JorgeGonzales
Well-Known Member
He is saying the Tj might be slightly hotter for the Citizen at the same Tc. Cree seems to be the worst of the bunch, but they don't publish specs.
...All Things Vero...
- Thread starter AquariusPanta
- Start date
guod
Well-Known Member
i edit my post to make it clearer....
kmog33
Well-Known Member
That's bridgelux at 1400ma vs the 1080 citizens test with. That's not really comparable. The charts I posted have the lpw at 1050 which is a much better comparison against the 1080ma current given for the citizens.
We're not comparing lpw at test current. We're comparing cobs at the same current output. The v22s test at 1400ma not 1080. Fortunately bridgelux publishes lpw at several currents. The 1050 being closest in comparison. The 5000k puts out more lpw in the last version. The 3000 and 4000k are all within 10lpw and the citizens have a higher blue % in the 3000k and 4000k than bridgelux, which would boost lumens. But hurt spectrum for flowering.
Sent from my iPhone using Tapatalk
Last edited:
Evil-Mobo
Well-Known Member
Thanks for clearing that up.
Evil-Mobo
Well-Known Member
Gotcha so it will work better with a pinned heat sink because it runs a bit warmer correct?
JorgeGonzales
Well-Known Member
I posted that for @guod, since he mentioned thermal resistance. Comparing Tc is the fair way to take that out of the equation.
Which is why you need to go back and focus on Tc=25C on both data sheets
kmog33
Well-Known Member
I'm focusing on lpw at current. At 1080ma on 36 volt cobs. The 5000k bridgelux put out more lpw. The 4000k put out 5 less lpw on a version you can't get anymore and the 3000ks put out 10lpw less on the same version. Do you not think the increase in efficiency in the new version bridgelux will make up the 5-10 lpw on the 3k and 4Ks, and increase the gap in the 5k? Bridgeluxs 4000k spectrum has more red than the citizens so the actual spectral output is better on the bridgelux than the citizens (for flowering cannabis) at a 5lpw loss. I would take 5lpw loss at a better spectrum. Adding blue to boost lumen output is nice on paper but not in practice when red is more beneficial than blue for our uses. The new bridgelux should pull even further ahead so this argument will be pointless lol. I wouldn't doubt that the new 54ish volt veros will probably give the 1818s a run for their money with lpw and a better spectrum.I posted that for @guod, since he mentioned thermal resistance. Comparing Tc is the fair way to take that out of the equation.
Which is why you need to go back and focus on Tc=25C on both data sheets
Sent from my iPhone using Tapatalk
JorgeGonzales
Well-Known Member
You know, it works as a system. Thermal resistance is measured in C/W in this case, which is how many degrees the temperature goes up per watt of heat dissipated.
The system looks like this:
led-> case -> thermal interface material -> heatsink-> ambient air
So you add up the thermal resistance between all those bits, and the lower the number, the cooler the cob stays. Bridgelux has a very low led -> case thermal resistance, but it's only a fraction of the whole.
Evil-Mobo
Well-Known Member
Ok so then I am confused again wouldn't the vero run cooler if both are cooled the same according to what you just explained?
JorgeGonzales
Well-Known Member
Do you have digitized SPD data for Bridgelux? Because you can't eyeball either relative or absolute amounts of color with any degree of accuracy. Just trust me here. You may be right, or not, but it's a minor side issue.
OK, I will use your first example. At 5000K, Tc=Tj=25C (which is a borderline dishonest way to rate lm/w on an LED) the V22 makes 5730 lumens, correct?
The Bridgelux makes 6382 lumens, or 170 lm/W. Got it? Good.
guod
Well-Known Member
at 50 Watt the vero29 has a temp-lift of 0.08x50= 4°C
the 1212 at the same watts has temp-lift of 0.34x50=17°C
the Cree cobs stay in the same range as the vero and nichia is like citizen on
thermal resistance
they maybe cheap but you have to pay for better heatsink to get the same junction temp
and by the Way we have only a handful manufactors of phosphors for leds. and they are all cooking the same water at a CRI of below 90.
the 1212 at the same watts has temp-lift of 0.34x50=17°C
the Cree cobs stay in the same range as the vero and nichia is like citizen on
thermal resistance
they maybe cheap but you have to pay for better heatsink to get the same junction temp
and by the Way we have only a handful manufactors of phosphors for leds. and they are all cooking the same water at a CRI of below 90.
kmog33
Well-Known Member
1, were you only talking abound bridgelux in that post because the v22 and bridgelux are one in the same.
2. Still not a valid point. If stick a 5000k 1212 on a driver that pushed on 1050ma it will put out 155lpw according to their data. If you put a v22 on that same driver at 1050ma it would be putting out 157lpw yeah? So exact same power consumed bridgelux comes out on top with the 5000ks. 5 under with more red in the 4000k and 10 under in the 3000ks with more red in the spectrum.
Have you experimented using both? I have and do run both so at least in practice I know which works better, but I also have the data. Again bridgeluxs new versions come out in the next couple weeks so will see if you still even have a dog in this fight at that point.
I'm not saying citizens are bad, just that bridgeluxs spectrum is a bit better for our needs(flowering) and i woud be very surprised if the efficiency across the board doesn't do a good job of overtaking the citizens with the v22s and possibly even on the vero 18s. The vero 29s should kill the 1212s and we'll see how the my do against the higher voltage 1818s and bridgelux is also doing a 54v version on the 29s.
Sent from my iPhone using Tapatalk
JorgeGonzales
Well-Known Member
It would in theory, yes. But only by .17-.08 C/W in the case of an 1818 vs Vero 29. So .07 C/W, so maybe 2 degrees running 50W at 50% efficiency. As much as .25 or so C/W with the 1212 vs Vero 29. But more efficiency can make up for greater tyermal resistance.
It can matter, but we cool for our cobs, not somebody else's, if that makes aense. The Cree PCT paints a pretty bleak picture for thermal resistance, from memory. Almost enough to matter, but they are very efficient so it's a bigger percentage if a smaller percentage of heat watts.
Anyway.
Evil-Mobo
Well-Known Member
So all this back and forth over minutiae at the end of the day
JorgeGonzales
Well-Known Member
It's worth it if you learned something.So all this back and forth over minutiae at the end of the day
JorgeGonzales
Well-Known Member
We are talking about the V22 here, and even at Tc=85C, which is as fair to Bridgelux as you can be with it's better thermal resistance, the 1212 poops on it.
So better temp droop? Don't know, don't care. You get it, I get it. Nobody else reading this has a fucking clue what we are talking about.
REALSTYLES
Well-Known Member
I can't help it but 155lpw isn't that great in my book that's probably why I haven't used my Vero 29 4k yet.
do you know what makes the cree and vero to have a lower thermal resistance than citizen/nichia?
JorgeGonzales
Well-Known Member
There is no dog fight. I don't have allegiance to corporations. They aren't sports teams.
I can't say I'm excited about the new Veros any more times or ways than I have. At the end of the day, the data sheets you pasted shows 170lm/W at 37W vs whatever the fuck lower number the V22 is and always will be. You don't think that's valid, then you can't read the data sheets. I legitimately give up.
I've done better than experimented. I've had a well known vendor who uses Cree and Bridgelux and has samples of Citizen tell me he's happy with his testing. He builds more lights than us, and confirms they are pretty OK.
The sad fact is testing doesn't matter, because you can't even get to step one, understanding the chart in front of you.
JorgeGonzales
Well-Known Member
Also, I think my blood sugar is getting low. I'm getting cranky. Apologies. But you are 100% not understanding what you are looking at.