Thanks guys for the good wishes and always good to know there's interest.
@Slater what I'll do is talk through a general overview here and see where it goes in terms of starting/joining discussions to expand on specific points. Once I get the building part done I'll maybe start a slightly preemptive grow diary or something like that.
This post is going to be pretty long but hopefully someone will find parts of it interesting, and it gives an idea of where I'm at right now.
Firstly then, what my design priorities are:
1. Safety - Essentially a micro-grow is a humid box of reasonably high wattage DIY electronics which is plugged into the mains, left unattended for extended periods and every couple of days topped up with water. One thing I want to keep a close eye on start to finish is not burning my house down.
2. Stealth - Obvious one this - I don't want anyone knowing this box exists. To achieve that for starters it needs to look innocuous and be low-noise and low-odour. There's other considerations playing into that like ways of monitoring without opening the case etc.
3. Fun - It's going to take a lot of time and money this project and realistically it's going to take a few years minimum to justify the investment in terms of yields. If I'm not enjoying doing it then I shouldn't be doing it, simple as. This includes choosing parts that are interesting to work with, using a platform that allows for future modifications etc etc.
4. Ease of Use - Not automation necessarily, but I want my first couple of grows to be as easy and low-maintenance as possible for obvious reasons. Maybe start with super soil, for example. Part of this as well is fast grows - it's a lot easier to find 8 weeks with no travel plans than it is to find 12+ weeks.
4. Yield - Ok, I'm a new grower working in a small space and I'm not looking to break any records. But I still want it to work, that's the challenge. And more to smoke obviously... I want to maximise yield firstly through the best possible use of my space and hopefully later on by learning a bit about plants too.
And considerations and plan for the main elements of the design:
1. Case and Raw Materials - For case I'm going to go with a Fractal Define XL - it's about as big as they come while still fitting under my desk and a massive bonus is the covered front panel meaning I can mount switches/hygrometers etc externally without them being in plain sight. Lining I'm going to use a thickish layer of rubber first of all which will be good for noise suppression, pretty much smell-proof and also a reasonably safe, non-conductive and heat resistant lining if something goes wrong. I'll then line that with diamond mylar for reflectivity. Haven't bought any of this yet but I'm pretty much set on what I'll do with it.
A lot of my electrics, as well as lights and outtake fans, are going to be mounted at the top of the case so I’m intending to use one sheet of MDF at the top to mount everything to. That should be easy to work with and should hopefully be sturdy with just a couple of bolts visible.
2. Airflow - Obviously carbon filters are a must for ultra-low odour. Unfortunately, they generally need powerful fans and powerful fans tend to be a)large and b)loud. The compromise I've come up with is a unit comprising two Noctua 80mm fans side by side pushing air into around an acrylic box with maybe 1" of carbon and two identical fans pulling the air out the other side. I've built one of these that I'll have at the top rear of the case and my overall design has the capacity to add another at the top front. If I do that I can run them both lower, meaning the carbon has more time to do its work, and I also have the capacity for increased airflow in the event of temperature issues etc.
First thing you'll note is that this setup costs a damn fortune - in terms of airflow per £ this is an awful solution but it does have a few massive advantages. Firstly, using two 80mm fans instead of, say, one 120mm fan means it fits into a small corner and doesn't encroach much on the growing space. The fact that two such units fit nicely in my design is idea. Secondly, the quality of the fans - the Noctuas still produce a vaguely respectable static pressure despite being quieter than a dead man snoring. Absolutely no noise concern with these whatsoever, really amazing little bits of kit.
For airflow in, my current plan is to keep it passive but I'm not sure if maybe one small intake fan would be beneficial. Trouble is, I'm also not too sure how to test and measure that without spending silly money on pressure sensors. But passive will do for now...
3. Lights - This is the part I've been thinking about recently. I want to use LEDs because they run cooler (which in my mind makes them safer), they're more efficient, more flexible, more interesting to work with and generally just better in my estimation.
