Drivers for 445

[vegasf6]

It looks like it should work fine to me. It is just a 7805 voltage regulator being tricked into a current regulator, no different than many here are doing with the LM317. You can see a basic schematic in the data sheet on page 21 to do the same thing.

R6 and R7 control the current. If you want a variable circuit you could use a 1 ohm resistor for R6 and replace R7 with a pot, maybe a 5W 100 ohm linear taper for instance.

Are you sure on the cap values? The schematic looks like 470uF on the input. I can't really tell what is on the output. Looks like C5 10uF, and then the tiny caps, rated in nano and pico to filter the ground path out for the ICs. Seems like this would be sufficient as a laser diode should be a very constant load? If not, I don't suppose it would hurt to parallel a 47uF or even 100uF cap (25V) but I don't think it would be needed.

Finally, what will you use to power it? This is a linear regulator so the greater difference between Vin and Vout the more heat generated. Even though the literature claims no heatsink needed, it wouldn't hurt to add one to T1 and the 7805. Perhaps if you are using a project box with a metal lid you could bond them to that. I would probably want a regulated DC supply despite the circuit including a rectifier.

 

[ZRTMWA]

I just set a V5 FlexDrive at 1100mA's (per request) for a 'Hot' option 18650 kit and it is stable at that current even though it does not recommend this in the documentation...

As far as whether the driver will start to have problems (with long 'on' time), I have not tested. But it works well with 445. I tested it on a diode. You can't set the current this high with a blu-ray test load. At that current, the driver won't work unless you are using a red test load... (but it runs the 445 diode just fine - I measured 1200mW's with a 405-G-1)

One thing that will make a huge difference is the quality of the battery. Not all brands will handle that kind of current draw. Probably between 1.5 and 2 Amps... (guessing)

I would recommend spending the money on an AW battery:
AW 18650 Protected 2600 mAh Rechargeable Lithium Battery - New Version

Those batteries can handle a crapload of current draw! I ordered these from DX: DealExtreme: $8.01 TrustFire Protected 18650 Lithium Battery (2500mAh 2-Pack Blue)

The reviews on that page state that they can handle anywhere from 1.6A to 2.8A of current draw. I bet the wide range is due to poor quality control. I'll test em out and if the laser flickers I'll just order quality cells. If those don't work for some reason, I'll just lower the current.

 

[Exerd]

Due to the amount of questions about batteries, I figured this test of a few brands may clear some confusion.

As you can see, the Tenergy is actually the best cell out there at near 1A current, defeating even the AW brand. At the highest of currents around 3A, the Ultrafire 3000 is rated slightly higher than the rest.

Source: http://www.candlepowerforums.com/vb/showthread.php?t=257543

 

[nospin]

My flexdrive needed to be sinked at full output to keep it from pulsing
now it seems to run fine

 

[LazyBeam]

Due to the amount of questions about batteries, I figured this test of a few brands may clear some confusion.

As you can see, the Tenergy is actually the best cell out there at near 1A current, defeating even the AW brand. At the highest of currents around 3A, the Ultrafire 3000 is rated slightly higher than the rest....

Keep in mind when you drive a ~4.5v diode @ 1A from a battery that outputs 3.7v you are drawing alot more than 1A, especially when you consider the inefficiencies. Assume a Flexdrive @ 87%... to drive 1A you would need to draw at least ~1400mA. (1000mA x 4.5v / 3.7v / 0.87 = 1398mA)

You'd have to assume performance around 2A draw if you want to output 1000+ mA.
IMO, based on that graph, the AW 2600 Black or Ultrafire BRC3000 would be equally good choices along with the Tenergy.
They seem to have better 3A performance. At 2A they'd all be so close you'd never notice a difference in performance.

Thanks for the chart... it should help alot of choose the best batteries!

 

[oic0]

So Flexdrive > Microboost for these diodes?
Has anyone seen how high the flexdrive can stably go with one?

My flexdrive needed to be sinked at full output to keep it from pulsing
now it seems to run fine

Your flexdrive is handling a full 1.5 amps? What did you use as a heatsink? a motherboard mosfett sink? Also what battery?

 

[Xplorer877]

The problem is the Boostdrive's efficiency. Depending on what your pot is set to you can have between 76-86% efficiency. That is pretty good but not good enough with super high currents now. If the diode is eating up 4.4W the the driver has to dissipate about .7 watts at best. That's alot for those little components.

Originally I was planning on driving my diode at 650mA but ever since I've seen that these diode can be pushed safely up to an amp I think I'll raise it up a bit.

