New Light

[ChiefRunningPhist]

Hey all!

Making a new light and was originally trying to be secretive about it lol, but recently started to be more realistic about my expectations and figured I'd try to get some input from folks. Idk if it's an original idea or even a good one but hoping to make it better.

I'm building a hollow center PCB light with Wi-Fi control. Using the ESP32. I'm making tuneable/dimmable main, IR, UV, and far red channels. The light will be plug and play right out of the box with a HLG 160p 60b included in frame. The android app will control dimming, scheduling, Emerson effect timer, as well as the RH, T, and FLIR sensors (for VPD). It won't have an external server to store my info, I won't need to have internet access. It will be daisy-chainable. The sensors let me make "rules" in the app to activate/deactivate a seperate Wi-Fi outlet juction box with 4 outlets, - initially designed for AC, heater, exhaust fan, humidifier, but I could plug anything in. It a 2 part light, 1 part hollow center PCB light, 1 part seperate outlet junction box, but controlled via light/app.

Working on the PCB layout right now as my IR & UV chips have not yet been selected, which makes PCB design change. I'm using EAGLE.

Was going with Nichia 757 V3 5000k 80CRI (400 per light) and then fine tuning total light Blue% by increasing far red respectively. Obvi tuning further with IR and UV.



I've moved on to designing dimming ccts because as soon as I settle on the specific IR and UV (and far red chips, are 660nm epistars currently) I should be able to transfer the light PCB into a Gerber for board fabrication. Dimming cct & ESP32 will be on a different PCB.

I'm trying to not have to buy additional drivers, but only use 1.

If you're interested in becoming part of this project pm me. If you have suggestions on chips Id love to hear them. Or any input! Thanks.

Im barely in the prelim of designing the 3D printable frame in freecad, which supports the driver and attaches it to the hollow PCB, so if you are a 3D printer guy hmu! Its been the most neglected part of this project so far.

 

[JSheeze]

Ill pm you

 

[Rocket Soul]

Subbed

 

[OLD MOTHER SATIVA]

way to decked out for me..

i support your enthusiasm..

go for it..

 

[ChiefRunningPhist]

LM301b & Nichia 757 V3 EAGLE libraries..
https://drive.google.com/folderview?id=1HGq6clKrR63rqt_kdpLqJY7FoizSQ2PS

I'm pretty sure it's going to be 5 channel. I keep flip flopping on many things lol but IR is one of them.

75% sure-ish lol...
Chan 1 - 160w (50:50 - 5k Optisolis / 3k 80CRI 757)
Chan 2 - 20w 660nm
Chan 3 - 10w 730nm
Chan 4 - 20w 525 nm
Chan 5 - 12w UV (4w 290nm, 4w 365nm, 4w 380nm)

Board dimensions are 19" × 19" w/ 8.5" × 8.5" square cut out of middle. ~288in^2.



I'm having to upgrade to a 240h or buy additional drivers to power the supp channels. :/ Though, I want it to perform over 1100+ evenly, should I want it too.

I've found MCPCB large enough to print the fixture in one piece, but the company I was originally speaking with has a 17" size limitation, so I'll be printing it myself.

Using this, or some variation of this method...

I don't think I can soak my entire board in etchant because the aluminium will also be etched, so I think I'll have to use a spray bottle and be more selective in my etchant application. Ill be using H202 + HCl. (There is a thin insulating layer between the aluminium and copper)

I'm having a local print shop print my "PCB glossy paper" transfer template. Ill be hoping they can cut me a vinyl solder stencil as well.

Fixture will have an averaged max thermal dissipation rate of .76w/in2 (220w/289in2), which I'm fairly confident can be thermally managed using only the aluminum backing from the MCPCB. For comparison, an HLG QB120 is ~104in2 and operates at 75w max, or .72w/in2.

FLIR is soooo cool but also an immediate $40 addition. Upgrading the driver adds cost as well. The MCPCB comes at a premium too, once you start to add up all these "cheap" things it really starts to snoball. It would be much easier to buy premade stuff, so this build is not prioritizing cost (well as much as I can), and I'm admitting a reality that I can achieve the same for cheaper. This is just a fun project for me. Ill add links to all libraries I make (EAGLE).

Why green light? Refer to attachment.

 

[Unorthodoxy]

I'm very intrigued. Looking forward to this build.

 

[Covetsculitvars]

OH goody!

 

[SSGrower]

LM301b & Nichia 757 V3 EAGLE libraries..
https://drive.google.com/folderview?id=1HGq6clKrR63rqt_kdpLqJY7FoizSQ2PS

I'm pretty sure it's going to be 5 channel. I keep flip flopping on many things lol but IR is one of them.

75% sure-ish lol...
Chan 1 - 160w (50:50 - 5k Optisolis / 3k 80CRI 757)
Chan 2 - 20w 660nm
Chan 3 - 10w 730nm
Chan 4 - 20w 525 nm
Chan 5 - 12w UV (4w 290nm, 4w 365nm, 4w 380nm)

Board dimensions are 19" × 19" w/ 8.5" × 8.5" square cut out of middle. ~288in^2.

