Light Panels specifications for PFC


#1

Hey everyone,

I have some questions about the light panels for the PFC:

  1. Are we at the point where we would care for spectrum tune-ability?
  2. What are the PAR levels that we want to acheive?
  3. Do we have any thoughts on the following lighting approaches:
    -light bars (individual linkable fixtures)
    -light panels (large PCBs)
    -built up light panels (several small PCBs)
    -light boxes (panel shaped fixture with LED board on edge(s)
    -light strips
  4. What is the state of the OpenAg-developed light panel?

Thanks!

CC @Webb.Peter @spartan081 @TechBrainstorm @Poitrast


Complete design of the personal food computer (and possible alternatives of some components)
Sensor Data Modeling
Requirements for controlling the LED Panel
#2

Adrian I would expand on the tune-ability question and further ask whether there is interest within the community to utilize specific wavelengths such as far red (730nm) and some of the new phosphor converted violet LEDs which have been developed by Lumileds and others specifically for horticulture. I would suggest that FC should aspire to follow the state of the art to enable research in this field.


#3

Couldn’t agree more. UV/far red LED has huge potential for manipulating phenotypical traits.

I’m interested to know your experience with these bulbs. How difficult would it be to build one in addition to an existing light source? Please link me to some good builds you’ve seen or done yourself.


#4

Hi Peter,

I develop the technology behind GrowFlux. I’ve collaborated with leading LED manufacturers developing phosphor converted LEDs for horticulture, have designed and prototyped horticultural COB devices, remote phosphor LED engines for horticulture, and package on board LED light engines with integrated drivers for horticulture. I am currently putting a 16 channel 600W horticultural lighting board with integrated drivers in production and have 7 patents and a few pending in LED lighting.

We are working on something that I believe might be a good match for the Food Computer if we can scale it down and cut cost a bit. Any thoughts on Adrian’s question about preferred form factor would be appreciated.

I’d also like to get a handle on cost - does anyone have cost targets in mind for a few scenarios such as those listed below?

  • Fixed spectrum lighting reaching 150µmol/m2/s
  • Fixed spectrum lighting reaching 300µmol/m2/s
  • Tune-able lighting reaching 150µmol/m2/s
  • Tune-able lighting reaching 300µmol/m2/s
  • Tune-able lighting incorporating UVA LEDs

(lets define tune-able as including, at minimum, white, far red, deep red, and blue LEDs)

Finally, is lighting exceeding 300µmol/m2/s of interest for the Food Computer?


#5

HI ericelsele,

I’m working with a group of volunteers in St. Louis to build two Food Computers based upon the Fenome FC kit. The lights as supplied in the kits appear to be under powered (i.e. low PAR readings) so we are searching for new lights that we can put into these computers. Do you have any samples or suggestions that you can donate to our effort?

Thanks

Joe


#6

Adrian, keep in mind these are 100% my personal opinions. I won’t go into detail about why, but if you’re curious would be happy to explain my reasoning.

  1. It depends, who “we” is. Undoubtedly this will increase price/complexity. Another option I plan to use for the MVP is just to provide two sets of full spectrum COB LEDs (2800k, 5000k). For years horticulture did this between Metal Halide & HPS.
  2. 200-300 umol.m.2s1

-light bars (individual linkable fixtures) - More practical for large-scale applications (generally longer)
-light panels (large PCBs) - Could you explain this more? Do you mean COBs or arrays of diodes?
-built up light panels (several small PCBs) - Same as above
-light boxes (panel shaped fixture with LED board on edge(s) - Probably not, please explain this as well, perhaps include a picture.
-light strips - These are pretty much what goes on a heat sink to make a “light bar” if I’m not mistaken.

@ericeisele

The only model I found on your site did not list a price, but judging by the quality sensor integration you’re developing top of the line products. Honestly, that product may be the most advanced I’ve seen to date, which while impressive means it also is likely way out of my price range. While I completely understand the value of tuneable spectrum lighting for research purposes and large-scale commercial operations I worry that at this point it’s just too expensive for the average “maker” to get their hands on.

I am a novice and the extent of my LED development is mounting COBs on CPU heatsinks. I went this route because there was no way I could afford $1000 on a KIND LED array or other comparable product. I’m curious about whether or not you think a fixed spectrum COB targeted at leafy greens is feasible. I was using 32V CREE CXB3590’s to grow peppers, but we wouldn’t need anything near that high powered for a PFC.

