Building a Home Made LED

[repuk]

I'm seriously tempted to get far reds, reds and blues in 3W a couple dimmable drivers and build another ~120W HGL to go side by side with the all white... (this time sourcing the LEDs from china) I'll have ~140W with the white HML for 4,5 sq feet, while the most successfull crops I've seen had always > 40W < 60W per sq feet.. plus in addition to check all white - white + RB I would be able to "tune" the whole using the dimmers...

ah, temptations, temptations

 

[vukman]

If you will use active cooling (fans) a standard Al plate (2-4mm thick) will do the job.

Can you tell us what kind of LEDs did you ordered ? 1W ? 3W ? What wavelengths ?

Sure thing..no secrets here my friends...

First off, they are 3W........I am still missing the 380 and 730 due to no stock at the place I ordered from.

40-660--30-40Lm
20-630--80-100Lm
10-470--30-40Lm
10-440--25-35Lm
10-2700-3200K--190-210Lm
10-6000-7000K--190-210Lm
10-8000-9000K--190-210Lm

I got that order for 96.00USD delivered to Toronto....

Not sure of layout or ratio yet either but at least I'm getting the stuff together and then I can play around to see what's what.

Also, thank you very much for the info about the sheet aluminium. I was thinking of having to go that way but was still holding out hope that I can find a piece of finned extruded for a decent price..I've seen a few video's on Utube of DIY projects and the one I liked was two sheets....one for the chips and the other spaced a bit apart with threaded 'rod' and nuts to hold the fans and drivers.....

Will see how it goes but thank you for the interest and if you need anything, just ask.....

Got the chips from hanhualed......dealt with guy named Jero..

<add edit>,,,,,,,,,shipment sent Monday, received Tuesday afternoon!!

Good Luck

 

[tenthirty]

Vuk,

Go to home depot or the likes and get some square aluminum tubing, drill some holes in it for airflow. Should be cheap enough.
If it doesn't stay cool enough, just pop rivet some fins onto it.

Now just for giggles.

http://www.cree.com/~/media/Files/C...d Modules/XLamp/Data and Binning/XLampXPE.pdf

Check out the 2 wave forms that are shown on page 6.
Yes I look at the light spectrum as a wave and not discrete colors or frequencies.

What I find interesting is the white led waves overlap and would mix nicely. They would be coherent.

The colored leds show many peaks and valleys.

If I wanted to jam a radio receiver, the bottom graph is the signal/wave form that I would want to use.
The peaks and valleys make it very hard for the gain control system in a radio receiver to pick out the desired signal. It is being swamped by the adjacent signals.

If we mix the 3 whites, the relative signal is relatively consistent across the band and the gain control system would be in a steady state more or less.
The plant has to work less to tune the antenna pigments that feed the reaction centers.

Based on all the above gibberish,
My vote is for an all white fixture for your first try.....

 

[hempfield]

Sure thing..no secrets here my friends...

First off, they are 3W........I am still missing the 380 and 730 due to no stock at the place I ordered from.

40-660--30-40Lm
20-630--80-100Lm
10-470--30-40Lm
10-440--25-35Lm
10-2700-3200K--190-210Lm
10-6000-7000K--190-210Lm
10-8000-9000K--190-210Lm

I got that order for 96.00USD delivered to Toronto....

Not sure of layout or ratio yet either but at least I'm getting the stuff together and then I can play around to see what's what.

Also, thank you very much for the info about the sheet aluminium. I was thinking of having to go that way but was still holding out hope that I can find a piece of finned extruded for a decent price..I've seen a few video's on Utube of DIY projects and the one I liked was two sheets....one for the chips and the other spaced a bit apart with threaded 'rod' and nuts to hold the fans and drivers.....

Will see how it goes but thank you for the interest and if you need anything, just ask.....

Got the chips from hanhualed......dealt with guy named Jero..

<add edit>,,,,,,,,,shipment sent Monday, received Tuesday afternoon!!

Good Luck

I also looked for a single piece of extruded aluminium for almost half a year and I couldn't find any piece of my desired size ( 40X80cm or 16"x32") so I decided to build my onw.

So I used a 3mm thick aluminium plate of 40x80cm on which I "glued" 32 pcs of "U" profiles ( 8x8x8mm). I glued the profiles with white thermal paste and I fixed them in place with pop rivets. It was a little tricky to do that work but I was very satisfied by the final product : a solid , massive heatsink (with fins).

