Introducing CobKits.com - specializing in DIY and Citizen COBs

OLD MOTHER SATIVA

Well-Known Member
so how did my nichia test post stack up...[svp]

i was not trying to clutter your thread..

yes first day with a new par meter and my first cob build..

i posted it a few places and got zero comment....if you please
 

CobKits

Well-Known Member
i had to search back to find the post but my gut tells me a single measurement is of little use. you need to measure something side by side with the same setup (in other words your 550 umol has no reference to my data. also if youre comparing optics you really need to look at the effect of average par over an area as spot values are misleading. remember- youre not "making" more light with a reflector, just (lossily) bouncing it around

i try not to even compare data from different days/different power supplies/etc on my own tests, even with the same chips/heatsinks/par meter. any graphs ive ever shown here were generated in one long sesh, some took many hours of just measurement

also any lenses or optics are a no-no. i test bare cobs either soldered or in ideals. not even BJB as they reflect a little per robin's testing- like 10%. even if you are using same optics and lenses on multple chips you can get a lot of variation just by moving an inch. ive found with the bare cob measurements at 12" i can move an inch in either direction and maybe move 1-2 ppf (<<1% in most cases) on the meter- so its a system that works for me for now until i get the sphere (which i bought... ). I have to figure out which method im going to use- the classic method of using the photodiode with a picoammeter for bulk lumen measurements... or using the ocean optics...to measure intensity and spectra and integrate PAR from that or just the Li-Cor directly. all have their ups and downs. the latter is more useful to us in terms of turning over more measurements quickly
 

CobKits

Well-Known Member
power supply fail.

got one of these to see how far i could push a 3618 and it wasnt up to the task

http://www.ebay.com/itm/182292429613

with no current regulation and limited voltage regulation i couldn't get it to deliver less than 4.3A to a single cob (out of spec), and it couldnt light up two in parallel. i guess well have to wait for that HLG-480 for that

so now i have 2 3618s humming happily on a 320h-1400b in series instead at 205.3V x 1.42A = 291W. seems like theyd run forever at this load. despite the case temps calculated, the active heatsinks are cool to the touch

upload_2016-10-13_22-56-47.png

so i guess that will be my tent PAR test when i set up an extra 3x3- 1 3618 on a 320-2800 vs 2 on a 320-1400
 

bleak

Well-Known Member
well (2) 2pole wagos

yes it will drive an 1825
Awesome, thanks for clearing this up. Can I run my shopping list by you one more time? Sorry to be a pain, but I can't afford to miss any small parts in the order (I'm in a very remote location). Heres what I'm looking at:

- CLU058-1825 light engine
- Meanwell HLG-120H-48A driver
- 2x Wago 2-pole connectors

Anything else I should add to the order? Last question, is the discount code still "RIU"? Thanks in advance for your help!
 

bleak

Well-Known Member
maybe one of these if you dont have a 9-12V power supply handy

http://cobkits.com/product/apv-12-12-12v-driver-for-fans/

you could use wagos for those or wire nuts, or just twist and tape them ill throw some wire nuts in if you remind me when you order
Awesome, I'll definitely add a 12v power supply. Thanks about the wire nuts also, I'll drop you a message once I've ordered.

Is the discount code still "RIU"?
 

kaivorth

Active Member
power supply fail.

got one of these to see how far i could push a 3618 and it wasnt up to the task

http://www.ebay.com/itm/182292429613

with no current regulation and limited voltage regulation i couldn't get it to deliver less than 4.3A to a single cob (out of spec), and it couldnt light up two in parallel. i guess well have to wait for that HLG-480 for that

so now i have 2 3618s humming happily on a 320h-1400b in series instead at 205.3V x 1.42A = 291W. seems like theyd run forever at this load. despite the case temps calculated, the active heatsinks are cool to the touch

View attachment 3804550

so i guess that will be my tent PAR test when i set up an extra 3x3- 1 3618 on a 320-2800 vs 2 on a 320-1400
What do you think is better? 2 3618s at 150w each or 4 1825s at 75w each?

Less wiring with the 3618, but more cooling required.

I wonder if you could kapton tape a Cryorig M9i CPU cooler to these. They're $20 coolers that perform good as a lot of $50 ones.
 
Hello, im very new at led's tech and to this forum too. I have a 10sq ft tent and i want to build a led light with the components in your website, but im very confused on what to buy to build a led light that fully supplies the space im growing. Also is there a thread or somewhere i can read a step by step to build the led light?

Thanks
 

CobKits

Well-Known Member
Hello, im very new at led's tech and to this forum too. I have a 10sq ft tent and i want to build a led light with the components in your website, but im very confused on what to buy to build a led light that fully supplies the space im growing. Also is there a thread or somewhere i can read a step by step to build the led light?

Thanks
if you search youtube for growmau5 channel he has some excellent videos covering the fundamentals of DIY LED. great place to start then i can answer any specific questions you might have
 

CobKits

Well-Known Member
Suggestions for a single driver to drive 4 of them at 1400ma? Dimmable a plus :D

Should be my last question though, I know I've been as nusiance. Will defiantly be directing friends to your site though.
hlg-320h-1400B is what you want. great match for the chips and 140mm heatsinks
 

HolyGhost23

Well-Known Member
if you play with the citi calculator the 1825 is the most efficient for lm/w... at least above 2500mah
 
Last edited:

CobKits

Well-Known Member
at all currents actually, its just not cost justified down low where its only a few percent more efficient than the cheaper chips

its all about the series : parallel ratio
 
Top