My Aurora C6 with LPC-826, Copper Heatsink and 1085 DIY Driver

[chefla]

I had a few parts laying around and decided to do a quick build using the following:
1. Aurora C6 host
2. LPC-826 diode (new)
3. copper heat sink
4. 1085 IC, resistor, ceramic cap
5. 2x CR123A batteries
6. glass lens

The host:
It is of decent quality. I ordered the bare host and not a complete flashlight. Threads well greased, nice surface finish, good form factor. It is nothing exciting but I like it.

The laser diode:
I was surprised to see that the diode looks exactly the same as the diode salvaged from an LPC-815 sled. I was under the impression that the emitter on the LPC-815 was shorter, but as a matter of fact, both are exactly the same length.:thinking:

The heat sink:
I got it from Moh. Good chunk of copper, very solid. A few minor burrs that I had to deal with prior to installing the diode, but nothing major. I like it a lot. Only the thread for the lens could be a little tighter. And I could not use the back end of an Aixiz module to set the diode. The back end was too large (or somebody made the pocket in the heat sink too small:spank so I had to file it down to make it fit inside the cooper heat sink.

The 1085:
I love this thing. I have build a few lasers with it and that piece is awesome. For only 50 cents it is easy to use, small enough to fit into many hosts and very stable.
The only thing you have to watch out for is the metal tab which is not connected to ground. This makes the heatsinking a tad more difficult. For this build I had to remove some material of the metal tab to ensure that it fits inside the host.
I used a small ceramic SMD capacitor soldered directly to the pins of the diode. I wanted to set it to 431mA but my local electronics guy had no 2.9 Ohm resistors so I used a 3.0 Ohm one resulting in 416mA driver current.

The batteries:
I used the gray UltraFire 880mAh CR123A protected ones from DX. So far no problems.

Glass lens:
Standard three element lens, with AR coating for red lasers.

Power Output:
You guess it! I still don't have an LPM. But this thing can burn white paper instantly. So I'd say that is enough power for me
EDIT:
I used my new LaserBee and measured 250mW with full batteries. A little dissapointing, I was hoping to get more out of it. I will check the driver, change the resistor and pump more current through it. I will update the post once I get it done.

Thanks for reading and here come the pictures:

The host and the heatsink.

The backside of the copper heat sink.

The 1085 and its new home.

1085 installed with current setting resistor.

On the bottom a piece of epoxy board with solder blob for battery (+) connection.

All wired up (notice the blue wire for minus connection).

Done!

Beamshot!

Excuse my handwriting

 

[Fiddy]

Good stuff mate!

 

[rhd]

I love your build style - it's totally my bag!

One thing to watch out for -

Almost an entire li-ion cell's worth of Voltage is going to be dissipated by the 1085. In my experience, you really need to find a way to get the metal tab of the 1085 heatsinked to something. You might be fine here since the current is so low (and thus less actual wattage/heat dissipated by the IC), but it's cutting it close. If you notice your diode "blinks" when you run it for 4 or 5 minutes straight, the heat on the IC is why.

Great build - I love DIY drivers, and especially in hosts that don't usually get the DIY driver treatment!

 

[Lotus_Darkrose]

Very nice build. That focus sure sticks out there though

 

[bobhaha]

You need a focus adapter.... but looks awesome mate! moh did a good job on that heatsink.

 

[chefla]

Thanks guys, I am flattered!
@rhd, thanks for the hint. Is the blinking the result of the thermal shutdown of the 1085 kicking in? So far I have not noticed it in any of my builds.
I did a lot of research on drivers and the 1085 seemed to be the best choice overall. Prior to installing it, I had it running at this current with no heatsinking for a few minutes and it only got warm to the touch. So I figure I am still on the safe side.

@Lotus_Darkrose and @bobhaha, there is a lathe sitting around waiting to spit out a few custom parts for me. I just need to find the time...

 

[LaZeRz]

Awesome build!

Make sure those parts dont short out in the pill

So the 1085 has a lower voltage dropout then the 1117?

 

[chefla]

So the 1085 has a lower voltage dropout then the 1117?

Where did I state that? But yes, if you look up their datasheets you will find that the 1085 has a lower dropout voltage at the same current than the 1117. But how does this matter? We are using them as current regulators, not for voltage. Therefore you will always see no less than 1.25 Vdrop.

 

[danefex]

Nice job! I was thinking of making a LOC from one of my C6s with moh's heatsink also, you just helped me make the decision rhd is right too, that 1085 will live a lot longer with some heatsinking :beer:

 

[chefla]

Thanks! In my opinion there is really no need for a better heatsinking at that current. It needs to dissipate around 2 Watt when the batteries are fully charged. As I said, I had run it for several minutes and it barely got warm. There is some learning involved with this hobby. So if the heatsinking results in a damaged 1085, then I have a lesson learned and wasted 50 cents. I am a risk-taker ... :beer:

 

[rhd]

Where did I state that? But yes, if you look up their datasheets you will find that the 1085 has a lower dropout voltage at the same current than the 1117. But how does this matter? We are using them as current regulators, not for voltage. Therefore you will always see no less than 1.25 Vdrop.

In fairness, that's kind of analogous to saying "Why do you need a lighter bicycle if every bicycle will always weigh at least 5 pounds?"
- the point is, some weigh a lot more.

It may be impossible to break the 1.25V barrier with a DDL style driver, but it's still worth getting as close to it as possible.

 

[chefla]

And, danefex, be advised that this is a very tight package there. Between the 1085 and the wiring coming from the laser diode are only a few tenths left. I recommend very flexible wires and try to keep the pins on the diode short. Good luck and let me know how it turned out.

 

[chefla]

In fairness, that's kind of analogous to saying "Why do you need a lighter bicycle if every bicycle will always weigh at least 5 pounds?"
- the point is, some weigh a lot more.

It may be impossible to break the 1.25V barrier with a DDL style driver, but it's still worth getting as close to it as possible.

Interesting analogy:thinking: But this reminds me of the first time I rode an ultra light racing bike. I was not able to drive it straight. It was so light it took me some time to get my steering inputs fine tuned.
But yes, in general I'd agree, the lower the voltage reference the better. BUT, we must not forget that there are limits. The more you reduce the voltage reference, the more difficult it gets for the IC to maintain it. You would probably see more transient spikes and other nasty things which we better keep away from our precious laser diodes. I believe there is a good reason why most companies chose to not lower their reference voltage any more. I think that may be a similar problem as the one I mentioned above. If the bike weighted nothing, there would be no gyroscopic forces to stabilize me and I'd fall on my butt. Ok, that my be highly simplified. But with all the part tolerances and temperature effects a larger Vref allows for better regulation. Maybe I am completely wrong here....

 

[danefex]

And, danefex, be advised that this is a very tight package there. Between the 1085 and the wiring coming from the laser diode are only a few tenths left. I recommend very flexible wires and try to keep the pins on the diode short. Good luck and let me know how it turned out.

Yes I completely agree, I was thinking of mounting my driver inside the very limited space in the pill on these units :beer:

 

[chefla]

Updated the post with power measurement.

 

[JBTexas]

<- Newbie here, and I lack experience. I'm suprised that it burns *white* paper. I've had a ~300mw 660nm (with a 650-G-1 Lens, so 300mw + ) for a while now, and I THINK it is based on the same diode. I was *assuming* that it wouldn't burn white paper. I will have to try it.