 

[vukman]

Vuk,

Go to home depot or the likes and get some square aluminum tubing, drill some holes in it for airflow. Should be cheap enough.
If it doesn't stay cool enough, just pop rivet some fins onto it.

Now just for giggles.

http://www.cree.com/~/media/Files/C...d Modules/XLamp/Data and Binning/XLampXPE.pdf

Check out the 2 wave forms that are shown on page 6.
Yes I look at the light spectrum as a wave and not discrete colors or frequencies.

What I find interesting is the white led waves overlap and would mix nicely. They would be coherent.

The colored leds show many peaks and valleys.

If I wanted to jam a radio receiver, the bottom graph is the signal/wave form that I would want to use.
The peaks and valleys make it very hard for the gain control system in a radio receiver to pick out the desired signal. It is being swamped by the adjacent signals.

If we mix the 3 whites, the relative signal is relatively consistent across the band and the gain control system would be in a steady state more or less.
The plant has to work less to tune the antenna pigments that feed the reaction centers.

Based on all the above gibberish,
My vote is for an all white fixture for your first try.....

Heya bro.. I've never considered anything you've said 'gibberish'..........

As you know, everything we've learned and tried is always in flux. The reason I got so many 660's was due to earlier thinking but then seeing people's results, I got the whites and now, believe it or not, I was going to make the first test mostly white light with blues and reds used as extra 'punch' power.
I still want and believe the 660's have a major role to play but waiting for the 730's to combine them with.....The 380's, well...that's in 'flux' as well...

 

[vukman]

I also looked for a single piece of extruded aluminium for almost half a year and I couldn't find any piece of my desired size ( 40X80cm or 16"x32") so I decided to build my onw.

So I used a 3mm thick aluminium plate of 40x80cm on which I "glued" 32 pcs of "U" profiles ( 8x8x8mm). I glued the profiles with white thermal paste and I fixed them in place with pop rivets. It was a little tricky to do that work but I was very satisfied by the final product : a solid , massive heatsink (with fins).

That is a lot of work.. Tell me friend...is that now a passive cooling unit or do you have some sort of active cooling for it. Not necessarily a fan attached but rather a fan say in a tent or something blowing air across the fins?

I was watching a Utube video where the person had 2 pieces of thin aluminium and spaced them apart and then cut holes and attached the fans to blow air across the piece with the chips on it.

I'll have to find that video again and link it...

Thanks again for letting me know about the issue with finding extruded finned Al.

 

[repuk]

Vuk,

Go to home depot or the likes and get some square aluminum tubing, drill some holes in it for airflow. Should be cheap enough.
If it doesn't stay cool enough, just pop rivet some fins onto it.

You can also weld aluminum with durafix with a simple torch...

Being the heatsink cost a good chunk of the total costs a cheap and efficient way to build a DIY heatsink would ease the costs for all...

 

[hempfield]

That is a lot of work.. Tell me friend...is that now a passive cooling unit or do you have some sort of active cooling for it. Not necessarily a fan attached but rather a fan say in a tent or something blowing air across the fins?

I was watching a Utube video where the person had 2 pieces of thin aluminium and spaced them apart and then cut holes and attached the fans to blow air across the piece with the chips on it.

I'll have to find that video again and link it...

Thanks again for letting me know about the issue with finding extruded finned Al.

It's not a passiv cooling unit because I want to be sure that this panel never comes hot. But event with the fans off, the temperature remain low (it feels just warm when I touch the panel, maybe a little warmer in center - because the upper fiberglass sheet that hold the fans is actually blocking the natural convection of warm air).

In the post https://www.icmag.com/ic/showpost.php?p=4825653&postcount=87 you can see both the front (aluminium sheet) and the back of the panel. Unfortunately I did not took pictures of the sides but I will do tonight and I will post them here for you (to see how I did the heatsink).

I have seen the movie you talk about it and I know why that guy used to sheets of Al - to increase the active cooling efficiency by using Coanda Effect , which cause the air flow to stick to the metal surface, causing a better cooling.

We you use active cooling you in fact want to extract as much heat as it possible and this is done in two ways : increasing the amount of air blown over the hot area or by increasing the total exposed surface of that area, which mean using heatsink with fins. The fins also direct the air flow, slowing down turbulence caused by fans.

I was inspired by the old heatsink models used on the Pentium 3 processors, like this one :

As you can see, on the main Al board a very thin sheet of folded Al is glued, increasing the exposed surface. I could not find any thin sheet of Al to make a similat heatsink (I could use Al from beer cans but I wanted a more professional look for my panel ).

If you want you can order a custom sample of folded fins for your heatsink here : http://www.tuckereng.com/custom-folded.html

Take a look at this article to read more about heatsink cooling.

I know that was a lot of work, but I don't care if in the final I am pleased with what I did. A special glue for aluminium would be nice, but I was also on quite low budget for thir project and is also very hard to find special products like this on my country.

L.E. In this picture you can actually see the 'heads' of the pop-rivets on the front side :

 

[VladTheInhaler]

I was looking at various Chinese manufacturer's leds on ebay and aliexpress and came to a stunning conclusion that they're currently too expensive for my needs. Now if I was building, say, 15 lamps.... economy of scale would kick in.
I'm recovering from mono, seems to me that for last 2 months the only thing I've been doing is looking at various led prices and specs.

 

[repuk]

hempfield, there's a lot of love in that work!!! congrats!

You say that w/o fans the temp was 50C? with almost 300W!!

Durafix is used for low cost repair of automotive pieces, specially broken sumps so getting some rods isn't too expensive. But I doubt it will be cheaper than your solution (thermal paste and pop rivets).

what did you use to drive the LEDs? I see only a power supply...

waiting for those pics...

 

[hempfield]

hempfield, there's a lot of love in that work!!! congrats!

You say that w/o fans the temp was 50C? with almost 300W!!

Durafix is used for low cost repair of automotive pieces, specially broken sumps so getting some rods isn't too expensive. But I doubt it will be cheaper than your solution (thermal paste and pop rivets).

what did you use to drive the LEDs? I see only a power supply...

waiting for those pics...

Yes my friend, all I do I do with love.

I didn't use any commercial drivers because I was on low budget for this project and in addition I wanted to be built entirely by me.

I use a 48V power supply ( 360W max load) and I use some DIY current limiters. We want the current to be limited on the LEDs, not forced into them

So I inspired myself from this instructable and the results are very good at a fraction of the cost of dedicated drivers.

I have 1W and 3W chips, driven at different current values : 350mA, 600mA and 700mA .

You can see the 11 'drivers' attached to the back cover with some 3mm bolts. Each 'driver' has a very small heatsink because I could not bond them to the main heatsink due some electrical path restriction (the drain pin of the MOSFET transistor is conected to the thermal pad, and the voltage on that pin is used for feedback loop. If I connect all that pins together - by bonding MOSFETs to heatsink - I have a single voltage reference and even some short circuit and this means failure).

The problem with this 'drivers' is their efficiency: if the supply voltage is grater that the total voltage drop on the LED chain + 0.6V on the feedback resistor , the difference is converted on heat by the MOSFET and this mean a greater heatsink should be used for each individual MOSFET. So the LED chain must be as long as possible for minimal power loss.

If we consider a maximul forward voltage on each LED to be 3.4V (for blue and white ones) , it means that for a 48V power supply we can have 14 LEDs on a chain : 3.4*14+0.6=48.2
I set the power supply at about ~50V so the voltage difference is ~1.8V. And the dissipated power on the MOSFET is 0.35*1.8=0.63W dor 1W chains and 1.26W for 3W chains. The total dissipated power on a chain is (0.6+1.8)*0.35=0.84W for 1W chains and 1.68W for 3W chains.

The comercial drivers have more efficiency but in the same time are more expensive.

In my previous version of the LED panel I used 12V power supply and the power loss was big because I could not put more than 4 LEDs in a chain.

At this moment my panel has 2 switches (one for blue and one for red) which gives me the ability to use only blue, only red or blue and red together.

But I want to modify a little bit that "drivers" and to add a PWM controler built with a IC 555 , for purpose of dimming and eventually to add a slow start/stop ( like a sun rise) .

 

[repuk]

Thanks for all the details!! Would you share the final circuit you used? How much did that PSU cost? Did you consider using a small PC one using the 12V outputs in series?? (dunno if that would be feasible)

 

[vukman]

AGreed..thank you so very much in disclosing everything you did there... That is what I like about DIY'ers..........there are no secrets withheld...it's all about the field getting better, not the individual....

Of course, if a person is going to make a living at this, then I would say sure, I understand why they would not reveal everything especially spectra and ratios...I mean that is what the meat and potatoes is off all this...figuring out the 'optimum' design...

As for myself, when I start going..that is when money allows me to get everything together to get going, I'll be giving full disclosure as well because I'll definitely be needing help!!! LOL.....

 

[hempfield]

Take a look at this album for a fresh lot of images of my LED panel : https://www.icmag.com/ic/album.php?albumid=40244

I have no secrets regarding this panel and I will share that info with you guys. At this moment I am a little busy but tomorow I will give you more details (all I can remember) and of course you are free to ask

Cheers !

 

[hempfield]

Builind a Home Made LED v3

Builind a Home Made LED v3

Hi guys,

As I promised, I will try to give as much information as possible about the LED grow panel I have built (last year).

This is the third version (and the most complex) of my LED panel. It was build partially with new and reused LEDs (from the previous 2 version).

I spent much time reading and thinking for this project in order to cover this aspects:
- low cost ;
- spectral composition ;
- light footprint ;
- light distribution ;
- heat dissipation ;
- power supply ;
- maneuverability ;
- reliability.

In my first version of LED panel I used 90 pcs of 1W chips :
- 50pcs of 660nm (deep red);
- 20pcs of 625nm (red);
- 20pcs of 460nm (blue) ;

In the second version I build two panels containing only 3W chips:
- 56pcs of 625nm (red) ;
- 10pcs of 460nm(blue);
- 12pcs of warm white ;
- 12pcs of cool white;

For the third version I bought some new 3W chips:
- 20pcs of 660nm (deep red);
- 10pcs of 430nm (royal blue) ;

Having all that LEDs available I had to figure how to combine them in a single led panel ( for maneuverability). My biggest concern was related to the heat dissipation so I looked for several types of heatsinks (or objects/materials that I could use as heatsink).
I found an article somewhere on the internet about heat dissipation on high power LEDs and I understood that for passive cooling 16~20 cm2 (I do not remember the exact value ) are required for each watt.

I decided to use 200 LEDs (1W and 3W combined) and to have 16cm@ of heatsink surface for each LED. I create a layout of 10x20 LEDs, which gave me the final size of the heatsink : 40x80cm ( roughly 16x32 inch).

So where can I find a heatsink of that size ? I looked everywhere and I could not find a heatsink of that size , or even half of that size.

In the end I stopped at two options :
- to use rectangular Al profiles bonded together somehow (with some threaded rod or something like this) . I found some 40x80mm Al profiles on 4 or 6 meters bar, but it seems a little overpriced and I always had in mind the "first amendment" : 'low cost'
- to use a sheet of aluminium, 3mm thick . I already had a big sheet of Al in my workshop so I had to find a way to cut it to the desired size. This was done with relative low cost (actually a couple of beers) with a hydraulic press.

But surface of the Al sheet was too smooth and I was afraid that the heat dissipation will not be very effective so I had to find a way to increase the surface - by adding fins. It's a simple and effective idea, but extremely complicated to solve in practice. How can I do this, was my question. I need a lot of fins, and I need to make air flow through them.

My first thought was to use Al sheet from the beer cans, but after spending a couple of hours trying to cut, to straighten and to fold the sheet to obtain the folded fins, I came to the conclusion that is not practical and the final look will be far from professional.

So I went to a second idea, to use aluminium U profiles bonded to the big Al sheet. I found some 8mm profiles at a hardware store at a reasonable price (always have the 'low cost' idea on my mind) and the guy at the store was so kind and helped me with cutting of that profiles at 80cm.

I drilled the U profiles (24pcs in total) and the Al base plate with a 3mm drill, I greased the base of the profiles with some (cheap) white thermal paste and I firmly bond them to the base using 3mm 'pop-rivets'. This phase took me several day to complete ( time was a restraint also for me) but the final heat-sink was quite impressive and it looked almost like an extruded one.

Now that the heat-sink was complete I could pass to the second phase, building the electronics.

On the previous two versions I used a 12V power supply recovered from and 1U IBM server ( low profile, high power and efficiency, extremely stable voltage) but I encountered some problems :
- at almost 300W powered by a 12V power supply, the sucked current is close to 25A, which can be very difficult to transport via a regular copper wire. The solution was to use a bundle of 2.5mm wires, but this was not a professional solution ;
- because I don't use current drivers, but instead current limiters, I have some power loss on each circuit. The more LEDs you connect in series, the smaller the loss.

