Minimalist Growing


@wsnook Here is the post you requested on how I do my growing.

First the disclaimer: I am very sloppy in what I am currently doing, and I have had only one ‘successful’ batch.

As you can see from the picture, I have two set-ups. One is a coffee jar with an airstone in the bottom, the other is a bus tub with a small pump that flows water up and onto the base of the plants.
Seeds are started in .75 inch (cut down from 1.5 inch cubes) blocks of rock wool.
Growing is in 3 inch plastic net pots, filled with aquarium gravel (I need to find something other than the gravel).
I started with 5 LED strips (recycled from refrigeration cases), but these have proven to be under power; and lacking easy access to more, I have suplimented them with GE Bright Stik (100w equiv) in a metal reflector. The strips by themselves were about 108 PAR at 3 inches, 92 at 6 inches and 38 at 12 inches. The GE bulbs were 575 at 3 inches, 182 at 6 inches and 48 at 12 inches (this was a quick and sloppy reading with a borrowed PAR meter). LUX (at 6 inches) is about 4,000. If starting over, for quick and cheap I would go with several GE bulbs ($7 each), or upscale to a CREE XLamp LED, or possibly one already in a module. I have one and it came in at 3,600 PAR at 3 inches!!!
With just the 5 strips my lettuce was very leggy and poor roots. I upped the light and now like the results.
This is just the house humidity, no supplementation: 35% to 40%. Leaves near the circulation fan can dry out on the tips.
I started with Jack’s 20-20-20, but thought it was part of my initial problem (poor roots, leggy, …) and switched over to Jack’s hydroponic plus calcium nitrate. Now that the lights are going, I don’t think there is that much difference. Mix according to directions. I have 1 qt in the coffee jar and 4 gallons in the bus tub.
I empty everything every 2 weeks (and clean things out) and make up a new solution. I need to check EC and pH sometime :wink:.
I punch a 3/16 inch hole in the rock wool and drop in a seed. You see two seedlings sitting in the tray near the lights. I keep them there for about 10 days before moving to the containers (per Cornell instructions). Right now, they get the same light (18 hour on/6 off) as the other plants. I start fertilizer after they emerge. Doing them this way I have to be careful that they don’t dry out (damage the roots).
After the 10 days I put them in the pots. I need to find something other than gravel, as the transplanting seems to damage the roots and put them in shock for a couple of days. I may need to not be so cheap and use the full rock wool cube. I don’t know exactly how long they take to mature. The one in the coffee jar was struggling under the old lights, and took off with the additional light. I ripped off some lower leaves for a salad, so there is no ‘maturity date’. I am guessing they take about 6 weeks. Looking at the coffee jar, I may need to get out the balsamic vinegar tonight.

I hope this is of some help to you.

MVP - Product Design
MVP - ppm levels and ph of water

Hello, Webbhm,What is the height between the light and the greens?
6 inch is the best distance? It seems so complex.

And I see the picture you provided, using white LED light? Regular LED lamps commonly used for lighting are NOT well suited for growing plants.

Below chart you can find the spectrum suitable for chlorophyll.

Below picture form 第3回植物工場ワーキンググループ.


From the lights to the base of the plant is right at 6 inches.


That’s great. Your description of PAR for the GE bulbs is interesting. I had my eye on some of the cheap Philips LED bulbs before I bought the grow light, but I didn’t have any way to measure or calculate the PAR I’d end up with.

I’ve been wondering about the math to convert visible light spectrum charts with a lux or lumen spec into PAR. It seems like it ought to be possible once you digitize a decent approximation of the spectrum chart from the marketing photos. Figuring something out for that is on my todo list.


I think this is an OpenAg project (for the MVP) , as only one person needs to do it, and do it only one time. The PAR sensors are not cheap ($300), but you only need one for a limited time to calibrate bulbs. I may be able to borrow one again, but I suspect MIT has one and could do this.
The easiest way would be to identify a set of MVP lights, then do a careful measure of the PAR and LUX an various distances. If we want to get fancy, we could create a conversion calculation based on the two curves for any given bulb. Otherwise, this becomes a lookup table for a quick guestimate of PAR based on a cheap LUX sensor. It would also be nice to get a spectrometer reading for each light. If you really want to go crazy, re-run all of these readings at different voltages (dimming of the light), and over time as the light deteriorates (in 10 years).
Given this data, any light could be correlated to spectral setting of the fancy PFC led array, allowing for combining the research data. The MVP just wouldn’t be able to do research with light spectrum as a variable. The MVP should be able to give just as good of research data as the PFC, just limited as to the scope of research, as some variables would be ‘fixed’.
This should all be meta-data about the lights that is available for anyone to use.

