Digitally Controlled Propagation - A newb geeks approach

[IndianDreaming]

Hi All - First main post for me to meet 'n' greet and let you know what I'm up to. I've found the Corroboree to be an absolute weath of information and humour and would love to share my thoughts with you all for both flaming and fair share.

My interests: Chemistry, electronics, robotics, computing, plants, people, parties, psychonautism and generally LTUAE (Life the universe and everything...)

Being fairly new to gardening in general, I'm learning very quickly the meaning of 'patience is a virtue'... Especially with L.W which should be the new saying 'Yeah mate, its like watching peyote grow...' With that in mind, I've decided to combine a few interests to come up with a digitally controlled (cactus) propagation environment that meets the following goals: Cheap, low maintenance, eco friendly & efficient, fully automatic.

Below is the list of things I have already and a brief explanation (and some pics if i can get them in here):

2 x Shelves (bunnings $15 each on special combined into one) = $30

1 x 12v 80psi water pump = $50

7 x Dan fogger jets 7 lph = $15

1 x AM2302 Temperature + Humidity sensor = $6

1 x Computer fan (free - recycled, have a few)

1 x 4 Bank 5v controlled solid state relay bank (rated 240v 10A or 30VDC 10A) = $6

2 x Arduino uno 328's (a Dev board and a mini Pro board) = $20

5m x LED Strip (3014, 180LEDS/m, Lum 2070, 21.6W/m - Approx 90Lm/w, 5000k) = $80

Some plastic, clips, prop trays and little seedling thingo's

Still to come:

1 x Water vapour mister thingo like you see in those foggy fountains

1 x Digital soil temp probe and maybe a digital PH probe

Lots of x Time to whang it all together and write the control software

Eventually all this digital stuff will go into a fully fledged greenhouse, but at the moment, I have neither the funds to purchase one or the space... I think thats enough of an intro, I'll whang in some pics and would love to hear any comments - I'll also be chasing some cactus seed eventually, and have some items to trade such as Entada Rheedi seed, P. Harmala sproutlings, etc... But thats a little ways off.

p.s. Wasn't sure where this sort of post belongs...

Edited April 6, 2012 by IndianDreaming

[Bert&Ernie]

Awesome idea. And welcome to the forum what a Shame your not in WA xD

Can I ask how big your rue seedlings are cause I've tried to grow them and the get about 2 cm high and they just die lol hope fully you will have better luck with them

[IndianDreaming]

Hi jwerta, they're only little atm and now that you mention it, they're 2cm high and they're not looking as good as what they were a week ago... They seem to spring up, bend and grow sideways then sort of lay there and keep growing. Fingers crossed though, I cheer them on daily and hope they'll keep growing! I'm going to re-pot one today and see if it'll stand upright.

[Bert&Ernie]

try not to water them to much cause i think that's what i did to mine and they just sorta went all yellow and rotted

so how long has it taken you to set this system up?

[IndianDreaming]

Ah yes, I have pampered them, i'll dry them out a bit and see, one has rotted where the stem meets the dirt so I think you may be right. (How do i attach a pic in here now?)

I've been gathering elecronics bits for ages, had the shelves for about 2 months now, and just bought the roll of lights yesterday and am waiting for them to arrive. I have many projects on the go, so I don't really know how long it'll take to get the automation going, but as soon as the lights arrive I'll stick them in and see how they shine.

[Bert&Ernie]

to attach a pic go to more reply options and down the bottom should be attach files

[IndianDreaming]

 

P. Harmala - Not doing the best but hanging in there...

[Amazonian]

Welcome to the Forum IndianDreaming

Looks like you are going all out hey. It makes me look like an amateur.... I just whack some seeds in soil, water them , leave them outside, and cross my fingers,lol. My success rate isn't great, but what does grow is usually hardy.

Happy growing

Edited April 6, 2012 by Amazonian

[Sallubrious]

When they rot at soil level most times that is caused by what's known as "damping off" which is caused by a pythium infection.

Pythium does thrive in overly wet conditions.

For an enclosed system like yours you'd be best to use sterile planting mixes or at least pasteurised mixes to try and keep fungal spores to a minimum. Outdoors you have the benefit of UV rays to keep parasitic fungus in check, but indoors especially in an enclosed system it can get out of hand.

I prefer to keep things in separate containers for each batch to cut down on cross contamination but that is an awesome piece of work, I played around with PIC control back in the day so I can really appreciate what you've done.

Was there any reason you chose the 3014's over 3528 or 5050 strip ?

LED's tend to emit light over narrow spectrum (unlike the sun) so you could tweak that a bit by adding some different colours if you feel you need to down the track.

Here's a calculator to convert wavelength to Kelvin, just type in the colour temp and select Kelvin, then look at the wavelength below in nanometers.

http://www.ajdesigne...en_equation.php

A quick check on ebay for LED grow lights should give you an idea of the wavelengths the trendsetters are using these days.

Welcome to the forum.

Edited April 6, 2012 by SallyD

[IndianDreaming]

Thanks for the welcome

Amazonian: It's a bit over the top i know but fun,i had most of the bits already from other projects, I'm as amatuer as you can get when it comes to growing though.

