Coming soon..Cree CMT drop in COBS for all you favourite packages
- Thread starter welight
- Start date
Randomblame
Well-Known Member
Hopefully the CMT series has better numbers,
CMA series seems to be still behind current Veros/Luminus genenrations.
CMA series seems to be still behind current Veros/Luminus genenrations.
Greengenes707
Well-Known Member
I would suggest using the 72v versions of both to compare as the current:wattage the same for all intents and purposes and don't need to look around much.
The whole point of cree move to aluminum line(s) was to accommodate high current operation. And you will see at about the 85w mark...the CXB and the CMA cross paths in efficacy/efficiency and the CMA takes the lead all the way out. Looking at the 72v makes this easier to see on the PCT.
Cree still swears by ceramic for low current/maximum potential. But also over values it as we know. Aluminum takes the cool factor out of low currents and just levels the bar across the currents and helps prices.
The CMA's aren't flip chips which I found very interesting, really expected it, but didn't happen. Possibly the CMT has them...just speculating. I just heard about the cmt's last week and am waiting on more data.
welight
Well-Known Member
Just fleshing out the ALU based product range, offers slightly higher current than Citi , CLU048-1212 is 2.76A, CMT2850 is 3.50A, CLU048-1812 54V is 2.76A CMT2890 54 is 2.80 , in all but 1818 at 4.14A whereas CMT2890 54 is 3.3A, big current is not a high priority for hort so will be interesting to test in the mid current, we have 2890 sample inbound so testing next week with any luck.CMA and CMT will come with the same premium colour offering, Fidelity and Speciality, limited samples available now with production starting early Q2 CY2018 and off course LM80 that you can can bank on, will advise when datasheet ready
Cheers
Mark
CobKits
Well-Known Member
i guess that woulda been a little easier to read. the cxb3070 would still be 36V data though
thats my experience with CXB they really dropped off at higher currents (fantastic at low currents obviously. I was never a fan of the alumina ceramic and wondered why that was a design decision. at the time i believe they were claiming it was superior.
why would they be better at dissipation at low currents but not higher currents? i feel like theres a material property im missing. if aluminum is better at high currents, wouldnt it naturally be an improvement across the whole range?
it seems like there are opposing forces at work in the
-conduction from dies to substrate
-conduction thru the suubstrate
-transfer from substrate to heatsink
maybe the way the dies bond to the alumina allow for better heat transfer but as the current ramps up the limitations of the other two factors overwhelm it.
Last edited:
NanoGadget
Well-Known Member
Not sure if this applies in this particular instance, but based on my work I do know that certain materials can have efficacy that is very similar in a particular range but diverge significantly outside that range. Could be the case here.
Greengenes707
Well-Known Member
Yes, you are getting it...it's not a one variable situation. When is it ever really.
My guess would be on the density differential and the thermal capacity/specific heat between the two materials. Combine those differences with the high wattage situation of a COB, and we see the discrepancies between one variable predictions and an actual system/end package. With smaller 3535 like packages, they seem to not reach the limits of ceramics capacity.
Here is one for you to toss around your brain...
Cree's new mid powers...all vary significantly in efficacy/efficiency/output/even voltage. Even the new 5630 on the way that is a "leap" above their 3030 entry...but actually ALL USE THE SAME DIE. Not a single difference other than final packaging(3030 vs 5630 vs 2835...). Same wafer and bins. Huge impact on output form the packaging. Samsungs big thing with the 301b is the reflectance of the package.
Last edited:
CobKits
Well-Known Member
well this may be a simplification but wouldnt a 5630 have almost twice the footprint to dissipate? i may be missing what youre laying down but that seems pretty elementary
im surprised voltage is affected significantly but droop is real when dissipation is lacking i guess
im surprised voltage is affected significantly but droop is real when dissipation is lacking i guess
GrowLightResearch
Well-Known Member
CoBs are history. Use strips. Samsung F-Series or Bridgelux EB Gen2.
CoBs are more expensive per lumen than strips.
CoBs have horrific uniformity compared to strips.
Cobs require expensive heatsinks where strips do not.
CoBs are more expensive per lumen than strips.
CoBs have horrific uniformity compared to strips.
Cobs require expensive heatsinks where strips do not.