parallel drivers

[jayrob]

I don't have much experience with power supplies so I can't give you any advice on which one to get...

But hopefully someone will chime in with some recommendations!

I would say that if you had one that you could adjust voltage output, as well as current separately. That would be what you need...

 

[Morgan]

A 37 second runtime is not much less than the first time I used a flexdrive maxed out at 1.5A without a heatsink. Unfortunately it does not seem as easy to mount a heatsink on these drivers.

I'm not convinced the drivers need a seperate heatsink in all cases. The 18650 host I built has no seperate sink for the Flexdrive and is running at maximum outputting at least 1,350mW with a 3 element type lens, (+15% to compare with a 405-G-1 and that's 1,550mW or so). The driver is bonded straight to the pill with thermal plaster and uses the host as a sink. It runs for 6mins and drops the somewhat expected ~30% but is too hot to handle after that. The driver doesn't drop out though.

This is the same host as Jayrob provides but a different design heatsink for the Aixiz, (I'll send you some photos Jay). It works really nicely.

I have another host on route and will try a parallel Microboost setup with a higher capacity 18650 battery and see what happens. I've decided I'm going to try and murder a diode! :eg:

M

 

[jayrob]

I'm not convinced the drivers need a seperate heatsink in all cases. The 18650 host I built has no seperate sink for the Flexdrive and is running at maximum outputting at least 1,350mW with a 3 element type lens, (+15% to compare with a 405-G-1 and that's 1,550mW or so). The driver is bonded straight to the pill with thermal plaster and uses the host as a sink. It runs for 6mins and drops the somewhat expected ~30% but is too hot to handle after that. The driver doesn't drop out though.

This is the same host as Jayrob provides but a different design heatsink for the Aixiz, (I'll send you some photos Jay). It works really nicely.

I have another host on route and will try a parallel Microboost setup with a higher capacity 18650 battery and see what happens. I've decided I'm going to try and murder a diode!

M

Yes if the driver was connected directly to the pill with thermal adhesive (5 pin chip), it would be a huge heatsink. (host) But I did not do it that way simply because I did not see a good way to do it... (not on the Stainless Steel 18650 kit)

Maybe if the top of the pill was not drilled out, and the driver was glued into the top of the pill on the inside??

But with the black 18650 kit you could glue it on top of the pill. Did you glue it on top? Ahh... I'll bet that's what you did huh? The black 18650 kit has room to mount the driver on top of the pill if you wanted to. So all you would need to do, is glue the driver on top of the pill...

Better yet, I could machine the top of the pill flat all the way across... (because there is a sunken area where the LED emitter was mounted...

 

[Morgan]

Yes if the driver was connected directly to the pill with thermal adhesive (5 pin chip), it would be a huge heatsink. (host) But I did not do it that way simply because I did not see a good way to do it... (not on the Stainless Steel 18650 kit)

Maybe if the top of the pill was not drilled out, and the driver was glued into the top of the pill on the inside??

But with the black 18650 kit you could glue it on top of the pill. Did you glue it on top? Ahh... I'll bet that's what you did huh? The black 18650 kit has room to mount the driver on top of the pill if you wanted to. So all you would need to do, is glue the driver on top of the pill...

Better yet, I could machine the top of the pill flat all the way across... (because there is a sunken area where the LED emitter was mounted...

Ker-ching! ^ This! The module heatsink bonds to the pill too and is clamped against it.

I should've clarified it is the black host, not the Stainless one.

M

 

[DTR]

Alright I set up 3 18650's and a 18500 in parallel for a little overkill. 40 second run. Diode was warm/hot and the drivers were hot/burning to the touch. Also took the lens out so you can see the smoked diode window on this diode.

 

[jayrob]

Yeah that dot looked weird...

Hopefully it was the lens and not the diode...

 

[DTR]

Yeah that dot looked weird...

Hopefully it was the lens and not the diode...

That is the raw output of the diode. It was burnt a long time ago and that is why I use it for testing instead of a good one. Something was in the module that burnt onto the lens. I tried to clean it up but this is as good as it gets and it does function fine for testing. I did test it on the LPM and it came in at 1444mW but I think a lot of light is not getting through.

EDIT

Here is a LPM test. 1444mW even with this messed up diode.