i’ll not buy another led until..

[VincenzioVonHook]

i said 10-20% but whatever its clear you have problems so sure do your thing i really dont care much.

Tell me you care without telling me you care

 

[crimsonecho]

The radiators in a custom cooled kit still need active airflow. My PC has a custom loop and still has

Tell me you care without telling me you care

sure have fun man i guess you need to feel like someone out there gives a fuck about you, i give a fuck about you. not more not less.

 

[MICHI-CAN]

Kids just entering the world of thermal applications in biophysics. As I just grin.

 

[lusidghost]

i said 10-20% but whatever its clear you have problems so sure do your thing i really dont care much.



still there will be an exhaust fan running in a tent. so air movement is there. just not pointing a cheap fan to run it cool. what is so hard to understand about that?

A really powerful exhaust fan right beside the lights would probably cool them. Those HPS tube lights work, right?

 

[HippieDudeRon]

so to bring it around to the topic once again the votes are 5 yes to 10 no. so i guess there are many people who does give a fuck about the lifespan of their expensive fixtures and want better thermal management. and considering the question is itself is may be a little flawed as better thermal management can be achieved by many means, though heatsink upgrades were the easiest route to take in my mind, i’m guessing a lot of people are bothered by how hot their fixtures run.

just a couple of hours ago @hotrodharley was posting info about his fixture which rans at scalding temps even with fans pointed on if i remember it correctly.

migros ir camera readings show that scalding hot temps on the heatsinks as well.

120F is scalding hot. so besides shortening the lifespan of these diodes these temps can also cause bodily harm for some people. i thought we passed this phase of grow lights when a lot of people abandoned hids and moved to leds. if the heatsink is scalding hot, the diodes are almost definitely too hot.



its not hot to touch its dangerous to touch. its not about migros video its about me buying plug and play for the first time and realizing this shit was on the fan already. i dont have a problem with manufacturer data the problem is heatsink temps are hitting 50 so what are the temps on the actual diodes would you like to contribute by taking these measurements?



again there are many ways to achieve proper thermal management. this is not about bad air exchange. i got adequate exhaust in my tents. more than adequate to be exact. complete air exchance 2 times per minute or something. if i have to hang fans on top of my lights then whats the point. thats exactly what i meant by the way, these lights should run cool to the touch with just exhaust fan running. do yours?

50C is not hot for an LED. 85C is not that hot for an LED. That is what they are designed to runat. What do you not get about that?
They aren't selling it as so cool you can drop your balls on it...they sell it as so cool it will last it's LM80 or better. Which 50C is plenty cool enough to do that.

 

[VincenzioVonHook]

till there will be an exhaust fan running in a tent. so air movement is there. just not pointing a cheap fan to run it cool. what is so hard to understand about that?

Some of us factor in price to performance before any purchase. It's simple. A $10 fan will do a better job than a passive heatsink by itself , and the same heatsink will work better with the $10 fan on top of that. For us advocating active cooling, the core of the matter is lowering the functional temperature, whereas with you it's about some bizarre personal attachment to heatsinks instead of the actual functional temperature.

We are trying to achieve a relative drop in functional temperature here, and will choose the most cost and performance effective choice plain and simple. The medium doesn't matter. Just price to performance.

 

[loco41]

I'm still curious why you bought premade fixtures when you are comfortable with building your own? You said earlier that is most likely the best option so why not stick with that instead of buying these 4 separate units to then complain about their heat management abilities?

 

[VincenzioVonHook]

sure have fun man i guess you need to feel like someone out there gives a fuck about you, i give a fuck about you. not more not less.

Why thank you. It's good to feel acknowledged.

 

[crimsonecho]

Why thank you. It's good to feel acknowledged.

you’re welcome

I'm still curious why you bought premade fixtures when you are comfortable with building your own? You said earlier that is most likely the best option so why not stick with that instead of buying these 4 separate units to then complain about their heat management abilities?

didnt have the time or the energy to build another one.