SallyD: Damping off - I've read a little bit about that, is there anything i can do now to get rid of it or are my babies doomed? I'll google about and see what i can dig up. I've been recycling my mix when things don't sprout, I'm assuming thats probably not a good idea?

The PIC and AVR's today are fantastic, the Arduino has basically made it all plug and play. USB upload of the program, pluggable modules and a thriving community make it accessbile to a much wider audience. There's everything from flashing leds to fully fledged gps!

I chose the 3014s because apparently they're better heat dissipaters than the 5050. The general plan being that once the strip is mounted up in the undersides of the tin shelves, they'll transmit the heat through the shelf to bottom heat the trays above. The strip I got was the highest power I could find (21watt/m) and the best lumens per watt in an IP68 strip. I'll be really interested to see what temp the shelves end up after continuous running.

I've had a thought about the spectrum... Leds are far from ideal but I was thinking of playing around with some luminous paint/plastic pieces in the light path or on the walls of the shelves. There's lots of things like luminous transparent perspex, cellophane and glow paints, both translucent for painting directly on a few leds per metre, or plain luminous paint for reflecting off the walls. Do you know if anyone has experimented in this direction? I haven't managed to find any info on this approach. It'd reduce the light output a bit, and I'll need to find a way to measure the spectrum (digital camera and photoshop maybe), but I think with the right materials it just might work... we'll see anyway, I'll post my results when the lights arrive. It'll give me something to do instead of (patiently ) watching my plants grow...

[Sallubrious]

Yeah don't go recycling your seed raising mixes indoors, especially in an enclosed system. You could do it after microwave pasteurisation, but really for an enclosed long term system you should try and keep things as clean as possible.

I've tried oven sterilisation and it didn't work for me, I've read (somewhere) that toxic compounds build up at high temps that are detrimental to the plants. That's for mixes containing organic humates and such.

A sterile soil less mix would probably suit your situation better. As much as I hate chemical agriculture, I can't deny the benefits in some situations.

As fas as reflectors go, they are a good idea but they can only reflect the spectrum that is supplied to them. If the LED doesn't emit it the reflector can't reflect it. White or silver is the best as it reflects everything.

LED's have a bit of a range in the spectrum they emit so try to reflect it all.

I don't think a digital camera and photoshop will be accurate enough for what you want, there are too many opportunities for errors to creep in. I t would be interesting to see some calibrated results though.

LED's are manufactured to precise tolerances so the peak output kelvin is normally very close to the stated output, they dope the chips with a specific chemical blend for each colour so they are for the most part a known quantity.

Some manufacturers could take shortcuts and use cheaper doping chems so there could be errors, but I just trust the propaganda and it serves me quite well.

[IndianDreaming]

Sweet, I'll turf the failures out the back into the unimportant house plants - maybe they'll damp off and i can use the space for better plants :-)

With the reflective surface, if it's luminescent or fluorescent paint, it'll absorb the light and re-radiate it at a different frequency, usually a longer wavelength but wikipedia is telling me it's also possible for some double photon on photon action for a shorter wavelength emission... Pity we can't get hold of some radon paint like they used to paint on clock hands, that stuff glowed! I'll probably end up buying more leds and having a mix of colours, but I'm just starting out need to watch the budget, and still have some leftover for buying seeds/plants.

I think you're right about the accuracy of the digital camera approach, I'm not too sure how i'll tackle that. I did see a camera once that had a live spectral analysis on the display, I wonder if that's any good. It's a damn expensive camera so i'd imagine it might be ok... Will be fun to try anyway. Thanks for the tips and advice, with a case of damping off and some tiny little white bugs on my P. Viridis, I think i'm going to need all the advice i can get.

Speaking of little white bugs... are they likely to be spider mites?? I've tried a garlic and soap spray, and a 3% peroxide diluted 10-1 but the little buggers are still there! The plants are looking very sick, is there anything else i can try before getting into nasty stuff in can?? I'm hoping to get my hands on some nicotiana seed/plants and eventually use an extract of that, but these plants won't last that long.

Edited April 6, 2012 by IndianDreaming

[Sallubrious]

Sweet, I'll turf the failures out the back into the unimportant house plants - maybe they'll damp off and i can use the space for better plants :-)

With the reflective surface, if it's luminescent or fluorescent paint, it'll absorb the light and re-radiate it at a different frequency, usually a longer wavelength but wikipedia is telling me it's also possible for some double photon on photon action for a shorter wavelength emission... Pity we can't get hold of some radon paint like they used to paint on clock hands, that stuff glowed! I'll probably end up buying more leds and having a mix of colours, but I'm just starting out need to watch the budget, and still have some leftover for buying seeds/plants.

I think you're right about the accuracy of the digital camera approach, I'm not too sure how i'll tackle that. I did see a camera once that had a live spectral analysis on the display, I wonder if that's any good. It's a damn expensive camera so i'd imagine it might be ok... Will be fun to try anyway. Thanks for the tips and advice, with a case of damping off and some tiny little white bugs on my P. Viridis, I think i'm going to need all the advice i can get.

