Switch Mode LED driver PCB is embedded in potting.....cool running?

[treez]

Hello,

The SD24 (9W) LED flashlight has its switch mode led driver encased in a tube of potting material.

SD24 LED flashlight
https://www.911lights.com/products/nova-sd24

The switch mode LED driver can be seen at 47 seconds in the following demo video......

https://vimeo.com/38638873


.....So the PCB containing the switch mode LED driver is encased in potting compound, the potting compound actually forms the enclosure of the PCB too, as the potting is poured into a dye containing the PCB.

Does this kind of potting mean that the cooling for the PCB's power components is better than if the PCB was simply screwed to the inside of a hollow plastic enclosure?

I mean, does the "encasement" of the PCB's components in the potting help to draw heat away from the power components?.......does it run cooler like this than if the PCB's components were just exposed to the air? (with no airflow other than convection)

 

[betwixt]

It depends on the potting material. Epoxy resin may conduct heat away more than air convection but silicone (RTV) compounds generally act like blankets and trap the heat inside.

Brian.

 

[treez]

Would you say that the potting at 47 seconds in the video is RTV or Epoxy? (the max voltage in the led driver is 60V.)

Surely, one would always use a potting compound that made the circuitry run cooler than it would in air?

 

[FvM]

I would emphasize other aspects in the potting discussion.

- Even standard epoxy or silicone resin has a much higher thermal conductivity than air. It will spread the power dissipation of small hot parts over the device volume.
- Potting doesn't increase the surface device much. All heat must be still transmitted to the embedding air unless the device is mounted to a heatsink or metal plate.
- If a heatsink comes into play, thermally enhanced potting resin is more interesting

 

[betwixt]

Unfortunately, my internet connection isn't fast enough to see video so I can't tell what resin was used but FvM is giving good advice.

Brian.

 

[ark5230]

Is it that the potting is intended for some other purpose like copy protection rather than handling heat related issues and care is taken that potting does not deteriorate the performance claimed.

 

[tpetar]

I will say its protection from water, dust and vibrations. Used diodes for this product are not power diodes and I dont think they have cooling in mind. I dont think that inside is something smart worth for cloning.


I saw video I dont think this product is something special "SD24 9W". Get red led UltraHell 35mA 10-15deg 5mm 35000mcd and see what is light. Before few years I make stop third light for my car with 15 mentioned leds, and I removed this after few days, this is very strong and blind drivers behind.

This product in video dont have 9W of led power, or have 9W but including power dissipation all others losses. This 9W mark looks like model ID or marketing trick as usual with power inverters and other products.

See CREE leds for 9W-10W you cant look in direction of light source.

 

[treez]

I dont think that inside is something smart worth for cloning.

..thanks but how do you know that what is inside the potting is not worth cloning, have you opened one up?

by the way, its 24 LEDs and each led has 150mA in it.......(ie when its flashed ON, each led has 150mA in it)

Even standard epoxy or silicone resin has a much higher thermal conductivity than air. It will spread the power dissipation of small hot parts over the device volume

....yes but won't there end up being a tiny air gap between the potting compound and the components...?...i mean surely you can't expect the potting compound to "hug" right up to the components when it sets?

 

[tpetar]

..thanks but how do you know that what is inside the potting is not worth cloning, have you opened one up?

No my friend, I didnt open this product, but based on video I conclude that. Nothing special.

by the way, its 24 LEDs and each led has 150mA in it.......(ie when its flashed ON, each led has 150mA in it)

Based on video I will say power is around 3W, no more then that. You should see real 3W, 5W, 10W leds then all is clear.

If I have time I will make picture of 5mm 10-20° Red UltraHell 15cd 25cd 35cd 50cd 35mA.

....yes but won't there end up being a tiny air gap between the potting compound and the components...?...i mean surely you can't expect the potting compound to "hug" right up to the components when it sets?

Post picture of air gap. Why air gap?

 

[FvM]

If I were in the LED flasher business, I would probably want to know the basic design parameters and device properties: LED power, input current, efficiency, linear or switched mode control (I guess the latter), switching frequency, EMI properties.

You don't need to open the enclosure to know these parameters, but of course a device for test.

 

[betwixt]

A good potting resin will flow as liquid before it cures and will fill all the cavities and gaps in, under and around components. It's usually a two part chemical which is mixed at point of use so it has chance to flow before the curing reaction has time to start. So without doubt it 'hugs' all the surfaces it touches and adheres to them. Usually, it's used for environmental protection, either against moisture or vibration but sometimes just as an insulator where high voltages have to be near touchable areas. For example, I worked on mains powered drilling equipment with a 'trigger' style speed control and 240V AC was present inside the handle. If it became damaged or wet (especially likely in a harsh working environment) there had to be no risk of conduction to the operator so after testing the whole handle was injected with resin.

And.... beware of claims about LED power. I have a '20W' LED spotlight here that lasted only a few seconds of use. I opened it up to see what had happened and found a burned out array of 10 tiny indicator LEDs, a bridge rectifier and a capacitor which was insufficiently rated and had gone short circuit! I would estimate it's real output to be less than 0.5W.

Brian.

 

[tpetar]

I bet that Treez will be first customer for Philips TLED (200lm/W).

 

[treez]

I bet that Treez will be first customer for Philips TLED (200lm/W).

..funny you should mention it, just about two weeks ago , i read in the sun newspaper that phillips had some new leds for release in 2015, which will have at least twice the lumens/watt of the best present leds.....i wonder if this is true?, ive heard it so often before from others

 

[tpetar]

..funny you should mention it, just about two weeks ago , i read in the sun newspaper that phillips had some new leds for release in 2015, which will have at least twice the lumens/watt of the best present leds.....i wonder if this is true?, ive heard it so often before from others

Yes, I hope it is true. I saw some video demonstration with measuring, I must say this is really impressive. For now CREE power leds gives enough lights. Check cheap Police hand flash lights 1000-1200lum per diode + lens...