View attachment 4305407

I'm having to upgrade to a 240h or buy additional drivers to power the supp channels. :/ Though, I want it to perform over 1100+ evenly, should I want it too.

I've found MCPCB large enough to print the fixture in one piece, but the company I was originally speaking with has a 17" size limitation, so I'll be printing it myself.

Using this, or some variation of this method...

I don't think I can soak my entire board in etchant because the aluminium will also be etched, so I think I'll have to use a spray bottle and be more selective in my etchant application. Ill be using H202 + HCl. (There is a thin insulating layer between the aluminium and copper)

I'm having a local print shop print my "PCB glossy paper" transfer template. Ill be hoping they can cut me a vinyl solder stencil as well.

Fixture will have an averaged max thermal dissipation rate of .76w/in2 (220w/289in2), which I'm fairly confident can be thermally managed using only the aluminum backing from the MCPCB. For comparison, an HLG QB120 is ~104in2 and operates at 75w max, or .72w/in2.

FLIR is soooo cool but also an immediate $40 addition. Upgrading the driver adds cost as well. The MCPCB comes at a premium too, once you start to add up all these "cheap" things it really starts to snoball. It would be much easier to buy premade stuff, so this build is not prioritizing cost (well as much as I can), and I'm admitting a reality that I can achieve the same for cheaper. This is just a fun project for me. Ill add links to all libraries I make (EAGLE).

Why green light? Refer to attachment.

So it would be your contention that even higher coler temp (4000K, 5000K) still dont contain a proper ratio of green even in high cri?
Given all led stats for individual chips seem to be given in relative numbers Im curious how you came to this conclusion if that is the case.
Not a shit post, curious.

 

[ChiefRunningPhist]

So it would be your contention that even higher coler temp (4000K, 5000K) still dont contain a proper ratio of green even in high cri?
Given all led stats for individual chips seem to be given in relative numbers Im curious how you came to this conclusion if that is the case.
Not a shit post, curious.

Haha! No offense taken! I think I know what you're saying, why not just use straight opti's? Achieving photon saturation or fluorescence with red & blue and then adding green is the same as having it to begin with, right?

I think it might help with "penetration."

When I used HID I felt I got better growth under canopy than with LED.



I wanted a red weighted SPD, but didn't want to neglect green. I liked the width of the SPD, and the added blue peaking around 450nm of the 3k, compared to just supplementing the opti with a narrower banded red. I liked that the opti peaked in 420nm and had WV farther red than the 3k. The overall width of the opti was a major part of its selection. It allowed me to cut out the 420nm channel.

Hopefully this is what you were asking about

 

[SSGrower]

Haha! No offense taken! I think I know what you're saying, why not just use straight opti's? Achieving photon saturation or fluorescence with red & blue and then adding green is the same as having it to begin with, right?

I think it might help with "penetration."

When I used HID I felt I got better growth under canopy than with LED.

View attachment 4306027 View attachment 4306026

I wanted a red weighted SPD, but didn't want to neglect green. I liked the width of the SPD, and the added blue peaking around 450nm of the 3k, compared to just supplementing the opti with a narrower banded red. I liked that the opti peaked in 420nm and had WV farther red than the 3k. The overall width of the opti was a major part of its selection. It allowed me to cut out the 420nm channel.

Hopefully this is what you were asking about

Thats exactly where I was going with it.

Is this for a 3x3 area? Didnt see if you posted that info.

UV is a bit further to uvc than I prefer, but wattage is low and are on separate channels, so assuming you can control them.

Hope you are having fun doing this.

 

[ChiefRunningPhist]

Thats exactly where I was going with it.

Is this for a 3x3 area? Didnt see if you posted that info.

UV is a bit further to uvc than I prefer, but wattage is low and are on separate channels, so assuming you can control them.

Hope you are having fun doing this.

You know it! Its ideally meant for a 2x2 but with the extra power could extend it a bit further (2.5) or raise it up a bit higher. I wanted to have the power but don't see me slamming it at 220w.

And ya, dimmable, Ive been thinking of moving towards the 310 nm peak. The supposed SPD of the 290nm I was looking at is decently wide and I'm not trying to get under 280nm. I don't want to cook them!

 

[ChiefRunningPhist]

Have prelimiarily finished dimmer cct. Prelim Wi-Fi connection code has been flashed to the ESP32.

ESP32 and dimmer cct are now comunicating and commands are coming over Wi-Fi.

Next steps, build light, tweak code, and test dimmer integration.

Most supplemental chips ordered.

 

[ChiefRunningPhist]

Waiting on parts... Working on app in the meantime... It needs some color and style lol but its coming along. Ill put the FLIR video where the blank spot is on the main menu so I can scroll from light to light using the buttons underneath it..

 

[SSGrower]

Digital Bagel? Is that a plug and play interface for the esp32?
Or is it what you ar building?

 

[ChiefRunningPhist]

Digital Bagel? Is that a plug and play interface for the esp32?
Or is it what you ar building?

Lol you noticed that? Haha Idk if I'll stick with that name or not. But no, I'm building it out.