50C is not hot for an LED. 85C is not that hot for an LED. That is what they are designed to runat. What do you not get about that?
They aren't selling it as so cool you can drop your balls on it...they sell it as so cool it will last it's LM80 or better. Which 50C is plenty cool enough to do that.

at the actual diode yes. but i havent taken a reading of the diodes have you? whats the lm80 for samsung diodes and whats the lm80 on a mars fixture for example. the data put forth by migro suggests the lm80 for most fixtures are not what its supposed to be.

 

[PadawanWarrior]

you’re welcome



didnt have the time or the energy to build another one.



at the actual diode yes. but i havent taken a reading of the diodes have you? whats the lm80 for samsung diodes and whats the lm80 on a mars fixture for example. the data put forth by migro suggests the lm80 for most fixtures are not what its supposed to be.

Just curious about your built lights. Is this what you're talking about? Are those standard LED bulbs with the diffusers removed?

 

[crimsonecho]

Just curious about your built lights. Is this what you're talking about? Are those standard LED bulbs with the diffusers removed?

View attachment 5139825

yeah pretty much.

well anyway they’re 10 cree cxb1830s 3000K 90cri vivid spectrum (oh and also have 6500Ks and 4000Ks they are not 90cri or vivid spectrum haha) cobs ran at 700ma. running in series. actively cooled with cheap pc fans. i still got a smaller version running in my 2x2 and heatsink stays at room temp. miracle of cooling i guess.

 

[crimsonecho]

Just curious about your built lights. Is this what you're talking about? Are those standard LED bulbs with the diffusers removed?

View attachment 5139825

also where did you find that photo from its looking gnarly haha i could have posted a new one if you asked.

 

[crimsonecho]

here is my 2x2 atm since you were curious. personal crosses vegged for a couple of weeks fast flip to flower.


here are some other grows under the bars you were curious about



i had others but have to do some digging and i’m tired.

 

[PadawanWarrior]

also where did you find that photo from its looking gnarly haha i could have posted a new one if you asked.

You're journal,

 

[crimsonecho]

You're journal,

haha i abandoned that journal sometime ago i dont even remember whats there
riu is not my journal forum i’m doing that stuff on the cp now

 

[MICHI-CAN]

You're journal,

Needs some IR and those enclosures trap heat. Drill holes in the top and sides. Yikes.

 

[cobshopgrow]

the ebs you quoted are nothing like what mars or any other bar type light manufacturers use. they cram those diodes so close to each other its running almost as dense as cobs.

i think 4 years is a decent amount of time not arguing that but everyone changes their light in 4 years is another speculation imo. maybe they wouldnt if they didn’t see significant par loss.

can you do the same math with a mars pcb for example? what 100cm long about 10cm wide? running at 100w? still a shit ton of heat. nothing like and underdriven eb strip.

in the original video migro runs the lights for 6K hours to reach at that conclusion. so its not a years based experiment. you can reach 6K hours in a year with veg and flower with photos, if growing autos faster. and the lights he used are mostly cobs with big chunky spikey heatsinks. but doesnt matter much, he sees 40-50C on those heatsinks and i definitely believe if i were to push my lumatek 100% i’d see similar temps and i’m pretty sure mars hydro runs even hotter because the diode density is incomparable to my lumatek etc.

4 years to q80 may be an overestimation of what it’ll produce over the years for most lights.

somehow somewhere along the way companies started pushing these light without proper heatsinks and thermal management because people thought hey these are not dense as cobs so they dont need much cooling i guess but your underdriven eb strips does not have much in common with what these companies produce

the numbers i was refernig to for the EB3 strips where the nominal, same for the 1818 Cob.
am too lazy too calc what Mars and else drives their leds with.
there is probably room for improvement in future generations.
while what they all do will be complete within specs.

the math for your example would be easy.
100x10 = 1000
100W divided by 1000 = 0.1W a sqaure centimeter.

no matter what you cant reach the density of a cob with mid power leds.
dont forget there is a casing around the led itself.
3030 is quite common these days, which are 3x3mm, the smd in below example will be smaller then 3030s.

the thing for me is mainly that using more leds is even cheaper then using more heatsink material and the benefit is much higher for the given cost.

other thing, i used COBs on huge pinhead heatsinks, theyre a pain to handle and hang, i really dont like their weight.