I'm reluctant to use COBs though because a lot of reports talk about light burn with plants too close - no good if I'm trying to maximise a small space. That’s something interesting on this forum though is that I’ve seen a few people growing well with COBs in seriously tiny spaces. I may consider a hybrid system maybe if anyone can convince me! Still open to ideas on this and I’m not sure exactly what kind of wattage will work for me either.
Plan a) currently is to use 18 1ft Samsung Q series strips with 6/8 at the top of the case powered by one driver and 10/12 on the side walls of the case on another driver. My main thought is that if I can have the same wattage of light coming from a variety of angles it will give better penetration and massively decrease risk of light burn. Also, the more strips I run the cooler I can run each strip meaning better efficiency and less heat.
I’ve been test running this week with a 25w driver (pulling 20w at my voltage) running 6 of these strips. This puts at max brightness, 0.3A through the strips when their rated safe max is 0.9A so there’s plenty of breathing room to upgrade to a 40W driver on the same circuit. I do like them running cool though as the heat produced is minimal – I’m using just a small C section aluminium as heat sink, which is nice and easy.
Question now is, do I keep the 25w driver powering the 6 on top and add 12 on the sides powered by a 40w driver (maybe 55W from the wall) or do I upgrade to a 40W driver for 8 strips at the top of the case and run 10/12 strips around the sides from a 60W driver (maybe 85W from the wall). Have to say the latter seems a bit like overkill but as you might have guessed, I’m a sucker for overengineering.
4. Temperature Control – Obviously airflow plays a big part of temperature control but Ialso have an idea in mind for controlling the intake air temperature, up or down by a couple of degrees which may make all the difference to react to weather fluctuations.
This is an idea I’ve never seen anyone use before but I’ve done some simple testing and it seems to work pretty well. Put simply, I want to have two heatsinks, one outside the case and one in the air intake channel. Connecting them are two thermoelectric panels, each wired to a 3-way switch so they can pump heat in either direction. Both run at 12V, one connects to a 2A supply and one connects to a 5A supply because I had suitable adaptors at those values lying around.
I’ve tested this and heating the intake air I can get +10 degrees C or more, which is of course total overkill. Cooling, I could get maybe -3 or -4 degrees C but in my test runs I was limited by the external heatsink being unable to cope with the heat that 7A was pumping. It would drop the temperature at first but after a while the external heatsink would heat up so much that it would effectively ‘back up’ and the temperature would rise again.
However, I know the principle works. If I install a thermoelectric system in the box I’ll use an active “external” heatsink (which will actually be inside the box but in a separate airflow chamber). I may not install this for my first couple of grows, I might just see how I get on at first. Really promising idea though and I’m quite excited that I’ve not seen anyone do it before.
5. Monitoring/Controls – I mentioned earlier I don’t really want to automate this unit. That’s because I want to be as close to the grow as possible while I’m learning the basics so other than light cycle being on a timer, I want to monitor temps, humidity, watering etc myself.
That said, I do want easy control and monitoring. I have a pair of cheap hygro-thermometers off Ebay – the probe kind thankfully so I can mount the displays to the front of the case (remember my choice of case has a covered front panel) and position the probes where I like.
The light dimmers I’ll keep inside I imagine as they won’t need adjusting often.
The switches for the cooling system can be wired easily to the front panel and I can probably have a controller for the external heatsink’s fan in the same place.
The main fan controller is a little Noctua PWM unit I already have, that again I can mount easily to the front.
I may need a separate controller for internal circulation fan(s), I haven’t really got a solid plan for that yet.
An additional thing I’m doing is mounting an old webcam to the MDF top panel – I’m going to keep it really simple and just hang the USB cable out the back of the case. I don’t plan on broadcasting but I can plug it in for a quick check during the day without opening the case is my theory.
EDIT - Where are my manners? I forgot pics or it didn't happen. Couple here of a prototype (empty) carbon filter, one picture of the lights as tested yesterday and one of the intake temperature control idea. What you can't see in that picture is the second (intake) heatsink and the thermoelectric panels connecting the two inside the cardboard.
we have a great lil micro section 
Hi and welcome to AFN 
