Maxflex measurements

[Erasmus]

I finally got to play with the Maxflex 2.0 driver, and I made some measurements to share :
(click for bigger size)


I also made a custom heatsink for the Maxflex. It is a piece of copper pipe to which the Maxflex is soldered. The whole thing fits right into the barrel of a D-size Maglite so the heat can be further dissipated to the Maglite's body :
(click for bigger size)


Cheers
Raf

 

[chimo]

Raf, thanks for sharing this info!

Paul

 

[wquiles]

Raf,

Have you emailed George? He should be able to help

Will

 

[yellow]

Troubleshooting anyone?

I think I've read somewhere that the black IC shown in the pic (the 8 legged one) should also be sinked somehow

... sitting here and thinking of how to include that to the actual poject :thinking:

 

[Erasmus]

One of the wires to the LEDs was not connected anymore... I did some further testing and assembling and so far everything goes fine It's the first time I used epoxy to fill up the heatsink and cover all wires, the result is nice but damn this epoxy stuff is something dirty!

 

[Erasmus]

I think I've read somewhere that the black IC shown in the pic (the 8 legged one) should also be sinked somehow

... sitting here and thinking of how to include that to the actual poject :thinking:

That 8 legged IC is a controller for the temperature sensing, thus this doesn't need to be cooled. The part you're talking about is IC 'U1' which is at the other side of the board, under the place where I soldered the heatsink to the driver. If the driver is dissipating more than 1W, it is recommended to cool the U1 from the top as well. I contacted George @ Taskled and he told me it is safe to run my setup as it is heatsinked now, because the maximum power dissipation in the driver is 0.6W.

Cheers!

 

[HarryN]

Thanks for posting the table - nice work.

Interestingly, while the driver and leds are definitely getting pretty good, if you add in the AC - DC conversion + driver efficiency + LED efficiency = still not very good, something less than 50%.

I guess there is still some work to do.

 

[Erasmus]

Thanks for posting the table - nice work.

Interestingly, while the driver and leds are definitely getting pretty good, if you add in the AC - DC conversion + driver efficiency + LED efficiency = still not very good, something less than 50%.

I guess there is still some work to do.

There's some work indeed, but the setup with the power adapter was only for testing purposes during the construction of the lamp. Now it's finished and running from 8 AA-size Eneloop NiMH's, with which I made a few extra measurements. I also connected the IC 'U1' to the heatsink with a piece of solder wick, thermal epoxied to the IC and the driver's heatsink. Some pics :

Measurements with Eneloops :

Solder wick epoxied to U1 :

Red arrow points to the solder wick which is epoxied to the heatsink :

Cheers
Raf