DIY Alkalinity Monitor

[certain_code]

I’d just buy this guys device . Based on Rasberrypi too.
https://www.reef2reef.com/threads/announcement-my-own-builded-kh-machine-is-ready.278352/page-32

Nothing stopping you from doing that. Personally, I like that and the Apex systems best. Apex is out because I just got the older system and I'm not paying for their Alk monitor and upgrading half my system. Reeftronics looks nice but I like an adventure

Great stuff.
Looks overall good to me. My advice will be not to spend a lot of time in getting the perfect circuit done. Whatever you build, it will likely go through multiple iterations before things settle down. I suspect a lot of it will involve having different calibration and validations in place.
I would consider a dual h-bridge darlington driver like l293d to drive the stepper or dc motor based peristaltic pumps. I am not sure how accurate is drop count via IR sensors, give it a shot. Using the peristaltic pump to pour measured amount is a common use, and we know that route will work, in fact some of the vendors (like atlas scientific) provide I2C interface specificallt geared towards that, just something to note.

I would also suggest using Pi over arduino. This will significantly reduce your development/iteration cycle time. Its cheaper (zero w is only 10$), and unless you need an analog input, you will have everything that you need (including two hardware backed pwm pins). Even for bare minimal testing, you 'll need to chart/analyze multiple readings, Pi and the entire linux ecosystem (python for example) will come very handy.

It will be very helpful if you keep the software under apache 2 or MIT and the hardware under openhardware licensing, that gives us (reef-pi and similar projects) an easy way to integrate with your project as well as vendor can consider this for their products without worrying for royalty. Its not a must, just a nice to have.

I have pondered about integrating a general titration module using ph/color sensors along side peristaltic pumps for reef-pi. But our current priorities are around 2.0 milestones, I hope we'll get back to this early next year or after x-mas. I hope we'll be able to use your learnings/project.

Godspeed & keep us posted. Any help I can offer around electronics/software/Pi let me know.

Thanks Ranjib, appreciate looking over my first chunks of work so far, I have been following reefPi here for a while now and love what you are doing with it. Happy to work on integration with both projects sometime for sure.

I will look into the H-bridge. Certainly not an engineer so might take a while to figure out why I would need them but the main key is that I can control the speed. The calibration plan right now is to figure out the fastest speed where the IR can accurately count drops but also not take forever to perform a test. This also would allow people to essentially use any decent pump with the system automatically adjusting.

I went the Arduino(ish) route just because it's easier for me to prototype with, honestly. The Particle Photon board has benefits over the normal Arduino which just adds to that (i.e. remote flashing). I know I will probably switch to something else eventually but for now I like the simplicity of the Arduino platform.

I'm not too worried about the software side license yet but thanks for the tip there. Certainly going to keep this open and make it easy for others to use, just don't want anyone profiting off of it really.

Right now I really like the idea of using a BNC for the output so essentially any controller can read the output. I am also going to try and keep the API open so the web based tools could be used separately as well. No idea yet if it makes sense to support other specific controller connections yet.

 

[certain_code]

Im really making two,points here. How to build... my point is, you can learn a lot from his blog, all the things to think about... OS upgrades, Bluetooth, WiFi, PH prope, reagents, pumps, historical data, cloud connectivity, etc. all the things you will want to consider and think about in your DIY project, all the feedback and ideas that others are giving him. Great ideas and things to consider yourself. and yes, I made a second point of, when you consider all the things to think about, all the things you may want to include, it might be easier to buy.

I get what you are saying and thank you for the advice. There certainly is a lot involved but that's why I created this post and want to work with others in figuring out everything as a community. Sure, it might be easier or cost less to just get a already final product but this is the DIY forum and it's about helping others learn something about the hobby.

For now, I'd say your point has been made and, again, I appreciate your insight. I'd like to keep the thread here on topic and not try to discourage the DIY mentality.

 

[Caswal]

I went the Arduino(ish) route just because it's easier for me to prototype with, honestly.

Often the best platform is the one that you are comfortable with and can do the job.

 

[kilgoretrout]

You guys are way too smart.

 

[rygh]

I was poking around with something similar a while back, although perhaps simpler.
The idea is : I don't really need true dKH, I just need to know when it drifts.
See enclosed pdf

 

[rygh]

Note on controller:
I have used Photon on a few things, and for me it has been unreliable long term if you push data to the cloud.
With this sort of thing, you definitely want logs and email/text triggers, so cloud capable.
Works for weeks, maybe a few months, but eventually hangs.
Seems to depend on Wifi stability though.

