What parts of Italy?...................I haven't been back to the motherland in awhile, can't afford it, happy the euro rate was good for you though.
........I'm trying to do this on my crappy android phone so bear with me. Edited some shit out that we're not privy to
I think we've seen enough garage testing of efficiency curves with Apogee meters, cross-referenced with data sheets and professional testing at this point to be fairly certain that halving the wattage and moving to 80CRI will give you more than a 6-7% gain in efficiency. It's just how the curve curves. Crush is subjective, but LEDs are still progressing nicely
Thank you for sharing. That clears up so much BS around here.Alright finally got permission to post the sphere results...............Big thanks to Nick for letting us use his UL testing facility in OH again and Dan for sending in the gear, I sure as hell wouldn't do it
"We finally had time to get your fixture in the sphere and I was able to do all channels at the same time, as well as each individual channel. Everything labeled Timber V2 below is from this round of testing. These actually performed a little worse than the stuff you sent me last year, but I'm guessing that's due to the high CRI. You sacrifice 10-15% of your efficiency by going to high CRI. It's worth playing around with to see if there is any noticeable difference between 80 and 90 CRI, but from everything I have ever played with efficiency is king.
The build looks great and I like the extrusions you used for the frame, we have a ton of that stuff around here that we use for shelving. I think you could gain some efficiency by running these at 75 watts but your parts list will increase and possibly be too expensive.
As far as UL goes, they want to see all wiring terminal connections inside an enclosure, all wiring pass throughs need to go through a grommet and you have to fuse anything over 100 KVA (approximately 100 watts). If you only sell to home growers you don't need to worry about it, but electricians cannot touch anything that does not have a UL or ETL label for the entire fixture.
Let me know how the testing goes and if you see any appreciable difference between the high CRI and standard.
Power (W) Flux (lm) Radiant Flux (W) Efficacy (lm/W) Radiant efficacy (W/W) Waste Heat (W) PPF (umol) PPF/J
Timber 660nm 4 94.78 1257 25.9 13.26 0.273 68.88 143 1.51
Timber 3000K 4 212.7 27824 86.19 130.81 0.405 126.51 402 1.89
Timber 3000K 4 Ref 212.8 25463 79.11 119.66 0.372 133.69 369 1.73
Timber 4000K 214.4 29084 89.8 135.65 0.419 124.6 410 1.91
Timber V2 658.2 67706 235.44 102.87 0.358 422.76 1082 1.64
Timber V2 1750K 208.9 14603 69.42 69.90 0.332 139.48 315 1.51
Timber V2 3000K 216.9 26243 81.97 120.99 0.378 134.93 381 1.76
Timber V2 3500K 214.7 27192 84.42 126.65 0.393 130.28 388 1.81
"
First three fixtures were the cxb3590 from last year, top bin, mw driver, 1.4 amps 4cobs, 50w a cob, mechatronix 100mm pin sink. 3000k Ref==with reflector, rest where bare. Also the osram oslon 660nm "puck " , top group mid bin I believe, Dan can correct me. All cobs 80ra, cri.
All V2 is the vero 29c gen 7 fixture just sent in, three circuits, two cobs per circuit, 1.4 amps, 100w per cob. All bare no reflectors, 3000/3500k 90ra,cri. Makes nicks life easier and just can test each circuit/ kelvin separately without moving fixtures in and out of the large sphere.
I first was puzzled why he thought the veros performed worse than the cxbs, YES high cri will knock it down, we know that, also isn't factoring in that their getting double the wattage, he just looks at the total draw= which is right (doesn't care what theoretically can achieve at lower current, just what you sent him to test) . Asked him if maybe the hlg splayed sinks are garbage? Reminded him that the 1750k, 15- 18% of it's spd is above 700nm and isn't factored in the ppf. Did ambient temps affect the outcome, hot as shit in Ohio! Lol..... essentially asked to elaborate
He responded:
"We let it warm up in the warehouse before it goes in the sphere, ambient temps were in the mid 70s. I like looking at the radiant efficacy because it sums the entire output of the fixture from 380nm to 800nm and doesn't give any weight to any particular wavelength, unlike PPF. Radiant efficacy is the sum of all light energy captured in the sphere divided by the actual input wattage at the wall, and it is the true efficiency of the entire system. For comparison's sake the Mega does 309 watts of radiant flux and has a radiant efficacy of 0.46. This is what all companies should be advertising, not PPF/watt.
I don't think the splayed heat sinks are performing worse, I think you're just seeing how poorly high CRI performs from an efficiency standpoint. If you put those old CREE ones we did last year in that same rig I think you might pick up a point or two.
I'm pretty sure all 3 you sent this time were high CRI, though that 1750K really highlights the shortcomings of the CRI tile measurement system. Those things are obnoxious lol. If you are looking for higher efficacy you are going to have to spread the heat out and the best way to do that is with lots of LEDs on big huge boards. Just for fun you should try some liquid phase change coolers like they use for CPU coolers in big gaming rigs. They can't be any more expensive than those crazy pin things."
We get it, you really hate high cri! Ha. Hlg Splayed sinks don't suck . He also seems to think that the 1750k are 97+cri like I did? Strange, vero datasheet says otherwise. Nothing new here really with the test, can go high power its a watt to watt de hps replacement with lower lumen maintenance/ depreciation. Active cooling (don't like liquid) with some low watt cpu fans are becoming almost mandatory with these pin sinks driven hard imo. If chasing efficiency, electrical savings==more cobs, lower ma ,way more $$$. Also qbs are a good option no doubt. Can't recommend any other boards until tested/ verified. Sorry but China Factory can sell/ tell you anything, doesn't mean is true. Enough bull shit in this industry already.