Fixed spectrum lighting reaching 150µmol/m2/s - When would this low level of lighting be desired? In my reading, I haven’t seen anything lower than 200 be recommended.
Fixed spectrum lighting reaching 300µmol/m2/s - I know this is low, but I genuinely think that a small fixture designed for a 2x2 space needs to be >$200 in order to compete going forward. I may be totally off base here, in which case, please prove me wrong (that’s why I love these forums). Keep in mind, the lights we use for the MVP generate 264µmol/m2/s at canopy level and are just 4 sawed off GE bright stik’s.
Tune-able lighting reaching 150µmol/m2/s - Same question as fixed 150, though this one might be interesting as a supplemental fixture.
Tune-able lighting reaching 300µmol/m2/s - The first iteration of the lights for the V2 is 3 channel blue/red/white LEDs. For those interested in variable spectrums to create light recipes this is IMHO the bare minimum light. I would think that people would be willing to pay $400-600.
Tune-able lighting incorporating UVA LEDs - It’s hard to put a price on these because so few people even understand the potential. That’s why I initially left the comment is because I am excited by this area of research and would love to open people (primarily students) minds to all of the possibilities of CEA. That being said, I think you’ll have a hard time getting people to go out on a limb until you can link it back to yield or specific breakthroughs. I think if you could provide a $150 add-on module that could be used in addition to an existing fixture you would be able to find some interested individuals.

I have to answer your final question regarding lighting exceeding 300µmol/m2/s with a no at this point. While I think perhaps on larger systems like the Food Servers or PFCs that are designed for flowering plants this would be desired, at this point in time that has yet to be developed. I think as far as the CEA/Vertical Farming community is concerned they’re mostly focused on leafy greens, anything that high of intensity generally gets targeted at Cannabis.

Thanks for your contributions, I look forward to hearing back from you and appreciate your expert advice.

One final comment: as we’re working on CV for the MVP we’ve already realized our lights are too bright and wash out some of our images. We are using full spectrum COBs, but I know this is an even bigger problem with red/blue LEDs especially if trying to look for deficiencies using cheap cameras. If I’m not mistaken the V1 had a very simple white COB light strip in addition to the regular lights that came on just for taking pictures. I also think for beginners, even getting dimming capabilites will offer enough opportunity for experimentation.


#8

Two more comments:

  1. @adrianlu I recommend you read this post from awhile back, it provides some great perspective.
    Minimalist Growing

  2. I saw this article mentioning @iancollmceachern on LinkedIn earlier today. Looks like he’s working on finding the V2/MVP a light as well: http://www.crystallinlighting.com/light-polymers-growblade-evaluated-by-open-agriculture/


#9

@ericeisele I would be interested to get your feedback and recommendations for us V2 users. I’m very interested in adding on UV control as well - can you please provide your input and some links?

If anyone has a solution for lighting they’re using for the V2 right now please share, I don’t care if its 100% DIY or off the shelf. @TechBrainstorm @adrianlu @sllo @ferguman

I know @Poitrast mentioned he’d be putting out some designs with regards to the new lighting PCB/panels, but I haven’t seen one of those completed yet and am pretty sure they won’t be available to the public for some time. My goal is to get everyone outside of Boston growing who have invested the time and money into building a V2 and now need the most critical component.


#10

I’m helping to build two 2 V2 FC’s for the St. Louis Open Ag group. I
don’t have a lighting solution yet. The V2 FC light area is about 18" X 21"
I believe. I was thinking of using the V1 FC light panel. Or maybe the
MVP light fixture. I have some of the GE lights that were originally
spec’d for the V2 but the set I have doesn’t appear to emit enough PAR to
make them usable.


#11

We have been using a collection of lights for the PFC, but none of them are particularly exciting right now. We have used an Aerogarden 45w light (taken from one oftheir devices), a 45W R/B LED, a Growblade and then even the 4-GE LED bulbs.

Right now, the GE LED bulbs is actually the strongest, followed by the Aerogarden light. Would love to have something a bit more flexible and full spectrum for the future


#12

We haven’t started working on a lighting solution yet but we have started talking about developing a solution. Eric and I were trading messages briefly but I think he got swamped with work. For now, we are probably going to use the lights from the v1.


#13

We have been working on a vertical farming product, but it won’t be ready until early 2018, and is really more suited toward 4’ x 4’ minimum size racks.

We do have a 6 channel 24 watt full spectrum demo board of sorts some of you might be interested in testing. Its actually an internal platform we use to stress test our mesh networking and cloud features. The board measures 7.5" x 4" and can be driven as high as ~50W with some heat sinking. Standard configuration is 450nm, 450nm, 660nm, 660nm, 730nm, and 4000K 80CRI channels, but we can populate the monochromatic channels with pretty much anything, including UVA packages.

PM me if anyone is interested and I can send along some info. Also I will be in Cambridge this Friday/Saturday if anyone is interested in meeting up.