So the single solution was to use a power supply with a higher voltage, but safe enough. Working with DC, the highest safe voltage for human is 48V so this was the limit. Fortunately this voltage is quite popular and was not to hard to find a switching power supply on the Ebay, for about 35$ (incl shipping).



In almost 2 years of working with high power LEDs I did not had any single chip burned or failed, but I want to be sure that my panel will not became useless if a LED decide to suddenly die (I also read some posts of led light owners that were pissed off because some part of their units does not produce light anymore). So I decided to create several LED chains (11 in fact) and to evenly spread LEDs on the surface of the heat-sink to decrease the impact of one chain failure. This decision cost me more hours or soldering, but I think it was a good decision.

In the same time I took care to have different wave length chips evenly spread across the panel, and after one day of playing with colors on 'Paint Bursh' I created the final pattern which can be seen in the following picture in which the light red means 625nm, dark red 660nm, light blue 460nm , dark blue 430nm and white means warm white:



In addition I want to have this panel working in 3 modes: only blue, only red and blue and red together, so I also had this in mind when I created the pattern.

That gave me 11 LED chains : 4 blue (including some 1W red and 3W white) and 7 red, each chain controlled by a ON/OFF switch.

As a current limiter I used a very simple (and cheap) circuit (also named feedback loop) inspired from this Instructable.

On this image you can see how simple can be this circuit :



I will explain how this circuit works :

The resistor R1 determine opening of the MOSFET transistor T1. The current goes through LEDs chain, via T1 and resistor R2. When the forward voltage on the R2 achive ~0.6V, the transistor T2 opens and the gate of T1 is gradually connected to ground via T2, which cause the T1 to close, causing a closed feedback loop.

Because the voltage drop on the Base-Emiter of T2 is always 0.5-0.6V, it is easy to calculate the value of resistor R2, to get the desired current through LEDs chain. In the above picture you can see a table with some calculated values of R2 for different current values.

Because there is some voltage drop on T1 , this means that T1 will generate heat which must be dissipated. A small heat-sink (3-5cm2) will do the job pretty well.

I suggest you to test this circuit prior mounting it on the final assembly and to measure the voltage drop on the T1 . If it is greater than 2.4V I recommend that you add an extra LED to the chain. In this way the current will decrease a little, but this will extend the life of the LEDs and decrease the heat produces by them. In that case the T1 will work only as a over current protection and now power will be lost on it.

Some math :
In average, the blue and white LEDs have a voltage drop somewhere between 3.2-3.4V and the red and yellow ones have 2.2-2.4V, but this value depends from LED to LED.

Having a 48V power supply , the maximum numbers on a chain can be calculated with the formula :

N=(Vcc-0.6)/Vf, where Vcc is the voltage of the power supply and Vf is the forward voltage of a LED.

Assuming we have a Vf=3.3 (average) that give us a number of 14 LEDs that can be connected in a chain .

(48-0.6)/3.3=14.363636...36

Of course we can have just and integer number of LEDs, so the N will be 14 in that case.

Doing the math again, the forward drop on the T1 will be :

V=Vcc-0.6-N*Vf = 48-0.6-14*3.3=1.2V

The dissipated power in that case will be :
- for 1W chips : P=I*V= 0.35*1.2=0.42W
- for 3W chips : P=I*V= 0.7*1.2 = 0.84W

The lower the voltage drop on T1, the lower the power loss and heat dissipation on this transistor.

For R2 and adequate resistor should be used, due the high current traversing it. It must be over-sized in order to keep it's value on time.

The power dissipated on R2 is :
- for 1W chips : P=0.6*0.35 = 0.24W
- for 3W chips : P=0.6*0.7 = 0.42W

A value of 0.5W or 1W should be enough for R2.

But as you may know, resistors are available on a limited range on values ( measured in Ohms).
The calculated value of R2 is :
- for 1W chips : R=V/I=0.6/0.35 = 1.714 Ohm
- for 3W chips : R=V/I=0.6/0.7 = 0.857 Ohm

We can use several resistors connected in series or parallel to get a value close to the calculated value, but in general is a good practice to use a slight higher value, which cause a decrease of the forward current, having the effect of long lifespan of the LEDs and less heat.

Now that the electronics was solved, I had to bond the LEDs to the Al sheet. For this purpose I used the same cheap white thermal paste (as for the U profiles) and some universal epoxy glue (Bison in my case). I put a small drop of thermal paste on each led (they came already soldered on 2cm star Al heat-sink) and pressed against the Al sheet while rotating it slowly until I felt a strong grip. After that I put 2 tiny drops of epoxy glue on 2 corners of the star plate (actually on the small holes designed for bolts) and a few hour later the LEDs were extremely well bonded on the big heat-sink.

