MicrobeHunter.com Microscopy Forum

https://www.youtube.com/watch?v=-qyE-fHPZYk

So what would happen if you put this inside an LED lamp house that gets hot with a fanless heatsink?
Asking for a friend…

Apparently nothing much will happen. The LED itself and electrical components of its circuit warm the surrounding through convection, not radiation. The white paint reflects radiation.
Still, it is interesting and potentially important news. Although a high proportion of houses today have glass walls which will not be painted white, and will hence absorb heat. But this is stepping outside the scope of this forum...

Hobbyst46 posted whilst I was drafting this … but here’s my take anyway.

Probably not much … except that your friend might be able to turn the nominal brightness down a little.

The heat generated by an LED is mostly dissipated from the substrate … the light-beam itself is effectively at room temperature !

My suggestion would be a bigger/better heatsink.

MichaelG.

I am kind of curious as to why people worry about heat disappation from leds? Up to say 10 watts if led it's not an issue. 100watt halogen equivalent. I am seeing that there are getting to be some crazy high powered leds on the market but for microscopy above 20 waats of led is not necessary unless there is something I don't know about. I have been experimenting and currently have a 3 watt ebay China spot light on my standard junior with an Amazon dimmer for under $20, it works great but isn't elegant. I will buy one of these in 5 watrn for my Zeiss standard and a couple of the 10s for my Universal...

https://www.ledsupply.com/led-kits/10-w ... module-kit

If your LED gets too hot it may be a little dimmer and the emitted spectrum may not be as spec. Still, there's probably some folks that had 100w halogen for a long time and remember melted parts and small fires and other fun stuff

Gatorengineer64 wrote: ↑

Sun Oct 24, 2021 9:34 pm

I am kind of curious as to why people worry about heat disappation from leds? Up to say 10 watts if led it's not an issue …

It can be an issue if the heatsink is inadequate

Heatsinks are rated in terms of °C per Watt [or Kelvin per Watt for the SI pedants]
That prescribes the expected rise above ambient temperature.

If your room is at 20°C and you would consider 40°C a reasonable maximum, then a 10Watt LED needs heat-sinking of 2°C per Watt
… This can, of course include the body of the microscope, if you arrange things carefully : But this would be a typical passive heatsink of that rating: https://uk.rs-online.com/web/p/heatsinks/0158562

Other styles are available … but I think that one seems a reasonable point of reference.

MichaelG.

BramHuntingNematodes wrote: ↑

Sun Oct 24, 2021 11:23 pm

If your LED gets too hot it may be a little dimmer and the emitted spectrum may not be as spec. Still, there's probably some folks that had 100w halogen for a long time and remember melted parts and small fires and other fun stuff

Those 100W halogen bulbs are installed within a dedicated housing that is installed outside the microscope stand. The housing is a metallic box that is designed to dissipate heat outside, via cooling slots, aluminum fins etc. Some heat still enters the optical train as IR radiation, but the bulb is sufficiently distant from the glass to prevent damage.
If a LED is installed within the microscope base, directly below the field aperture, heat from the LED circuit components is conducted within the base. This heat must be efficiently eliminated. The LED itself does not radiate IR so the light beam does not heat anything. These considerations are relevant for LEDs of a few watts and north.

I mean i don't know i guess. I got here this old Zeiss pol microscope with an 8 watt bulb in the base and it has essentially no consideration for heat and its fine. Good enough for the engineers at Zeiss anyway. More relevant for LEDs is the die itself is tiny and as I said the light will dim or be off spec if it gets too hot in that tiny area. Could maybe shorten life too I guess. That's more of a secondary concern as far as I can tell as the life could still well be much longer than any incandescent. I really don't think a 10w led is going to cause any damage to a scope.

.
A 10W LED produces a lot of heat. That is not something you want inside your microscope.

An incorrectly installed high Watt LED light might not cook your sample - but it could well cook your microscope.

Quote below is from this link: https://www.lrc.rpi.edu/programs/NLPIP/ ... D/heat.asp

"The need to ensure heat sinking with LED systems is also important to consider when these systems are installed in applications. There must be sufficient means to conduct the heat away from the system, or ventilation to cool heated surfaces by convection."

Good link ^^^

MichaelG.