 

[SSGrower]

Googled as a digital tape measure.
https://www.kickstarter.com/projects/bagel-labs/bagel-the-worlds-smartest-tape-measure

This is more my speed

Digital Beagle

 

[ChiefRunningPhist]

Googled as a digital tape measure.
https://www.kickstarter.com/projects/bagel-labs/bagel-the-worlds-smartest-tape-measure

This is more my speed
View attachment 4310153
Digital Beagle

Haha that's crazy! Ya I thought about digital donut too but digital bagel sounded techier lol I'll have to come up with something that isn't taken

That's your graphic design?

 

[SSGrower]

Haha that's crazy! Ya I thought about digital donut too but digital bagel sounded techier lol I'll have to come up with something that isn't taken

That's your graphic design?

Nah, google image. I put effort in when necessary.

Just a comment on the use of monos and penetration/distribution as it relates, more of an observation with no technical backing. I notice spots on the floor of my garden from distributed monos (amare se350), this is not apparent with the monocobs (cobs of red mono leds) there is a red shade but no spots like with individual monos. I am of the opinion an intense difuse light is ideal and if the light source is too much like a point source, it seems to have a damaging affect. I feel like clusters of same color monos might mitigate this, so with all that said, how many of each mono and what wattage range per?

 

[ChiefRunningPhist]

Nah, google image. I put effort in when necessary.

Just a comment on the use of monos and penetration/distribution as it relates, more of an observation with no technical backing. I notice spots on the floor of my garden from distributed monos (amare se350), this is not apparent with the monocobs (cobs of red mono leds) there is a red shade but no spots like with individual monos. I am of the opinion an intense difuse light is ideal and if the light source is too much like a point source, it seems to have a damaging affect. I feel like clusters of same color monos might mitigate this, so with all that said, how many of each mono and what wattage range per?

I'm not sure yet on wattage requirements. I'm still trying to figure how many channels and what WV per channel. I may split every WV into its own channel. Right now I have 9 supp WV and only 5 supp channels so some WV are combined per channel. I'm distributing every channels' chips with even spacial distribution. (No clusters, except the UV will be split to 4 corner areas) The supps are 1w chips, compared to the .5w main light chips. That may need to be adjusted down to .5w chips too, if what you're saying ends up being an issue.

Ill post screenshots of the dimming menu when I get it ironed out a bit more, I deleted the previous pics because it'll have changes made and I don't want to confuse and make people think it's finished or that the wattages I have listed is what they'll be in the end.

Thinking..
6w of IR 850nm
8w of 730nm
8w of 680nm
16w of 620nm 660nm
16w of 525nm
4w of 310nm
8w of 365nm 390nm

150w of main light, 66w of supplementals..

I ballparked that it would take 7 white LEDs to match the single WV emission from 1 narrow banded LED or single color LED.

So looking at the 16w of green for example, I went 150w ÷ 2 because 3000k is lacking almost all green, that's 75w ÷ 7 colors ~11w of green is lacking, so I added 5 more watts to 11w to get a prominent peak over that SPD region if I so desired. I'm using the ×7, or ÷7 idea to get me in a region that I can then further control with dimmers.

I'd like more wattage per each channel but it starts to add up fast and then when you're cramming 90w of supp onto 150w of main for 240w total, and it's for a 2× 2, it makes you scratch your head, "how many watts do I really need?"

This is what I've put together as far as channels and WV goes, but even now I'm making changes to it..

 

[SSGrower]

I'm not sure yet on wattage requirements. I'm still trying to figure how many channels and what WV per channel. I may split every WV into its own channel. Right now I have 9 supp WV and only 5 supp channels so some WV are combined per channel. I'm distributing every channels' chips with even spacial distribution. (No clusters, except the UV will be split to 4 corner areas) The supps are 1w chips, compared to the .5w main light chips. That may need to be adjusted down to .5w chips too, if what you're saying ends up being an issue.

Ill post screenshots of the dimming menu when I get it ironed out a bit more, I deleted the previous pics because it'll have changes made and I don't want to confuse and make people think it's finished or that the wattages I have listed is what they'll be in the end.

Thinking..
6w of IR 850nm
8w of 730nm
8w of 680nm
16w of 620nm 660nm
16w of 525nm
4w of 310nm
8w of 365nm 390nm

150w of main light, 66w of supplementals..

I ballparked it would take 7 white LEDs to match the single WV emission from 1 narrow banded LED or single color LED.

So looking at the 16w of green for example, I went 150w ÷ 2 because 3000k is lacking almost all green, that's 75w ÷ 7 colors ~11w of green is lacking, so I added 5 more watts to 11w to get a prominent peak over that SPD region if I so desired. I'm using the ×7, or ÷7 idea to get me in a region that I can then further control with dimmers.

I'd like more wattage per each channel but it starts to add up fast and then when you're cramming 90w of supp onto 150w of main for 240w total, and it's for a 2× 2, it makes you scratch your head, "how many watts do I really need?"

^^^^this last bit shows how much fun youre having.
Group uv, green, red (620-660), fr, ir.
Maybe group the greens with the white main? Then you could isolate a day night initiator with the extra channel, or split that 310 to its own. Whats your timing plan with uv btw?