LEDgardener

How do you guys measure the temperatue on a heatsink?

And what temperatures are acceptable and why?
(I guess, I can use my QB288 heatsinks as a point of reference)

As I understand, the cheap IR thermometers are set to a threshold that doesn't necessarily work with polished aluminium at higher temperatures ???

I never measured from the heat sink. Aluminium does not melt that easily. The temperature you need to follow is your led, the strips usually have tc points meant for measuring, not sure about cobs. If you don't have something to measure the tc point, you can use a lazer thermometer to measure the strip casing.

My strips can handle 80c, and they run at under 40c.

On higher-end IR thermometers, you can change the emissivity value to something that works better for anodized aluminum. I've read .77 is a good place to start.

Aside from that, taping a thermocouple on there would be a good way to do it.

unkle_psycho wrote: ↑

Mon Jul 02, 2018 7:44 pm

I never measured from the heat sink. Aluminium does not melt that easily. The temperature you need to follow is your led, the strips usually have tc points meant for measuring, not sure about cobs. If you don't have something to measure the tc point, you can use a lazer thermometer to measure the strip casing.

My strips can handle 80c, and they run at under 40c.

Thanks for the feedback

Melting point is like 600-700°C , so it's more about efficient cooling and dissapation.

I think the H inFlux strips are rated for 90°C.

I'm trying to figure out if tit's worthwhile getting a bigger heatsink.
I have 30x20x2mm U-channel, and I want to campare it to 30x30x3mm which is about twice as expensive

LEDG wrote: ↑

Mon Jul 02, 2018 7:56 pm

On higher-end IR thermometers, you can change the emissivity value to something that works better for anodized aluminum. I've read .77 is a good place to start.

Aside from that, taping a thermocouple on there would be a good way to do it.

Thanks!

Great idea using a thermocouple


I think I'll use a cheap IR thermometer for the strips and a thermocouple for the heatsink.

I use a DS18B20 probe on my heatsink that's hooked up to an Arduino. My heatsink usually reaches about 50°C when the light is cranked all the way up.

PurpleGunRack wrote: ↑

Tue Jul 03, 2018 4:28 am

unkle_psycho wrote: ↑

Mon Jul 02, 2018 7:44 pm

I never measured from the heat sink. Aluminium does not melt that easily. The temperature you need to follow is your led, the strips usually have tc points meant for measuring, not sure about cobs. If you don't have something to measure the tc point, you can use a lazer thermometer to measure the strip casing.

My strips can handle 80c, and they run at under 40c.

Thanks for the feedback

Melting point is like 600-700°C , so it's more about efficient cooling and dissapation.

I think the H inFlux strips are rated for 90°C.

I'm trying to figure out if tit's worthwhile getting a bigger heatsink.
I have 30x20x2mm U-channel, and I want to campare it to 30x30x3mm which is about twice as expensive

LEDG wrote: ↑

Mon Jul 02, 2018 7:56 pm

On higher-end IR thermometers, you can change the emissivity value to something that works better for anodized aluminum. I've read .77 is a good place to start.

Aside from that, taping a thermocouple on there would be a good way to do it.

Thanks!

Great idea using a thermocouple


I think I'll use a cheap IR thermometer for the strips and a thermocouple for the heatsink.

Sorry, just messing with you.

Heatsinks have serious diminishing returns, so if you double your heatsink, you don't get anything like double cooling. In that sense it would be a pity to pay much more for a little bit more capacity.

I would'nt expect anything like double coolling properties, but if it's what keeps the strips under the nominal performance temperature of 55°C it's well worth it

I found some tape you can put on glossy surfaces that adjusts to the standard emissivity value 0,95

Here there is a calculator, which I imagine can calculate how much they differ in capacity. Don't know if this would be useful?

https://www.myheatsinks.com/calculate/t ... plate-fin/

PurpleGunRack wrote: ↑

Sat Jul 07, 2018 5:38 pm

I would'nt expect anything like double coolling properties, but if it's what keeps the strips under the nominal performance temperature of 55°C it's well worth it

I found some tape you can put on glossy surfaces that adjusts to the standard emissivity value 0,95

For heatsinks, a little bit of airflow goes a long way. So rather than increase the size of the heatsink, place a small fan somewhere nearby.

If you want to get into the weeds a bit, look up free convection vs. forced convection.

Here's a chart of typical convective heat transfer coefficients: https://www.engineersedge.com/heat_tran ... _13378.htm

The h of free convection of a plate is ~5 W/m2-K, while low speed airflow over a surface is ~10 W/m2-K. Basically, if all other things were constant, your heatsink will be twice as effective with just a bit of airflow.