Uno. Going to Mega would simplify things, but I already have the Uno and want a proof of concept first. I think I need a startup sequence where I use a doser pump to fill a small container with fresh saltwater each time I run a new test. This is because I want to separate the main tank from the test setup... Still thinking this through. I may have to jettison the RODI to flush and export and just use saltwater. Any input on this would be welcome...

06/15/2015, 07:21 PM	#58
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033

06/15/2015, 09:29 PM	#59
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	I can probably get this to work with an Uno, but it's too constricting. I'm going to the Arduino Mega instead. If anyone has experience with sensor shields to convert quadrature encoder output, please share it. I'm going to code it for now, but I would like to learn more about it.

06/15/2015, 09:41 PM	#60
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Here's the map of channels ... this is why I had to go Mega

06/15/2015, 11:29 PM	#61
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	It's not complete, but it's a start

06/15/2015, 11:30 PM	#62
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Since I'm using a relay board, I'm skipping the components to drive the relay and making the "connection" straight to the board. Also not including cleanup capacitors to ground. That'll come after I get a clean concept up and running.

06/16/2015, 06:41 AM	#63
formula409 Registered Member Join Date: Dec 2013 Posts: 37	Can you link up the actuators, Ill have a look

06/16/2015, 06:47 AM	#64
formula409 Registered Member Join Date: Dec 2013 Posts: 37	Have a look at how this guy has made his microfluidic chip/LOC Interesting approach to colorimetric tests His detector system is nice too https://hackaday.io/project/4693-medical-tricorder

06/16/2015, 11:31 AM	#65
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	http://www.naimotion.com/mm3mex.htm Here's the datasheet http://www.naimotion.com/PDF_folder/...ges/mm3mex.pdf Mine is the 6V 0.5"

06/16/2015, 03:38 PM	#67
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	I usually wash with my new saltwater before starting a new test. This way, I don't contaminate my tubes with tap water or even dilute with RO. I would flush at the end of a test with the current batch of saltwater. I would also flush before the next test with new saltwater.

06/16/2015, 03:51 PM	#68
jeromeit Registered Member Join Date: May 2012 Posts: 81	woah... riding along.. .this looks amazing!

06/16/2015, 03:54 PM	#69
jeromeit Registered Member Join Date: May 2012 Posts: 81	Hmmmm... I wonder if the output info can somehow make another controller like the apex respond.. I guess it could via IFTTT

06/16/2015, 05:33 PM	#72
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	So I was planning on using an output to close a relay to act as an input short to the Apex. The idea being to eventually be able to control dose kalk based on a calculated need from the results. So, let's say I want an Alk of 9 and I'm at 7, I know the amount of kalk needed to get there. I also should know the rate of my dosing pump. Based on that, I can determine the length of time I need to dose. Break that into hourly events - say 2mins every hour and create triggers for the Apex. pH probes would act as an alarm to override in case of a problem. It may even be possible to add part A and B independently to dial in both Alk and Ca. I'll need a lot more confidence in the accuracy and repeatability over time first.

06/16/2015, 07:51 PM	#74
karimwassef Registered Member Join Date: May 2004 Location: Dallas, TX Posts: 11,033	Agreed. The good thing about having the arduino is that I can create more sophisticated alerts than I can with the Apex. For example, capping a max daily dose on Apex is impossible to program. It has no way to sum up on time based on a measured variable. All I could do was limit the output per hour and then check hourly. This means that if I need a little more late in the day, I can't- even if I haven't dosed at all during the day. I can add cumulative time on the Arduino and trigger an alarm and stop any more dosing until I override. I also tried to program 'sudden change' alarms in Apex... Impossible. But I can program nearly anything for the arduino. I also have two pH probes on the Apex that act as an alarm with a backup. For an overdose to happen, this tester would have to fail, not catch the sudden change, not limit the daily dose and both pH probes would need to fail... Not impossible, but very low probability. If all that does happen... Someone has it in for me!