georges80
Flashlight Enthusiast
I've been working on a new high voltage current regulated boost converter and finally got time to assemble a prototype yesterday. More work to do to decide on dimming/pwm input control, but the basic driver is performing well with the multiturn trimpot.
The hyperboost can run from a nominal 6V to 50V input. Output voltage is up to 70V and open circuit protected (maxes out to 70V).
Board diameter is 1.4" with components only on the top side. Heatsink attachment is the same well proven scheme as for hipCC and hipFlex, a piece of thermal pad material. Running at >60W input, and 28C ambient the hottest area on the driver (via IR thermal gun) was 50C. So the driver is very temperature stable with the thermal path I've provided. The poor LEDs though were getting sizzling hot...
I've set max output current to 1.3A and as implemented is dimmable via a trimpot (much like CCHIPO). I plan to provide a PWM input so it can work with d2Flex. I "may" also do a Flex version - but that will come later if I decide there's a market for it.
Efficiency is around 88 - 92% for the tests I performed.
Here's a picture of the prototype driver:
Some efficiency readings here:
It was amusing to set up a load. I had 7 series connected Luxeon3's, a K2, an old Lamina white array and Lamina Red array and some other Red array. All that to get around 50Volts worth of Vf to test with. I had to cover the LEDs to stop from getting blinded by all the light
cheers,
george.
The hyperboost can run from a nominal 6V to 50V input. Output voltage is up to 70V and open circuit protected (maxes out to 70V).
Board diameter is 1.4" with components only on the top side. Heatsink attachment is the same well proven scheme as for hipCC and hipFlex, a piece of thermal pad material. Running at >60W input, and 28C ambient the hottest area on the driver (via IR thermal gun) was 50C. So the driver is very temperature stable with the thermal path I've provided. The poor LEDs though were getting sizzling hot...
I've set max output current to 1.3A and as implemented is dimmable via a trimpot (much like CCHIPO). I plan to provide a PWM input so it can work with d2Flex. I "may" also do a Flex version - but that will come later if I decide there's a market for it.
Efficiency is around 88 - 92% for the tests I performed.
Here's a picture of the prototype driver:
Some efficiency readings here:
It was amusing to set up a load. I had 7 series connected Luxeon3's, a K2, an old Lamina white array and Lamina Red array and some other Red array. All that to get around 50Volts worth of Vf to test with. I had to cover the LEDs to stop from getting blinded by all the light
cheers,
george.