Round 3 - Capn Style - Automated

[Skybound]

Which PH sensor did you get? Atlas Scientific makes a whole line of probes and circuits to work with Arduino and example code to get it up and running right away. I've had my eye on them for over a year now, but the high price has kept me away thus far. Hopefully after a harvest or two, I'll be able to start down that road.

 

[sleepless]

I found one on Amazo* that a guy used in an instructable.

I decided to hold off on EC.

I don't use it now so I'll save some cash and skip ec.

 

[Skybound]

For the price, ya can't go wrong! How long till ya have everything delivered? With no immediate plans to tinker over here, I'm hoping to assist you if possible. Are you planning on soldering anything for permanent use? Don't get me wrong, you can definitely achieve reliability with breadboards, and they're as simple as plug and play, but if they get used a lot to experiment with different circuits, the grabbers begin to lose their grip and short circuits can occur.

 

[sleepless]

For the price, ya can't go wrong! How long till ya have everything delivered? With no immediate plans to tinker over here, I'm hoping to assist you if possible. Are you planning on soldering anything for permanent use? Don't get me wrong, you can definitely achieve reliability with breadboards, and they're as simple as plug and play, but if they get used a lot to experiment with different circuits, the grabbers begin to lose their grip and short circuits can occur.

No breadboard. All soldered.

Amazon says things will be here mid week.

I need to find a project box now.

I can't stop reading.

This thing could run my yard and garden inground systems as well.

 

[Skybound]

Your project can be expanded as far as you can afford and desire. You can control your entire home, everything electrical if you want. HERE is my project box, though it's only good for one board with no shields. I'm looking to upgrade to something larger as well, but it's a starting point and was cheap so I got it for now.

 

[Latitude17]

...Atlas Scientific makes a whole line of probes and circuits to work with Arduino and example code to get it up and running right away. I've had my eye on them for over a year now, but the high price has kept me away thus far. Hopefully after a harvest or two, I'll be able to start down that road.

Have you tried Phidgets? Atlas Scientific isn't the only player out there.

 

[Skybound]

I believe I heard of phidgets when reading all that I have regarding automation, but for whatever reason went another route. Is phidgets a microcontroller, programming language or actual sensors? I didn't see any sensors on the link you shared, plus couldn't find any direct links to them on that page.

 

[Latitude17]

I believe I heard of phidgets when reading all that I have regarding automation, but for whatever reason went another route. Is phidgets a microcontroller, programming language or actual sensors? I didn't see any sensors on the link you shared, plus couldn't find any direct links to them on that page.

Sensors and IO boards (all hardware), didn't see the menu on the left side of the page?

 

[Skybound]

I found it now, thanks for pointing that out. Do you know if products from other manufacturers can be used like my relays? I see they have relays, but they are very expensive.

 

[Latitude17]

I found it now, thanks for pointing that out. Do you know if products from other manufacturers can be used like my relays? I see they have relays, but they are very expensive.

As long as the voltage and amperage align, I don't see why not. I haven't used them for any hydro projects, but they do have a nice set of sample code for things like relays.

 

[Skybound]

As long as the voltage and amperage align, I don't see why not. I haven't used them for any hydro projects, but they do have a nice set of sample code for things like relays.

I think I might try the PHprobe & Circuit in my existing project as the combo is cheap and I like cheap. Can you tell me which file in the library has the example code? I need to transfer the code into the Arduino programming language to be able to use sensor. As far as using any other phidget stuff, I can't justify the cost, lack of support and inferior IO count. It seems the Phidget board is more of a computer like the RasPi, but running a grow room doesn't take excessive amounts of computing power, it's simply test and execute. There's not too much beyond those 2 conditions. Those issues aside though, I hope I can sub out the Atlas Scientific PH probe and circuit successfully.

 

[Latitude17]

The thing with the probes you should note are the "industrial" vs "scientific" models - one you can leave in the water and the other is meant for spot testing. They really do not highlight that fact enough on the various sites. Still, the Phidget industrial probe is substantially less than the Atlas model.

 

[Skybound]

In my system, the probe will only be exposed momentarily or short durations at best.

Edit - I wish phidgets had the EC probe instead of ph as the atlas EC probe and circuit is the more expensive of the two. Are you running any phidgets?

 

[Latitude17]

DFRobot has some components as well, few industrial however.

I have used components in the past from all three of the mentioned vendors, currently using the Atlas Scientific items.

 

[Skybound]

Which Atlas stuff are you using currently and which MCU do you use to read them?