I built a new house last year, and it is almost all LED lighting. I spent a bit of time looking at bulbs. The real cheap ones don’t do what they say (either not the lumens, or they give off a lot of heat), you have to go with good name brands. I just looked up at my ‘can’ light and noticed it has a COB (chip on board), I may have to unscrew it and see what LUX it is giving off.
If you are willing to build your own, available components are unlimited; though I am a bit biased in thinking CREE has the best selection, price and documentation (disclaimer: I do not own stock in this company). I may be surprised, but I think we will quickly find that tweeking spectrum does not give that huge of (cost effective) benefit. Once the basic spectrum is identified, I think we can find a component that matches, and build a cost effective light that most of us can use. Those who want to do exotic research (and are willing to pay for the instruments) are welcome to it.

Light Panels specifications for PFC

Yeah, I like the sound of that. @Caleb, this is just a thought, but you might be able to give the community a big boost by having somebody in your group write a paper on PAR including:

  1. Links for finding background information on PAR and how much light different common crops need
  2. PAR measurements of the sort @webbhm described above for a sample of widely available brand name LED bulbs at common color temperatures (3000K, 5000K, 6400K, etc.)
  3. Raw spectrometer data for each of the sample bulbs so people can have a go at doing their own calculations for red vs. blue recipes.
  4. Recommendations on how construct a lighting configuration that will likely perform well, including an opinion about what “well” ought to mean.

[edit: Related threads: this one about PFC v1 lighting has a discussion of PAR and light timing cycles. This one about COB lighting has a discussion of quantum meters vs. spectrometers and converting between lux and PAR.]


Here is a good start on the topic of lighting. Oops, just realized this is the same one you mentioned.

I would suggest rather than some of us make suggestions of lights to test that we are using or are considering using (community input!!). This might be more useful than asking @Caleb to pick some,
My picks are the the GE Bright Stik and a Cree COB. @rbaynes: I believe the OpenAg LED array is from GE, do you know who interfaced with them for it, and could they talk GE into doing the research and giving us this data (which they probably already have). It would be good to see if they have any suggestions (number of bulbs, spacing) for making it similar to the PAR of the array. I need to research which bin of the Cree would be the best fit, I have a lead on this.

@rbaynes - side issue. If we produce a MVP, and intend it for school use (lower grades), I think there should be a STEM curriculum developed. There is already some rumbling about this; but I would like to see the projects formally joined.

Another aside: A major foreign company I am not allowed to name has a formal definition of ‘mature’ as a 100g head of lettuce. ‘Baby lettuce’ is defined as a 50g head. This works for me until something better comes along.


@webbhm I don’t have the details on where we source the LEDs from for the PFC, but I think you are correct. I have followed the PAR discussion and I don’t think we have any measured values of the different lighting sources yet. That is a good side project all on its own.

For the MVP, we probably want a simpler/cheaper grow light like the two you are suggesting, right?

Regarding curriculum, we have several pilots going on with schools in the Boston area using the PFC 2.0. I’m hoping that a result of them will be some teaching materials that can be further refined for various levels. We do have a teacher part time in the lab who is leading the pilots and I’m sure curriculum is part of the end game.


@wastonchen No argument that white LEDs may be wasting energy. As long as it covers the photosynthetic spectrum it will do a good job of growing a plant. As far as waste; if a $7 LED is only 60 PAR efficient, it still saves me a lot of money over a $400 or more custom growing setup. I am not trying to outfit a whole vertical farm with these, so the efficiency issue in not a significant concern.
Please remember the goal of a MVP (minimal viable product), to see how economically we can create a simple, small growth chamber out of commonly available parts that will produce meaningful research data.


@webbhm Hmm, you are right.
$7 for GE LED white bulb!
But I can buy the Grow Light Bulb(PAR 38) in $6 in my location.


@wastonchen Can you give me a specific brand, model number, and where to buy this? We need to look into various options, and if you can come up with a better bulb it would be good to check it out.
You say it is PAR 38; what is the distance from the bulb at which this was measured? The GE Bright Stik was giving me (quick rough measurement) 575 PAR at 3 inches, 182 PAR at 6 inches and 48 at 12 inches. We need to get beyond advertising and look at numbers to determine the best option.


No Ad sure.

I am in GuangZhou City, South of China, you know, most LED products make in China, so I have lots of options, PAR 38, PAR 30, PAR 20, fixed LED lights, etc! It seems not easy for you to buy from here.

PAR 38 is the standard incandescent type, 4.75 inches(120 mm) in diameter. I read some papers, and need some special wavelength, blue(450 nm) and red(650 nm), so I choose this type. Though some LED Grow Lights with CE/GS/UL/ETL/DLC/ES certification, they do not meet plants’ need in my opinion. I do not tested the precise spectrum.(As you know it need money to test!!).

The data 575/ 82/48 you refer to is illuminance Meter? I think there is some good products, but need some time to check.

Here is the LED Grow Light product I bought: