new ledengin 365nm LEDs

[phenol]

yes, the leds do have a hefty price tag, but the heatsinking pcb wasn't all that pricey, couple of pounds or so.
Curing glues with that one is really convenient; the uv glue im using cures in just a couple of seconds with the led 1-2 inches away. But then again, the collimated spot of a <100mW 405nm laser is just as good with that particular glue.

I got one of the older 10W LEDEngin 365nm LEDs a while back and found it helped a bit to use a UV-pass filter to keep all that visible light out. My filter wasn't perfect though -- a dichroic "woods glass" from Rosco that isn't right on 365nm. Using a photographic filter would be better, or an old Blak Ray filter on eBay as suggested on CPF.

Overall, I'd probably stick with fluorescent black lights next time. Those LEDs are really expensive, and the 10W heatsink for it was pricey too.

 

[phenol]

This pricey LZ4-00U600 4-die 365nm LED mounted on a MCPCB just arrived today. The tiny component is a zener esd protection diode.
It is probably similar or identical to the uv torch Ionlaser555 published earlier this year somewhere in this section.
I'll post more pictures when it's properly heatsinked and powered by the same Zetex driver as before.

 

[DashApple]

Nice : D , it looks to be near identical to the one I have except the board colour

IMG_0493 by TwirlyWhirly555, on Flickr

 

[phenol]

What is the part number of the one you have? yours seems to have thermistor pads, too. Mouser supplies randomly binned parts, but the typical power is supposed to be around 1.8W at 0.7A. I powered it up and it's bright enough to char white paper placed 2-3 mm in front of it at slightly less than 700mA.

While I got ZXLD1360 to make my own buck driver, those Recom bricks you're using look rather neat and I think I'll give them a try.

 

[DashApple]

I have a look and let you know soon .

How random is the binning , I was looking at the LZC-70U600 365nm , Bin U . Is it possible they would sent a different bin led ?

I do like the recom buck drivers , simple and a few of the have dimming and are small in size too

 

[phenol]

As per the datasheet, they only seem to offer full flux distribution parts and optionally bin them into 2 wavelength bins. I couldn't find any flux distribution charts, but the typ power is 5w

the 12-die led is a beastie, but the unit price is hard to swallow...

 

[DashApple]

As per the datasheet, they only seem to offer full flux distribution parts and optionally bin them into 2 wavelength bins. I couldn't find any flux distribution charts, but the typ power is 5w

the 12-die led is a beastie, but the unit price is hard to swallow...

Ok that makes sense , Thanks .

Yeah , I try and avoid looking at the price : P , I should have one of them with me in a week : D

 

[phenol]

how are you going to drive this?

 

[DashApple]

how are you going to drive this?

Recom brick , they do one with a 9-56 Volt output @ 700mA , 60V max input

Supply will be lithium cells ( 15 - 16 ) , Will probably use 16 Cells for 59.2 -67.2V and a LDO regulator to get it to 58V for the most part and as the recom is a buck current input will be lower at 58V assuming a 50V max led forward voltage so not to much power will be wasted in the regulator

Not the most efficient way to do it but its simple and will work perfectly

 

[phenol]

the recoms seem to have analog control inputs. you could make use of it (and a thermistor)to roll off the led current at elevated temperatures. it only makes sense to pamper such an investment and treat it carefully.

 

[DashApple]

the recoms seem to have analog control inputs. you could make use of it (and a thermistor)to roll off the led current at elevated temperatures. it only makes sense to pamper such an investment and treat it carefully.

That is true and in fact one of the recon PDFs have a circuit showing thermistor based temperature current limiting .

Yeah I fully agree there , ill get photos of it when it shows up .

 

[phenol]

it would be interesting to measure the led ripple and startup current with a current probe or sample the voltage across a sense resistor in series with the LED. Those uv leds have very limited tolerance for pulsed fwd current overshoot (850mA).
i have 3 recoms coming next week and i'll probe them with a dummy load first.

 

[DashApple]

Nice let me know the results , ill test the recom 48 0.7A when it arrives. The speck for the recom 48 is 300mVp-p max according to the data sheet .

What models did you get ?

I will try to add the temperature feedback current control into the build .

 

[phenol]

RCD-24-0.70/VREF. it has a reference voltage output, which can be used to power a thermistor dimming circuit.

 

[DashApple]

RCD-24-0.70/VREF. it has a reference voltage output, which can be used to power a thermistor dimming circuit.

Nice , I believe the RCD-48-0.7 I ordered has a VRef as well .

 

[phenol]

I mounted the 4-die led to this chunk of aluminum. the Recom brick with some extra rfi filtering and esd protection is tucked in the cavity of the heatsink.
There is also a PTC thermistor bolted adjacent to the led. The current to the led rolls off rapidly above 60c, but with just 10W dumped into such a massive chunk, it never gets much hotter than 40-45c even without that oversized cpu fan