CobKits
Well-Known Member
look at test report (or put a meter on it), meanwell datasheets are always underspec'd
Introducing CobKits.com - specializing in DIY and Citizen COBs
OLD MOTHER SATIVA
Well-Known Member
i am getting 295 w from a MW 240 48A using 4 cxm22's
thanks Cobkits..they are good gear!
DesertHydro
Well-Known Member
CITI 1212's driven at 37.5w each and one per sq ft get you this.......... that whole table which is just under 4x8 is only 1200w!
Attachments
sixstring2112
Well-Known Member
Nice man,how far along?
sureshot138
Well-Known Member
do you hang them about 7/8" from tallest bud? hard to tell from pic.
HydoDan
Well-Known Member
Looking good DH... looking good!!
DesertHydro
Well-Known Member
Day 35 today. Thanks. I'm bumping up the food as I was seeing deficiencies at my normal hps rates. These seem to be insatiable under the cobs. They are Calcium fiends
DesertHydro
Well-Known Member
Overall they are about 12-16" away I think but some of the bigger buds are about 4-6" away and no problems yet.
DesertHydro
Well-Known Member
Thanks. Super happy with this run so far. Just getting my drippers dialed in to take some of my workload away. I have to leave town for a week during my flush week so I'm kinda worried but I think it will be ok. I have a friend who grew for a dispensary who can come by and check up on them. Looks like my gpw will do quite well this time. These are all still untested seeds so it's still a roll of the dice at this point. Think I have a few keepers though. I will be getting these plants tested and papered when done
Manfromthenorth
Well-Known Member
A+++++++ seller quick shipper and stuff is getting here in one peace thanks a lot will be doing more business in the feature for sure.
Great, thank you. I have never seen these test reports before. Am I correct in assuming most people running a HLG 185 48A are wiring their COB's in parallel and using constant voltage mode (where all the typical warnings around thermal run-away and variable current apply)?
CobKits
Well-Known Member
youd have to unless you were running low voltage cobs (12V, 24V) or strings of individual 3V diodes
DesertHydro
Well-Known Member
mine are in parallel running 12-20 hours a day and no issues at all. i run good sized heatsinks, and have plenty of airflow in my room and all is well so far. with the new air cooled hood setup i will have so much air flow over them they will be running cooler than ever. i too was worried about the warnings about that but so far so good.
CobKits
Well-Known Member
"A" drivers prevent thermal runaway by limiting voltage.
CobKits
Well-Known Member
watched the most excellent techtalk with @Greengenes707 and @Growmau5 today, (http://tinyurl.com/techtalk5) and was thinking about their observations on applicability of certain chips depending on application current
their consensus was cree has an advantage at low current, and veros/citi/luminus had an advantage above 2A or so. This motivated me to get going on the sphere (new setup, identical active cooling on real heatsinks on all chips to give most accuracy esp at high currents. I figured out a way to slightly modify the heatsinks instead of attempting to cut the sphere. I plan on starting a new thread for just sphere data and discussion soon)
1216 vs. 1818gen6 vs. CXB3590DB (cree was fresh from cutter this week and is highest bin available in that color AFAIK)
I thought the 1216 would do a little better TBH. about 3-4% less relative efficacy than the CXB in the usual 30-60W range. 1216 needs to be run at about 80% of the wattage of a CXB to have an equivalent efficacy (not bad for 50% of the price tho, if you can keep heatsink costs under control)
CXB with a very slight edge in efficacy (1-2% relatively) below 30W. Not as pronounced as i would anticipate.
1818 essentially equal to CXB from 30-50W where then 1818 pulls ahead above 50W
as we often see, chips are converging below 30W, the 1216 and 1818 are almost identical.
so this is congruent with GG's observations (though the crossover is probably shifted, he was probably talking gen5 citi and not sure which vero. I have gen7 V29 and will run soon on new setup. i expect vero C to be right there with 1818 and Vero B and D to be closer to CXB, just a guess)
Worth noting that CXB has the least dies of all of those chips (144). A testament to Cree's die quality for sure, and a solid performer esp at low current.
from here on out i will include spectra with my tests. raw data available in excel if somebody wants to get all QER with it. 1216 and 1818 similar as expected. CXB with an ever-so-slightly broader phosphor peak and the blue peak is a little lower than the citis
their consensus was cree has an advantage at low current, and veros/citi/luminus had an advantage above 2A or so. This motivated me to get going on the sphere (new setup, identical active cooling on real heatsinks on all chips to give most accuracy esp at high currents. I figured out a way to slightly modify the heatsinks instead of attempting to cut the sphere. I plan on starting a new thread for just sphere data and discussion soon)
1216 vs. 1818gen6 vs. CXB3590DB (cree was fresh from cutter this week and is highest bin available in that color AFAIK)
I thought the 1216 would do a little better TBH. about 3-4% less relative efficacy than the CXB in the usual 30-60W range. 1216 needs to be run at about 80% of the wattage of a CXB to have an equivalent efficacy (not bad for 50% of the price tho, if you can keep heatsink costs under control)
CXB with a very slight edge in efficacy (1-2% relatively) below 30W. Not as pronounced as i would anticipate.
1818 essentially equal to CXB from 30-50W where then 1818 pulls ahead above 50W
as we often see, chips are converging below 30W, the 1216 and 1818 are almost identical.
so this is congruent with GG's observations (though the crossover is probably shifted, he was probably talking gen5 citi and not sure which vero. I have gen7 V29 and will run soon on new setup. i expect vero C to be right there with 1818 and Vero B and D to be closer to CXB, just a guess)
Worth noting that CXB has the least dies of all of those chips (144). A testament to Cree's die quality for sure, and a solid performer esp at low current.
from here on out i will include spectra with my tests. raw data available in excel if somebody wants to get all QER with it. 1216 and 1818 similar as expected. CXB with an ever-so-slightly broader phosphor peak and the blue peak is a little lower than the citis
Attachments
Last edited:
CobKits
Well-Known Member
1216x2 simulated. for same chip cost as cxb you could put 2 on a pinfin at half the wattage per chip (same total wattage per heatsink) for better efficacy. im not sure why multiple chips on a heatsink arent a thing. if you utilize the lens holes in a 133mm heatsink, you can put 4 35mm holders on it (on the 133s i currently have, the 4th chip falls over the 7mm cable hole- i will offset it 45 deg next time.
i only know of one person doing this, and its 2 regular cobs+1 blue cob on a pinfin
i only know of one person doing this, and its 2 regular cobs+1 blue cob on a pinfin
Last edited:
MrTwist1
Well-Known Member
I hear what you're saying, but doesn't driver cost increase significantly?
Silver or lead
Well-Known Member
I plan on doing this in about a month, 2 cxm22s on a single heatsink, one 3000k 80 and a 3500k 90 to get two peaks in the red.
What does the simulation for that look like on that graph? I suspect it is a bit better than the 2 1216s.