After the epoxy has dried, I started to connect the LEDs using thin insulated copper wires. This phase seems to me to be endless and at some point I thought that I will never finish. But I finished after several hours of intense soldering, spread over 3-4 days.

Because I wanted to add active cooling using fans, I need to add a second plate, to create an inside for more effective air flow.

At the beginning I wanted to use 3 fans of 15cm powered at 12V, but this would require an additional power supply, which was not accepted by me. So I thought to use and additional fan and to connect them in series straight to the 48V power supply. Again, a new issue has raised : the fan must be quiet, almost impossible to be heard, so they must be powered at a lower voltage. Doing some math, I came to the conclusion that 6 fans of 80mm connected in series will be the best solution. 48V divided at 6 give 8Volts for each fan, which mean, in theory, that each fan will run at 66% of it's speed (actually they run a a higher speed because the speed is not proportional with the voltage and I adjust the power supply to deliver ~51V instead os 48V).

The fans and connected to the power supply via 2 diodes trough 2 switches, so they run whenever one of those switches are turned ON.


For the back cover I used 3 fiberglass sheets, 1.5mm thick, connected to the main heat-sink with 5cm long bolts and some spacers. Also I installed 4 hooks (2 on each side) for hanging.


This LEDs have a wide angle of view (120-140 degrees) which is not desirable, because the penetration into the canopy is not to good. To change this situation I ordered 200 lens of 45 degrees. I believe that 60 degrees lenses could be a better choice , but at that moment I couldn't find so many lenses on the same seller at a decent price.

I glued a lens on each LED using the glue gun (by putting two small drops of melted plastic on the edge of the LED and the pressing gently on the lens against the LED until the plastic hardens). This was the less professional step I did for this panel but I had no other option.

And that my friends was the story of 'Building a Home Made LED v3'.


Peace !

 

[repuk]

Whoa hempfield!! thanks for such a detailed post!!! including nice hand drawings!

I assume you did grow with all versions, what differences did you notice between each design iteration?

For that price that PSU is definitely more worthy than a PC one!

With the first design constraint in mind I can assure you solved the challenges with tons of creativity and elegantly!

I have read the same rule of thumb per each watt, but 2cm (I'd swear cubic) per watt, not 20...

 

[vukman]

agree fully with you there repuk.......definitely a major bonus to all of the DIY community who might be a little wary of getting into this.....myself included since I've done mostly large scale construction work most of my life. I've never undertaken something like this where I might have to look through a magnifying glass to solder!!!!! LOL....the eyes just ain't what they used to be!! ahahahha

Anyway,,,,thank you again so much hempfield and be ready for my questions once I get started,,,,..........

 

[hempfield]

Yes, I did grow with the all 3 versions .

The first version was used together with some CFLs, being more like an experimental grow. The yield was not impressive, the buds were a little fluffy but the potency was quite high. I have to mention that this plant (a SuperSkunk , regular seeds from Sensi Seeds) was grown on a very small pot (2.5L if I remember correctly)

Here are some pictures :



In the next post I will show pics from the second version.

 

[hempfield]

V1+V2 (first grow)

V1+V2 (first grow)

Ok, so here are some pictures from two separate grows done using version 2.

The first one was kind of a failure because 2 major factors :
- genetics ( SuperSkunk femised seeds from Sensi Seeds, indica phenotype, low yielder, average potency);
- as you may see on the picture this grow was done in hydro (DWC) but in the middle of flowering period my pH tester failed and stressed the pant very hard with to low or to high pH values . It was my first DWC attempt and I learned my lesson, but now is my favorite way of growing.

Actually this was a transition grow (started veg under V1 and finished flowering under V1+V2).

Unfortunately I don't have any bud porn pictures on this grow :

 

[hempfield]

V2

V2

This is the second grow under the V2.

It was done in summer, on extremely high temperatures (35 degrees Celsius or higher) and I was must of the time away from this ladies so I did not had the opportunity to take to much pictures.

The plant on the left is a personal cross of Skunk#1xSuperSkunk, the one in the right is a clone of SuperSkunk from the previous grow.



The plant on the left was the smoothest weed I ever grow and smoke. I could inhale an entire joint from one hit without any discomfort on my throat.