ttystikk
Well-Known Member
Can you copy n paste the chart on the thread, so we can reference it?
COB Talk
- Thread starter GrumpyToker
- Start date
I see what your getting at, but your nitpicking. I designed the first iteration of the unit as passive (with hope) and found that it was running too hot (>50C). so I redesigned for Version 2 and have shown that it runs passively ok mid 40s.
BUT if I add a few watts of fan it runs 10 degrees C cooler. Your right we are not that far apart, I really was designing for passive. But that 10 degrees seems important to me and it is cheaper to add a bit of active cooling, to get it.
The heat sink thickness is only 0.098", so we have the thermal paste resistance plus the heat sink resistance. making my test point really not far from the Tj, since all measurements were made consistently at the same point, the point is moot.
with extruded heaatsinks from heatsinkusa the cost to lower the temps to the same level is MUCH higher. With the better anodized black pin fin heatsinks the cost is much closer. BUT have you seen the difference in cooling power between active and passive for a state of the art pin fin heatsink ? I recently tested a small 60 mm square sparsely populated pin fin with and without fans and got a 3X increase in performance with a minimal active air flow. pleasantly surprised me at just how well the pins perform.
Last edited:
uhhhhh, that would be the data sheets for your specific cobs ...
guod
Well-Known Member
a change of 10°C results in 2% flux variation
Vero 29 25°C(100%) to 35°C(98%) - this change is near linear up to 90°C
ttystikk
Well-Known Member
1% efficiency improvement for each 5C temperature reduction at Tc. I'd imagine the ratio would be similar for similarly designed chips?
guod
Well-Known Member
Its the material or wavelength here Gan or 450nm.
JorgeGonzales
Well-Known Member
Speaking of cheap and efficient, I started a thread about the CLU048 here, but here's a sweet snippet from the 1212 version's datasheet:
Full stats here, and a Citizen spreadsheet/simulator right here.
Running at 14.1W like @PurpleBuz, the CLU048-1212 would see 184lm/w for the 4000k 80 CRI version at Tc=25C. I don't think 25C is realistic in a grow room, but thats how Bridgelux advertises it's lm/w.
Anybody have a ballpark on the LER of Citizen COBs? I suppose these could still be less than impressive for eff %. That's where my head starts to hurt a little.
Full stats here, and a Citizen spreadsheet/simulator right here.
Running at 14.1W like @PurpleBuz, the CLU048-1212 would see 184lm/w for the 4000k 80 CRI version at Tc=25C. I don't think 25C is realistic in a grow room, but thats how Bridgelux advertises it's lm/w.
Anybody have a ballpark on the LER of Citizen COBs? I suppose these could still be less than impressive for eff %. That's where my head starts to hurt a little.
Rahz
Well-Known Member
Just depends on how you're using it and how much of it you're using. It's possible to keep temps at the test point 10C and less over ambient using fans. The emitter closest to the exhaust on my T2-2100 runs about 7C over ambient. That's just under 4 watts of fan power split between two emitters, 2 additional watts per emitter. If we split the cost of the fan, heat sink and fan driver ($14, $31.5 and $5) the cooling cost per emitter is $25.25
Airwalker16
Well-Known Member
How is it $25.
JorgeGonzales
Well-Known Member
$50/2.
Airwalker16
Well-Known Member
I'm an idiot
I see now.
I see now.
Last edited:
GrumpyToker
Well-Known Member
What do you fellas think about this strategy.
Use the hlg-185h-c1400b running 4 COBs for now until next round of upgrades then get another hlg-185h-c1400b running 4 more COBs in the same space but now I can dim both down to 700ma and have the extra efficiency. 67%
Seems like using the hlg-185h-c1400b can give you the higher output when you need it or you could double them up and run them both low at 700ma at get the efficiency when you want it.
hlg-185h-c1400b Is pretty versital drivers then get the best of both worlds with one driver.
Use the hlg-185h-c1400b running 4 COBs for now until next round of upgrades then get another hlg-185h-c1400b running 4 more COBs in the same space but now I can dim both down to 700ma and have the extra efficiency. 67%
Seems like using the hlg-185h-c1400b can give you the higher output when you need it or you could double them up and run them both low at 700ma at get the efficiency when you want it.
hlg-185h-c1400b Is pretty versital drivers then get the best of both worlds with one driver.
Airwalker16
Well-Known Member
Damn that Sink is perfect for a peltier cooler.Gonna play with this a bit.
Sent from my iPhone using Tapatalk
kmog33
Well-Known Member
I'm just wondering what kind of load I can stick on it passively.
Airwalker16
Well-Known Member
Well, it isn't black anodized so I'd say probably no more than 50ws.
ttystikk
Well-Known Member
The only way you get that efficiency number is if you keep your Tj at 25C.
ttystikk
Well-Known Member
Last pearl, then I give up; WTF is so great about passive if it lets the chip run hot?
kmog33
Well-Known Member
I run most of my stuff actively. But if I can run passively with the airflow in my space that's constantly running, its seems like less less parts is less work and leaves less room for equipment failure. It's was more likely a fan fail than an LED that's effectively cooled passively.
In a sense it will still be actively cooled. It will just be cooled using my whole setups cooling rather than a 90mm fan.
robincnn
Well-Known Member
simple and reliable. No mechanical parts to fail.
Chip hot/cold all subjective. 45C cool enough compared to 85C...but yet hot compared to 25C