Edit........ phone, sorry I'll fix the results/ table when I'm back home again. Getting 4+ inches of rain tomorrow, crazy.
Sure can. I think you already know how too. Also, wanted to say, I recognize your hard work @ trying to develop better products & share w/ people what you've learned in the process. Props. Not joking.Well we have been talking with a lab or two and just like anything else once Wall Street moves in the deceptive testing will be over. Goniophotometer is not perfect but apparently they are working on a near field light distribution system that will give us the ability to accurately simulate real world light distribution.
Can we have our high CRI and Eat our Efficiency cake too? Time will tell soon
What BS? Those numbers are right in line with basically everybody ever for those wattages. From @SupraSPL to @CobKits , @alesh ... pick your guy. Right around 1.8 PPF/W for relatively hard driven cobs with driver losses taken into account, CXB3590s coming in right where you'd expect as well. I think PLC tested theirs at 1.8 PPF/W.
i still have yet to do the big chip shootout, but multiple vero Cs tested right on the money with citi1825 gen5. i have both 1825 gen 6, 3618 gen 6, and luminus cxm32 gen 3 to test against them, the three latter ones are all the same 38mm size.
Apparently they will test over a pretty large area 6x8. If the fixture is symmetrical they will actual test 1 quadrant against that area and then duplicate it 4x for a total test area of 12x16. Apparently they have software that will take that data set and generate the ies file. The only difference is with this method if you test at say 18" the simulation will only be accurate if you position the fixtures in the software at the tested height. So you will want to have 3+ mounting heights tested Imho.
Yes 400- 700nm cut, "definition" of PAR......... also his sphere is conservative?, ul power supply gives sort of a worst case scenario, idk what that means...... his words not mine.hey @PSUAGRO. do you know if he used a sharp cut at 400 and 700 nm to define PAR?
some testers use 350-750 nm. obviously makes a huge difference with far red in there
i still have yet to do the big chip shootout, but multiple vero Cs tested right on the money with citi1825 gen5. i have both 1825 gen 6, 3618 gen 6, and luminus cxm32 gen 3 to test against them, the three latter ones are all the same 38mm size.
i think as good as the splayed heatsinks are, the veros would really benefit from active cooling at 100W. im thinking adding a <1W fan would return 2+% in efficacy. Im running an experimental setup in my tent, 4 cobs on a 133 with a 50mm fan, im running at 120 total watts and the heatsink is under 45C
@robincnn did a fan test on heatsinks a few months back with same fans and he had same conclusion- the fans are worth it. i think if you can get yourself in a situation where you are depending on the fans for an efficacy boost, but still have enough mass on your heatsink to dissipate the cob's heat with the fan off, its a win-win. if you design the latter situation to be 85-100 Tc with the fan off, you just have to notice the fan is off sometime in the next 50000 hrs or so to avoid cob damage
I didn't see that test, but I have an IR thermometer and even the slightest air movement is enough to knock 10C off a heatsink hitting 60-70C in still air. Real air movement will cause 2-3x that temperature drop. It's significant, and can be the difference between 70C and 40C at the heatsink.
...some testers use 350-750 nm.... obtained with a spectroradiodemeter... we can obtain data outside de 400-700 par zone from what i saw on youtube vids from testlab on monstergarden....hey @PSUAGRO. do you know if he used a sharp cut at 400 and 700 nm to define PAR?
some testers use 350-750 nm. obviously makes a huge difference with far red in there
i still have yet to do the big chip shootout, but multiple vero Cs tested right on the money with citi1825 gen5. i have both 1825 gen 6, 3618 gen 6, and luminus cxm32 gen 3 to test against them, the three latter ones are all the same 38mm size.
i think as good as the splayed heatsinks are, the veros would really benefit from active cooling at 100W. im thinking adding a <1W fan would return 2+% in efficacy. Im running an experimental setup in my tent, 4 cobs on a 133 with a 50mm fan, im running at 120 total watts and the heatsink is under 45C
@robincnn did a fan test on heatsinks a few months back with same fans and he had same conclusion- the fans are worth it. i think if you can get yourself in a situation where you are depending on the fans for an efficacy boost, but still have enough mass on your heatsink to dissipate the cob's heat with the fan off, its a win-win. if you design the latter situation to be 85-100 Tc with the fan off, you just have to notice the fan is off sometime in the next 50000 hrs or so to avoid cob damage
He still gives total radiant flux which is how efficient the light is. Yes it still doesn't answer how many umols you have between 300-800 but at least it gives us a good idea where the light stacks up. Radiant efficiency will go down regardless when going more red shifted/higher CRI as you are measuring mw in mw out and we know there are less mw in a red photon +. So we can still not be sure that higher CRI infact produces significantly less photons than the 80CRI counterparts. Perhaps @alesh will work his magic and give us the missing photon count above 700nm. He has the SPD and total radiant flux and it should be doable for the master.
Another Philly grower, nice. How about the Snyder family getting in the med/canna game?
...........remember I'm a lazy trimmer and didn't break down some of the large buds either; had to post something before cure. Round 2 is where I kind of have a better feel of the light and give it a proper go; also off season for me with more time on my hands. yeah, haven't seen @alesh since that egomaniac trolled up his thread...........hopefully we can get some chilLed #s! lol