I got some photos of my bloom plants. Bubblegum in week 7 is still recovering from early deficiencies. Shortly into bloom, I noticed my meters were acting crazy, but I wasn't able to resolve the problem right away. Everything is cool now, but the damage is done. A half decent mid bloom defol released a lot of light to the lower areas while increasing air flow. The humidity is generally 5% lower in A than B, I think because of the extra foliage in B, but that's just a guess. The Bubblelicious in B is in week 3 and has stretched out like a bastard. She crowded the light with roughly a dozen tops. I pulled the light back the remaining 6" but she just kept marching up. I foresee a lot of super cropping to manage all that canopy. Based on what I'm seeing in A, I will also need some kind of trellising or yoyos to support lagging branches. I really miss having space to walk around the plant to tend to it like I had in my last grow.

I'm still doing tests with the dosing pumps. There's a certain number in the code that represents how long a pump needs to turn to produce 1ml of solution. I'm finding that this number needs to change depending on the requested dose to remain accurate and this presents a little problem from a coding aspect. Still more testing is needed to be certain what course of action is appropriate.

The Bubblegum week 7

The Bubblelicious week 3

 

[Latitude17]

Which Atlas stuff are you using currently and which MCU do you use to read them?

Currently playing with the EZO™ Dissolved Oxygen Circuit on an Arduino UNO. I'm waiting for my Tentacle Shield to arrive before I spend too much more time on it.

The issues you are experiencing with your peristaltic pumps is interesting - I could probably help you work that out if you want to post the code.

 

[Skybound]

Sure, HERE's a link to my github. Also here is a photo of my test results. the number on top is the amount of milliseconds that is multiplied by the bottom dose value. All these different top numbers each represent a range of ml that said number will accurately pump out. I am thinking I need to write a struct, but I'm still learning those in spare time. I already have a bunch of arrays, but they work well when [x] determines the position. Note; this project requires using Blynk unless you want to supplement everything in circuitry.

 

[Latitude17]

900/40, 900/50, and 900/60 are all the same...? ...so I thought I was following your math description from above until I read that. A couple questions to get me caught up if you don't mind:

• How many ml are pumped at 1000ms, 5000ms, and 10000ms? How are you measuring?
• Are you compensating for different viscosities?
• Are your pumps pre-primed (no air to cycle out of the line)?

 

[Skybound]

900/40, 900/50, and 900/60 are all the same...? ...so I thought I was following your math description from above until I read that. A couple questions to get me caught up if you don't mind:

• How many ml are pumped at 1000ms, 5000ms, and 10000ms? How are you measuring?
• Are you compensating for different viscosities?
• Are your pumps pre-primed (no air to cycle out of the line)?

If you look through the code, you'll see that a dose request comes from Blynk as a float(). Traditionally, this is multiplied by the multiplier which itself is supposed to represent the amount of millis needed to produce 1ml. Previously, I pumped 100ml and counted millis while pumping. Add all the millis, divide by 100, and that gave me the original multiplier. Having done these tests 3 time for each pump with water, I knew viscosity would become an issue, so that was just my starting baseline if you will. The numbers pictured above are my findings when pump just GH Flora Gro. Those can be my baseline to test the other pumps as viscosity likely is different nute to nute. Yes my lines are completely primed, well the Gro one anyway, but I'm not testing the others yet till I resolve all issues with Gro, than I can duplicate the process 7 more times. For reasons unknown to me, different multipliers are needed to achieve accuracy. I would think a motor turning at a fixed speed for a fixed time would produce equally, but perhaps these cheap motors, or something in the circuitry is causing the motor speed to rev and slow, causing the output to be finnicky? Whatever the case may be, I see I need some "if" tests to interpret the desired dose and apply the appropriate multiplier to it. If you have an ethernet shield, you can launch the project on your phone with no hardware connected beyond the Arduino and shield to get a better grasp, but without actual pumps connected, it will be impossible to understand what I mean, but if I had to title it, I would call it an accuracy curve or band, like on a chart or graph.

Edit - I don't know why 40ml, 50ml and 60ml all are fine with a 900ms multiplier, but lower doses definitely need the higher multipliers to produce accurately.

 

[Latitude17]

It looks to me like either the power curve on your motor doesn't come up to full rotational speed until somewhere shortly after 905 or, more likely, the lines you are using are still flexing and haven't regained their shape until about that time (suction pressure, viscosity, etc.) - which would explain why the higher dosing doesn't vary.

Do you have some tubing that is a little more rigid that you could do a test with? Some silicones have a wide range of consistency and rigidity. Is your setup pulling upward over a larger distance that could add a gravity and siphoning component to your equation?