Announcing - Fatman lives!

[georges80]

In a fit of boredom I decided to build a few prototypes of a driver that will be well suited for multiple Lux 3's or Lux5's. I was running some other prototypes & knocked together the Fatman schematic and layed it out onto a portion of the PCB panel.

Here is the first Fatman prototype just before firing it up. And, since it worked (just minutes ago) I figured I'd announce a few specs and see what the interest is in running a production board.

Some basic specs:

[*]0.8" diameter
[*]boost converter
[*]nominal 600KHz switching frequency
[*]center rear nubbin (Vin) rest is ground plane to the edge
[*] adjustable via onboard trimmer (for configuring)
[*] Vin (min) = 2.7V, Vin (max) 12V
[*] Vout (max) = 16V
[*] Fully current regulated and fully adjustable either on board or via external pot (in parallel with trimmer)

I did some very preliminary testing and measured over 90% running from 3.30V - 5.0V while driving 720mA into a 5W luxeon with a measured Vf of 6.46V. The current output was just where the trimmer pot happened to be set to.

This switcher IC has a 2.55Amp (minimum) switch current limit so is well suited to higher current outputs. Fatman would be well suited to driving 3 Luxeon 3's (series connected) at 1A from say 3 123 cells for example.

Remember, the driver is fully dimmable and is configured to put out around 30mA at the fully dim setting.

Anyhow, much more to come as I get time to take more measurements and in response to any question you guys come up with. I also need to characterize the driver over voltage range and loads (lots of stuff still to do).

Right now things look good, but until I do the rest of the measurements and tests...

george.

 

[wasabe64]

I'll be looking forward to the production run! /ubbthreads/images/graemlins/thumbsup.gif

 

[HarryN]

Hi George

Looks interesting, but what got me going initially, was the idea that maybe you had a driver board for multiple 3 watt Luxeons in parallel, in a buck type format. Sort of like your other boards but in a 3 x or 4 x configuration.

By the way, the other boards I bought from you work great.

 

[georges80]

[ QUOTE ]
HarryN said:
Hi George

Looks interesting, but what got me going initially, was the idea that maybe you had a driver board for multiple 3 watt Luxeons in parallel, in a buck type format. Sort of like your other boards but in a 3 x or 4 x configuration.

By the way, the other boards I bought from you work great.

[/ QUOTE ]

Parallel Luxeons have an issue in trying to match the Vf. By putting the Luxeons in series, their individual Vf's don't change the current going through each of them - they obviously all get the exact same current.

Glad my other driver boards are performing well for you.

george.

 

[moraino]

Is it dimmable thru monentary switch like uFlex?

 

[georges80]

[ QUOTE ]
moraino said:
Is it dimmable thru monentary switch like uFlex?

[/ QUOTE ]

Nope, Fatman is fully analog, so the dimming is via an external potentiometer. Probably a good choice would be one with a log taper to approximate the eye's response to light intensity.

The onboard trimmer lets you set the current limit (1A max - or more if the input/output voltage differential isn't too great). The onboard trimmer can also be used to 'calibrate' the external pot. Both are in parallel. Most pots aren't that accurate - something like +/-20% is common, so the trimmer allows you to set your maximum current more accurately. Minimum current is set by an onboard resistor to around 30mA (which is reasonably dim for a Luxeon).

The external pot would be in the 20k ohm range. Of course you can also use Fatman as a fixed output driver for those multiple Luxeon lights some CPFers seem to be wanting to see in fully regulated form. The onboard trimmer would then allow you to set the precise current you want Fatman to provide - whether it be 300mA, 400mA, 823mA /ubbthreads/images/graemlins/grin.gif or 1A etc.

george.

 

[unnerv]

Very cool. This I think would make an excellent board for all of the Mag mods out there, esp since it can handle 1A driving a 5 watt. With a good heatsink it sounds like we could easily have a fully regulated SNII type light.

 

[georges80]

Ok, a few more tests. Will take pics later. Anyhow, got a pot hooked up and I can dim from 30mA to full drive at 1A (it can go higher...) into a 5W Luxeon with a 6.6V Vf. Feeding Fatman with around 5.1V (just what the supply was set to) I measured about 95% conversion efficiency!!

Based on these preliminary numbers Fatman will happily drive HIGH powered luxeons without itself melting down.

I used an antilog taper pot (since Fatman drives less current when the adjustment resistor is lower in value). The antilog taper gives a very nice light output change vs pot position - seems very 'linear' to the human eye.

george.

