What case temps (Tc) are you folks seeing with your citizen 1212s?
- Thread starter Hairiest_Stamen
- Start date
CobKits
Well-Known Member
what a great fit -i have a customer that did a whole room with coolers like these (maybe not exact model) but he had 100s of them
it would be preferable it you could rotate it to fit the square portion as the metal not directly under the LES also cools. but if your fan is blasting on the little bit of overhang that cools it a little bit too.might make for better airflow and less noise
not sure how many you have of these but if i was doing it id consider ditchiing the holder and mount the chips directly. $2-3 holders add up when your heatsink is only $2 and the chips might be as low as $7 each. or heck you could save the tapping and use thermal epoxy
you should be able to run that fan at 5V or 9V and make it quieter
im still not clear where youre measuring the temp tho. if you got a way of measuring tc with holders on id love to see how its done
if you download citi calculator you can plug in different temps to see the effect of thermal droop at various temps. its also handy for back calculating delta T vs voltage. i say delta T because you cant absolutely compute temp due to forward voltage variations from chip to chip. but you can start up cold take a quick voltage reading and call it 30C. then estimate Tc by plugging in your delta voltage at load relative to what it was cold
it would be preferable it you could rotate it to fit the square portion as the metal not directly under the LES also cools. but if your fan is blasting on the little bit of overhang that cools it a little bit too.might make for better airflow and less noise
not sure how many you have of these but if i was doing it id consider ditchiing the holder and mount the chips directly. $2-3 holders add up when your heatsink is only $2 and the chips might be as low as $7 each. or heck you could save the tapping and use thermal epoxy
you should be able to run that fan at 5V or 9V and make it quieter
im still not clear where youre measuring the temp tho. if you got a way of measuring tc with holders on id love to see how its done
if you download citi calculator you can plug in different temps to see the effect of thermal droop at various temps. its also handy for back calculating delta T vs voltage. i say delta T because you cant absolutely compute temp due to forward voltage variations from chip to chip. but you can start up cold take a quick voltage reading and call it 30C. then estimate Tc by plugging in your delta voltage at load relative to what it was cold
Last edited:
Hairiest_Stamen
Well-Known Member
Yeah, the rotation is due to the nature of the where metal is available to tap for the holder... I would have loved to have it line up, would have been super clean. Without the rotation the holes float above fins.
It's a small enough percentage of area that I'm not going to fret about it. Was happy that all of the area under the LES (and more). At 2+ amps
I'd probably be rethinking the mounting.
Good suggestion about the thermal epoxy. Going to order some of that now and try it out.
May try and play around with the cnc and machine some water blocks and ditch the fan. got a bunch of left over plate and some small pumps. Not sure if potential lower temps will buy me any meaningful performance.
No really accurate method with the holder I don't think... tc measured by wedging the thermocouple end in between the holder contact and the negative pad. With a bit of luck you can get the holder arm to apply pressure to the thermocouple end right against the pad. I joined the ends of the thermocouple with a dab of silver solder and made a small ball for the holder arm to push on. Had to massage the solder ball with a small hammer to get it to work. Without a decent self standing magnifying glass it's almost impossible to get this right.
Also, adde some copper solder (old ass stuff, no idea its origins) to another thermocouple and then flattened it out into a thin pad with the hammer. then wedged this under the holder arm and that produced similar results... a lot more time consuming. Had to use a butane torch to get it to flow lol.
Attachments
CobKits
Well-Known Member
thats interesting. ive often wondered about soldeirng a tc right into the joint (obv wouldnt work with a holder)
Hairiest_Stamen
Well-Known Member
Would be nice if citizen had a seperate pad like the crees but what do you want for less 50% of the cost.
Greengenes707
Well-Known Member
Various 3rd party certifications will test Tc via thermocouple. ETL, TUV, DLC...
And when issues arise, Tc measurements are part of warranty and safety claims. When you show Cree, Bridgelux, Citi...a blown/failure/DOA chip, they want a full inspection of the entire system...Tc the biggest metric.
Otherwise, they tell you to pound sand and buy more chips.
DrBlaze
Well-Known Member
Wow, its like you read my mind. I was using the citizen tool last night to estimate a build using the gen5 1212's you have (was thinking about the epoxy as well). I got the idea from this post: https://www.rollitup.org/t/driver-selection-help.945316/#post-13668467, where Danielson999 pairs 6x Samsung 1120mm F Series with an HLG-240 and due to high efficiency, gets enough light for a square meter.
So using the citi tool to try and match it, I found that 12x 1212 gen5 at around 20w and 35c will give 170l/w and should fit on an elg-240-36a. A super cheap build with these heatsinks and extremely efficient.
My question is, in your experience how close is the citi tool to your real-world measurements? Is it pretty much on, or a bit conservative like the info from the spec sheets?
CobKits
Well-Known Member
to me their calculator seems low relative to other manufacturers (i.e they seem to beat the performance predicted by calculator)