But now that I'm seeing this I'm not so sure anymore! haha! Anyways, when I bought my driver from Dr. Lava I had him preset the current to 650mA. I did that because I'm not super comfortable with that precision soldering. When I got the driver from Lava I noticed that the solder jumps didn't match any of the permutations on the instructions. I PMed Dr. Lava about it and he said it's actually an even higher range. He did it that way because it's easier to solder and will yield slightly higher efficiency. The higher range is about 560-1100mA.

left is 560-1100mA, right is 540-1000mA

So maybe the added bit of efficiency might alleviate some of the problem. I think I'll turn mine up to 800mA but I'm not sure.

 

[LASER66]

Considering the earlier posts of using an LM317, when the voltage is available for "buck' mode, as I know a lot of attention here is on boost drivers...

I sell a small kit uisang the LM317T. However, for some sales, buyers were asking about driving multiple watt 808nm diodes. We sold them our kits using the LM350T, same TO-220 package, but with a heat sink, the unit will drive 3 amps! Same 1.2 volt reference, etc. Maybe this is an alternative to the LM317 discussed in earlier posts.

Great thread
Will

 

[jmgallego]

Considering the earlier posts of using an LM317, when the voltage is available for "buck' mode, as I know a lot of attention here is on boost drivers...

I sell a small kit uisang the LM317T. However, for some sales, buyers were asking about driving multiple watt 808nm diodes. We sold them our kits using the LM350T, same TO-220 package, but with a heat sink, the unit will drive 3 amps! Same 1.2 volt reference, etc. Maybe this is an alternative to the LM317 discussed in earlier posts.

Great thread
Will

Will,

Are you saying that LM350T would this work on the 445nm? Do you have them available?

Thanks

 

[ReNNo]

Why not...
LM350 is the same as LM317... only difference is that LM350 can give up to 3A.

But it requires very good heatsink, specially if we use 2xLi-ion to power it on.

 

[BShanahan14rulz]

AW's LiMn is better for high current draws. LiMn in general is better at that than LiCo. If it's capacity you're after, though, stay with your LiCos. Lasers generally don't seem to require the high current draws that flashlights do. The AW LiMn is capable of 10C, and at 1600mAh, that's 16 amps safe continuous output from a single cell. Pretty amazing, and for the best of the best lasers possibly necessary. For most people, though, a good name-brand high capacity cell would work better.

Also, any word about the 7135's suitability for driving 445nm diodes? I found a gem buried in a post about red LOCs. You can get around the low voltage requirements of the 7135 by dropping extra voltage with diodes on the Vin of the 7135. Now we just need to figure out if the output is stable enough. Most 7135 boards don't come with the filter cap (none that I've seen, at least), but I'd be curious to see how stable the output is when the suggested cap is in place.

 

[jmgallego]

Great! Thanks ReNNo.

In my case, I was looking into using the LM350 for a lab bench version and really pushing the diode. As you said, the only thing I would need to worry is the heat generated not only from the diode, but also from the driver. I seen a couple already good ones using large heat sinks I will take a good look at those and start working on it this weekend.

THANKS!!

 

[vegasf6]

Also, any word about the 7135's suitability for driving 445nm diodes? I found a gem buried in a post about red LOCs. You can get around the low voltage requirements of the 7135 by dropping extra voltage with diodes on the Vin of the 7135. Now we just need to figure out if the output is stable enough. Most 7135 boards don't come with the filter cap (none that I've seen, at least), but I'd be curious to see how stable the output is when the suggested cap is in place.

There are pages and pages on Candlepower forums about using the 7135 with multiple leds to drop the voltage (as opposed to zener diodes or other silicon diodes) but you are still burning battery power elsewhere. I believe the output voltage will be as stable as your input voltage pretty much, with the exception of any noise picked up by long leads, which aren't being used here. The suggested cap on the output, .1 to 1uF is a bypass cap to eliminate signal noise, I don't believe not using it will allow any significant amounts of current that would damage laser diode. At least not with batteries as the power source.

But, the voltage requirements of the LD and the max input voltage of the 7135 make them a poor match for lithium battery. Best bet would be 4xnimh. That will net the maximum efficiency.

 

[vegasf6]

Great! Thanks ReNNo.

In my case, I was looking into using the LM350 for a lab bench version and really pushing the diode. As you said, the only thing I would need to worry is the heat generated not only from the diode, but also from the driver. I seen a couple already good ones using large heat sinks I will take a good look at those and start working on it this weekend.

THANKS!!

Why not look at the LM338 as well? A LDO regulator with a lower dropout would be nice. I am looking at the LM2941 myself, but it is max 1A.

 

[Kevlar]

Have you seen page 4 of this thread? I thought the shark buck looked pretty promising, it even comes with a nice heatsink. What are other peoples opinions on the shark buck. I thought about using this instead of the AMC7135

 

[oic0]

Anyone else try a flexdrive at 1.5amps on this diode?