 

[jtice]

So glad to hear this all worked out for you George! /ubbthreads/images/graemlins/bowdown.gif /ubbthreads/images/graemlins/buttrock.gif

These will be VERY nice, and make the world of modding alot more fun and creative.

Cant wait till they are for sale.

Thanks for the hard work

-John

 

[Flatscan]

Awesome. I've been looking around for 3x3W regulated solutions, and this looks perfect.

Is it possible to adjust brightness by using an external circuit to swap between resistors instead of a pot?

 

[georges80]

[ QUOTE ]
Flatscan said:
Awesome. I've been looking around for 3x3W regulated solutions, and this looks perfect.

Is it possible to adjust brightness by using an external circuit to swap between resistors instead of a pot?

[/ QUOTE ]

Sure - the pot just gives infinite control. Nothing stops you having a switch with a few positions that switches alternate resistors in circuit. Basically 2 wires from Fatman, one is ground, the other the 'adjust' pin. You put your resistor(s) across those 2 wires and you change the current level.

Just took some pics. First one shows Fatman pumping 30mA into a 5W Luxeon and the second with the pot at the other end pushing 1A into the 5W Luxeon. Exposure lock on the camera.

30mA

1A

The head that contains the luxeon has a piece of LSD (light diffusing material) hence the soft light.

george.

 

[Deleted member 684]

I've been looking for a 5 x 3 watt solution and this looks like it can do it.. now I need to find LS with close vFs....

Paypal on the way pal! /ubbthreads/images/graemlins/grin.gif

 

[georges80]

[ QUOTE ]
Nerd said:
I've been looking for a 5 x 3 watt solution and this looks like it can do it.. now I need to find LS with close vFs....

Paypal on the way pal! /ubbthreads/images/graemlins/grin.gif

[/ QUOTE ]

The Vf's don't need to be matched (as long as the light output is similar) - Fatman will drive 3 Lux3's in a SERIES connection. It can output up to 16V. Even 4 Lux3's if they have a Vf <4V at full drive (which should be easy to find).

Sounds like there's some interest out there - so I'll start a production board run in the next few days after a bit more testing (so far Fatman works perfectly). I'll need to order parts too - ouch$ /ubbthreads/images/graemlins/grin.gif That means I should have some production Fatmen in 2wks. I don't think I can post the price (target) in this thread??? CPF rules?? /ubbthreads/images/graemlins/thinking.gif

george.

 

[PhotonFanatic]

Well, can you PM me with a guesstimate on the price?

Looks like a great driver for a multiple LED headlamp, assuming we can heatsink it enough. /ubbthreads/images/graemlins/grin.gif

 

[georges80]

[ QUOTE ]
oeo2oo said:
Well, can you PM me with a guesstimate on the price?

Looks like a great driver for a multiple LED headlamp, assuming we can heatsink it enough. /ubbthreads/images/graemlins/grin.gif

[/ QUOTE ]

PM sent. The driver has the grunt. As you say - can you heatsink the Luxeons enough... Though, you can always run the dimmer - so a blast of light for when you need it and then dial back the current for the rest of the time.

george.

 

[McGizmo]

Very cool George! /ubbthreads/images/graemlins/thumbsup.gif /ubbthreads/images/graemlins/grin.gif

 

[hotbeam]

Excellent!! /ubbthreads/images/graemlins/grin.gif

 

[Ocelot]

I have a question ever since seeing all of the (small) designs which use extremely small inductors.

I wouldn't think that you could get such high efficiencies and large currents with such small coils. Smaller coils mean higher series resistance and therefore less efficiency. In addition, you are more likely to saturate them and cause efficiency to go down again.

What am I missing?

Scott

 

[georges80]

[ QUOTE ]
Ocelot said:
I have a question ever since seeing all of the (small) designs which use extremely small inductors.

I wouldn't think that you could get such high efficiencies and large currents with such small coils. Smaller coils mean higher series resistance and therefore less efficiency. In addition, you are more likely to saturate them and cause efficiency to go down again.

What am I missing?

Scott

[/ QUOTE ]

A lot /ubbthreads/images/graemlins/grin.gif

These switchers run at 600KHz or higher and thus the inductor value is only 4.7uH. The cores are also designed for high frequency and high currents. The smaller the inductance the less wire - which means less resistance. For the particular inductor on the Fatman proto the series resistance is only 0.05 ohms and the saturation current is 2.6Amps.

The advantage of these modern high frequency switchers is low inductance requirements (thus less resistance, thus more efficiency) and smaller capacitance requirements, thus small low ESR ceramics can be used which leads to higher efficiency.

george.

 

[dat2zip]

Nice /ubbthreads/images/graemlins/thumbsup.gif