How hard can you push the SST-50?

MikeAusC

Enlightened
Joined
Jul 8, 2010
Messages
995
Location
Sydney, Australia
To solder to your Copper adaptor, I would make a clamp so you can attach it to the barrel of your soldering iron - I assume you have a Temperature adjustable Iron.

Then look at the SST-50 datasheet and see the recommended temperature ramp-up ramp-down profile for maximum reliability.
 

darkzero

Flashaholic* ,
Joined
Oct 7, 2003
Messages
4,459
Location
SoCal
I've ran 9-10A to a SST-50 a few times by accident during some testing. Turned angry blue but never killed the emitter but it was reflowed onto a MCPCB which was epoxied to a copper slug.
 

Roland Gama

Newly Enlightened
Joined
Aug 29, 2008
Messages
83
Just tried Der Wichtel's 9.1 amp rated driver for the sst-90 on a KD SSR-50.
The SSR-50 is mounted with arctic silver on to a copper slug and epoxied around doughnut style.
Working great for short bursts till now.
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
I will reference this thread with regard to using stars vs not using stars.

https://www.candlepowerforums.com/threads/150244

Wow nice info. Guess the insulating layer in the MCPCB really hoses stuff up. Good news to that is the max specs for most the LEDs are probably calculated with it on a MCPCB since thats how most manufacturers would be putting them down. Guess that means direct soldering or AS/AA would let you push it a bit harder :)
 

Der Wichtel

Enlightened
Joined
Mar 12, 2007
Messages
829
Location
Germany
That's why some companies are using ceramic pcbs. They don't conduct heat as good as copper but they perform better without the need of an insulation layer
 

CKOD

Enlightened
Joined
Aug 3, 2010
Messages
708
That's why some companies are using ceramic pcbs. They don't conduct heat as good as copper but they perform better without the need of an insulation layer
Woah, ceramic PCBs to mount the LED's on? or do you mean a ceramic carrier like luminus has on the SST chips?
 

Der Wichtel

Enlightened
Joined
Mar 12, 2007
Messages
829
Location
Germany
it's a pcb. Just like a aluminum star. You can get those in various thicknesses. And with vias which are not possible with metal core pcbs.

I saw them using this stuff to build tiny antennas at our university as well.
 

IsaacHayes

Flashlight Enthusiast
Joined
Jan 30, 2003
Messages
5,876
Location
Missouri
Very interesting things in this thread. Looks like using bare emitters over stars is not better like the luxeon days when it comes to epoxy, since the stars on most SMT LEDs can be soldered to the star instead of thermal adhesive. Even the substrate of the star is better than epoxy right to the heatsink for heat transfer it seems.

Also, the 4500k LEDs probably use a warm white phosphor, which those are typically more fragile to heat, so the diming on those is probably due to the phosphor heating up and loosing efficiency.


Over driving a large die that is already at a high wattage is going to get tricky compared to a smaller die (like the XR-E) that is in rough the same size package. If the SST-50 was in a larger package, then it might bet easier to over drive it.

Looks like soldering it directly to a copper heat spreader, then fixing that to a large aluminum heatsink would be the best. I'm still wating for someone to come up with a copper tower that we can solder the LEDs to, then attach that to whatever flat heatsink in the mag tube that we want to continue the thermal path. That would rock. It would just need to be a tower with a flat/round base with 2 holes for screws. Then you could put your aluminum slug however deep you want in the mag for focus (mag rebel LED,etc) screw down the copper post and you're done.
 

DIWdiver

Flashlight Enthusiast
Joined
Jan 27, 2010
Messages
2,725
Location
Connecticut, USA
There's something fishy about the Thermal photo in this link - the temperature of the MPCB on the right is at the same temperature as the LED on the left ??? The heatsinking behind them must be very different.

The caption says "With no thermal emissivity coatings". In my experience this makes any temperature evaluations pretty worthless.

I once evaluated a product which had 6 LEDs on an aluminum PCB. The backside of the PCB was bare aluminum. While the board was uncomfortable to hold, it looked the same temperature as the desktop in the thermal image. I put a piece of electrical tape on it, and that looked the same temperature as the thermocouple indicated. This is because the emissivity of the aluminum is very low, while the tape is very high.

To make accurate temperature measurements with thermal imaging, you MUST either use emissivity coatings, or know the emissivity of the surface you are evaluating. Don't count on the imager's default assumption that everything has an emissivity around 0.7
 

MikeAusC

Enlightened
Joined
Jul 8, 2010
Messages
995
Location
Sydney, Australia
Vias are definitely possible on metal core PCBs. See here

They show an excellent cross-section photo that shows how they do it - they coat the inside of the hole to insulate the copper from the metal core, then do the usual through-plating. It won't be a cheap process - you need circuit board on BOTH sides of the metal core.

Interestingly they also offer electrically-insulating Thermal Vias (filled with thermal compound) as well as the usual copper electrical vias. Of course Thermal Vias are not used through the Metal Core, because the Metal Core conductivity will be 20 times better than any Thermal compound.
 

Notsure Fire

Enlightened
Joined
Jan 7, 2008
Messages
272
Sounds like quite the project you got going on.
ae59Dd
 
Top