Well sure it's a fine link but again this is discussing junction temperature which is a tiny area a ten watt led produces tem watts of heat which is not a lot especially when housing it in a massive metal object. Tell you what I have plenty of operational 10w LEDs and a thermocouple let me run some experiments.

BramHuntingNematodes wrote: ↑

Tue Oct 26, 2021 1:26 pm

…
Tell you what I have plenty of operational 10w LEDs and a thermocouple let me run some experiments.

That would be great … Thanks Bram

Could I please ask that you make sure the devices are running at full power throughout the test.
For background; here is a rough check I did on the current drawn by a nominal 10W LED at various Voltages:
. .

MichaelG.

Of course at full power. The pot is log so at any other position I would have trouble estimating what the wattage is. It is being powered by a 1000mA driver, the famous Buck Puck in fact.

1000mA is typically far from maximum power for a 10w led. My Cree XML2 has a rated max current of 3000mA. So even the 2100mA buckblock is not pushing it to the max. It still pumps out a ton of light at 1000mA, though. And, as you say, heat is not a problem when you are running it at such low power.

Here is my experience.
A retrofit 10W neutral white LED illuminator is stuck in the rear of the microscope. Made by a professional, and includes an aluminum heat sink, with fins. The heat sink itself is not inside the microscope base.

When it runs at full power and room temperature is around 30C, the heat sink is very hot to the touch. I believe, close to 50C.
The LED itself rests inside the base of the microscope, and the base is slightly warm. Not more.

The microscope base can be an adequate thermal mass or not, depending on material and mass, but the LED illuminator produces a lot of heat.


BTW, typical forward LED voltages are ~3V AFAIK, so for 10W the current should be ~3A.

Hobbyst46 wrote: ↑

Tue Oct 26, 2021 2:33 pm

BTW, typical forward LED voltages are ~3V AFAIK, so for 10W the current should be ~3A.

For clarity … mine was a COB LED , with multiple LED sources in a common matrix
Rated at 9 to 11 Volts
Similar to the one shown in this listing: https://www.ebay.co.uk/itm/154451507928

MichaelG.

Ah yes, I looked up the documentation and I guess it really is running at a little less than 5w. It was the amperage all the spectrum curves were specified at so that's what I went with. That would, then, be my recommendation as they are bright enough for immersion phase you would likely need to go higher only if you bought a cheap led die or are doing fluorescence. After running about half an hour at full power, the pad immediately behind the board was 131 f, which is not comfortable to touch but is not particularly dangerous. After letting it go another 45 minutes or so under an unventilated rubber bucket, the temp reached 141 f, also not disturbing. This was with no attached heat sink and the area directly behind the die on the board. The ambient temperature in here is around 74 f. Probably we could figure out the heat transfer coefficient and work out what would happen if we doubled the energy in the system or we can add the disclaimer that Bram's blase attitude towards heat management is only verified to be applicable in systems under 5 watts

MichaelG. wrote: ↑

Tue Oct 26, 2021 2:46 pm

Hobbyst46 wrote: ↑

Tue Oct 26, 2021 2:33 pm

BTW, typical forward LED voltages are ~3V AFAIK, so for 10W the current should be ~3A.

For clarity … mine was a COB LED , with multiple LED sources in a common matrix
Rated at 9 to 11 Volts
Similar to the one shown in this listing: https://www.ebay.co.uk/itm/154451507928

MichaelG.

Like the features it lists :)

Though the most important point is of course the last one.
.

Gatorengineer64 wrote: ↑

Sun Oct 24, 2021 9:34 pm

I am kind of curious as to why people worry about heat disappation from leds?

Also: Halogen bulbs are designed to work at high temperatures 260 degrees ish. A LED is not, its output is going to be affected by higher temperatures as well as life expectancy (or they can burn altogether).

75RR wrote: ↑

Tue Oct 26, 2021 3:23 pm

Like the features it lists

Though the most important point is of course the last one.
.

I bought mine before planet-saving was completely de rigueur

For interest though: Here is the makeshift heatsink that I was using for that test
… it’s a channel with flanges [sorry, no three-quarter view]
As photographed, it is being under-run to show the nine LED sources
… presumably wired in three parallel strings of three to get the Voltage rating:
. .

MichaelG.

who the heck is "mHuntingNematodes" I dont remember you being at the family reunion

Should call this variant SpamHuntingNematodes

maybe ScramHuntingNematodes.