I would suggest off the shelf relay+transistor modules. Not only do they have diodes, etc,
some also have opto-isolation, reducing risks in our salty environments.
You can connect those directly to micro-controller.

 

[JimWelsh]

Reaching for the popcorn, and enjoying my two year head start....

 

[tango0102]

There was also a paper done by Mark Johnson detailing how measuring dissolved CO2 directly, and with a known pH value you can calculate the alkalinity. Although I'm uncertain of the type of accuracy need in measuring these 2 variable in order to have an accurate alkalinity value. The paper is titled "Direct and continuous measurement of dissolved carbon dioxide in freshwater aquatic systems—method and applications". It looks doable as some CO2 sensors are arduino based, but sensitivity of the sensor matters.

 

[rygh]

Reaching for the popcorn, and enjoying my two year head start....

Well...
I first posted the doc above Jan 2015.
http://www.bareefers.org/forum/threads/automated-alkalinity-measurement.17726/
But ... I have made zero progress.
So not sure a head start matters all that much.

 

[GrandeAzul]

Following

 

[JoeyRod_Dev]

There is a ton of info and probes with scripts already out there for the RaspberryPI. I honestly think it’s your best bet a fast and easy project. Let me know if you run into any development issues. GL!

 

[theatrus]

There are stepper motor driven peristaltic pumps on eBay for about $30. I’ve used them on a dosed system before and they are very repeatable. Much better investment than the super cheapy ones.

I worked out a concept similar to yours and even went as far as 3D printing a reaction chamber. Never got too much further . I’ll post some images of the design.

I’d still count drops with an optical sensor even with a stepper.

 

[JimWelsh]

Well...
I first posted the doc above Jan 2015.
http://www.bareefers.org/forum/threads/automated-alkalinity-measurement.17726/
But ... I have made zero progress.
So not sure a head start matters all that much.

Yeah, but I've had a functional unit running in my fish room since June 2016. Still do, but now mine does Ca and Mg, too! I made good use of my head start.

 

[BigJohnny]

Your proposed design is very similar to the focustronic alkatronic. Rasberry pi and 4 pumps etc. You will need steppers for accurate titration. Magnetic stirrer is something else you'll need to incorporate. If you haven't already, I recommend checking out the alkatronic manual. May help you visualize what you are trying to do better.

I have one and I love it btw. Stand alone unit, although it communicates with my controller but you certainly don't need one. It comes with a Bluetooth controlled outlet to shut off dosing pump/calcium reactor, and can dose carbonate on its own to raise dkh, so it can do everything without a controller. Has its own cloud and app for remote access etc.

If you have a controller it can send dkh through bnc to your apex/ghl/etc and you can do anything you want after that ; )

Good luck, following.

 

[theatrus]

Right now I really like the idea of using a BNC for the output so essentially any controller can read the output. I am also going to try and keep the API open so the web based tools could be used separately as well. No idea yet if it makes sense to support other specific controller connections yet.

Emulating a pH probe... will be interesting. You're going to go through a lot of work to fake a linear calibration, and hope the other input doesn't do any temperature compensation, etc.

Second vote for I2C, or CAN. https://stonyx.github.io/AquaBusLib/

Or since its just an RPi, add an HTTP API and systems like ReefPi just make an RPC

 

[sonny316]

Thanks for sharing! I'll be following...

 

[Dr.Xipoles]

That's a nice and simple algorithm to implement in arduino or Pi, I wish I had thought of that. Thanks for sharing.

 

[peterj66]

Just a thought, did you consider using a precision scale instead of a drop counter? I was thinking about it this morning and if you can measure the weight of the container used for testing with high enough precision, you should be able to dose reagents etc without the need for other calibration. That said I am not sure how often you would need to calibrate the scale but combined with stepper dosers you should get pretty high accuracy.
Also the colorimeter version may be less exact but is on the other hand the easiest method to adapt to other testing such as Ca and Mg.

 

[preston398]

I am interested in this as well and I will be building a dosing pump sometime in the future as well. I have found the dosers Neptune uses but the china shipping is expensive.

I am a soon to be EE graduate going to work at Nasa this summer and should have some free time once this term is over to work on projects like this if you need help.

 

[s2nhle]

Very interesting. following along.