 

[crimsonecho]

the numbers i was refernig to for the EB3 strips where the nominal, same for the 1818 Cob.
am too lazy too calc what Mars and else drives their leds with.
there is probably room for improvement in future generations.
while what they all do will be complete within specs.

the math for your example would be easy.
100x10 = 1000
100W divided by 1000 = 0.1W a sqaure centimeter.

no matter what you cant reach the density of a cob with mid power leds.
dont forget there is a casing around the led itself.
3030 is quite common these days, which are 3x3mm, the smd in below example will be smaller then 3030s.

the thing for me is mainly that using more leds is even cheaper then using more heatsink material and the benefit is much higher for the given cost.

other thing, i used COBs on huge pinhead heatsinks, theyre a pain to handle and hang, i really dont like their weight.

i’m really not trying to dispute that smds are not dense as cobs. of course they are not. just saying on some of these boards smd replacement is super close and very dense and we went over this and i know for you its a given but without altering the pcb dimensions just cramming more diodes will not make much of a difference.

huge pin heatsinks are definitely less convenient than the ones being used on the bar types today, of course. but the airflow between the fins and the total surface area are so superior that it can cool a cob with that density, i wonder how well it’d work out for pcb board used in the bar designs. plus if i have to hang a fan over the fixture and lose headspace i’d really prefer a pin type heatsink protruding from the sides of the bars to catch the air i pull in from the bottom of tent with the exhaust fan.

i’ve done both pcbs and cobs as a diy and while a not actively cooled cob at the same wattage and heatsink burned out on me, pcb kept working but got scalding hot on the heatsink. so yeah less density helps with better heat transfer and dissipation and i use cpu thermal paste on these heatsinks after drilling holes for the led holders and leveling the surface with very fine sandpaper for cobs and pcbs so the heat transfer is pretty much at its best.

your eb3s probably do run cool to the touch on a relatively small heatsink especially when underdriven because of the led replacement and low current going through it but thats really not the story for many fixtures i’ve seen. really crammed in diodes. plus 50C on the heatsink as seen on migros video is medically scalding hot temperatures. i just expect better or at least not misinformation like it runs cool to the touch.. if i point a fan at it. that if i point a fan at it part is somehow a given for some where i find it totally idiotic when one of the selling points for most of these lights are “no moving parts”.

yeah no moving parts in the box but you gotta have moving parts to have them work correctly. brilliant.

 

[Horselover fat]

Mass of the heat sink doesn't dissipate heat. The surface area does. Why cram the diodes and spend money on big heat sinks when most of us don't want hot spots anyway?

i've show these boards many times, but these four boards run at 80w each. Each has 308 diodes and they have no additional heat sinks as the large board is enough to dissipate the heat. Those fixtures are very light and yet they are usually only luke warm. Oh, and that is 2*4 tent.

 

[crimsonecho]

Mass of the heat sink doesn't dissipate heat. The surface area does. Why cram the diodes and spend money on big heat sinks when most of us don't want hot spots anyway?

i've show these boards many times, but these four boards run at 80w each. Each has 308 diodes and they have no additional heat sinks as the large board is enough to dissipate the heat. Those fixtures are very light and yet they are usually only luke warm. Oh, and that is 2*4 tent.

View attachment 5139929

you got a fan pointing at it?

dont really care about the heatsink mass per se but as i said these heatsinks are basically aluminium slabs with no real fins and stuff so the surface area is abysmal compared to a pin type heatsink used on cobs. as i said they can be designed better and sure the pcb itself can be designed better not arguing over that either but if insisting on using these smd placements on these boards maybe use a better designed heatsink to catch the updraft created by exhaust fan pulling air from the bottom up. so maybe that way no fans are required on top of the fixture and it really can run without additional fans that way is my point.