Speaking of little white bugs... are they likely to be spider mites?? I've tried a garlic and soap spray, and a 3% peroxide diluted 10-1 but the little buggers are still there! The plants are looking very sick, is there anything else i can try before getting into nasty stuff in can?? I'm hoping to get my hands on some nicotiana seed/plants and eventually use an extract of that, but these plants won't last that long.

 

I didn't even consider luminescent paints, fuck I was drunk last night, that bit went straight over my head . Maybe it would be OK, I don't know how it will go in terms of efficiency compared to a white or silver reflective surface.

Maybe the spectral analysis on the camera would be OK for what you want, I'd be interested to see the results.

I could only speculate as to what those white bugs are, but I'd want them gone before you get the setup in full production mode. Bugs are natures way of cleaning up sick plants.

I'd suggest to move the plants from that growing area and spray the area with some normal household insecticide maybe twice,once to kill the bugs and then again in a week or so to kill any eggs that might hatch. You want to break their breeding cycle and start pest free if you are going large scale.

Then when you start again try your hardest not to introduce any pests or eggs into the system. If you choose to use a soil based mix maybe put it in a plastic bag and microwave it until it gets too hot to touch and then give it another 30 seconds or so, that should kill any bugs and their eggs. the soil has to be damp for that work properly. Pests can be fine outdoors with natures checks and balances to keep their numbers down, but bring them inside and their populations can explode.

Good luck mate keep us posted.

[IndianDreaming]

Will do, congrats on the inebriated posts, they're informative and articulate even when you're slideways :-) Luckily only 2 plants are infested, they're separated from everything else. After a lot more reading, multiple coloured leds are definately the go, but in the meantime i'll do the luminescence thing. Apparently some paints can crunch the spectrum and emit 300% in their band by conversion of other spectra... wouldn't be more efficient than multi coloured leds though by a long shot.

[tipz]

Speaking of little white bugs... are they likely to be spider mites?? I've tried a garlic and soap spray, and a 3% peroxide diluted 10-1 but the little buggers are still there! The plants are looking very sick, is there anything else i can try before getting into nasty stuff in can?? I'm hoping to get my hands on some nicotiana seed/plants and eventually use an extract of that, but these plants won't last that long.

sounds like it may be mealy bugs to me IndianDreaming. http://www.cactus-mall.com/pests.html

welcome to the forums man. Very interesting little project there mate. Personally, Im with amz and reckon mother nature has it downpat allready. But sometimes there is need to create artificial conditions for stealth or maybe even just for the sake of experimentation. ( i have been partial to a few controlled indoor experiments myself in the past )

Im looking forward to seeing how you fare with this project.

kudos

[IndianDreaming]

Thanks tipz! Here's a picture of the little buggers. This is after i soaped and garlic'd em, which i now believe worked quite well since they're all on their backs. But as mentioned above, I didn't repeat the soaping a week later so more of the little buggers would have hatched... clever little things :-)

Yes, I think mother nature has it spot on. The shelves are just an interim till i can scale up to a greenhouse in the yard. The control system is my main interest there (with a side interest in spectra), once I get that going smoothly the sky is the limit.

What's in your garden?

[IndianDreaming]

Well, after letting this project fall by the wayside for a while, I've inched forward a bit. The lights are now attached to the shelves. Here's a few pics because they save me typing so much

• The chopped lengths of LEDs
  
• 
  
• The soldering of the ends (in series)
  
• 
  
• Attaching of the strips to the shelves with mould resistant silicone
  
• 
  
• The shelves with the lights attached
  
• 
  
• The shelves as they are now filled with all the stuff that they've collected since I took the photos and uploaded them
  
• 
  

 

The light output seems ok, but it's nothing special. I think it'll need more LEDs to bring it up to the point of a nice t5 equivalent.

As far as the controller goes I have two ideas in mind:

• A state machine with in's triggering outs and a table of acceptable ranges to cause things to happen (needs me to work out the best ranges).
  
• A neural network/genetic algorithm to learn the best output to turn on based on previous input levels and output states. (works everyting out itself but will take a while of running to learn to do so)
  

 

The inputs being humidty, temperature & light sensor and the outputs being an internal circulation fan, internal<->external fan, mister, automatic vent.

The state table is by far the easiest to program, but it means I have to hard code the rules for what happens when the temp is say 30 degrees and the humidity is 10 for all ranges of values. Not the easiest thing to do really. Whereas with the AI approach, the programming is a little fiddly, but the end result is that I set the humidity and temp that I want and leave it. It will randomly turn the fans ,misters, sprinklers and open and close the vent while sampling the humidity and temperature. After a quite a few cycles of sampling and doing random things, it'll 'learn' what works to keep it in that temperature/humidity range. After that, it'll open/close, spray, blow or whatever it needs to whenever it see's fit, and it will get better and better at it as time goes on.

If I go the AI approach, which I think I will (after testing with the state machine) - it'll be able to learn any new environment that the sensors and fans/misters/vents are installed in, in any climate, in any season. I may have my sights set high, but if I get off my butt and finish it, it'll be an awesome gadget