3V 3.5A Driver Needed

I'm looking into building a 3watt 808nm laser and am researching some ideas for a driver. The LD has an operating voltage of 2.2V and current of 3.5A

What I have seen (that is available and affordable) is the LM350T (adjustable volt from 1.2 - 33V up to 3A with 2.5 Vdo).

With the Vdo of the LM350 and what I need for the diode, I will be needing at least 5V to barely run. Does anyone suggest using a 3.7V battery and boosting (how do I boost? Do I need transistor? inductor?), or use 7.2V and regulate down? Using 3-3.5A would drain the batteries rather quickly if I'm not mistaken... unless I use 4 batteries(2 series of 2 parallel)?

I'm a bit new to this, so I apologize for the ignorance!

Thanks everybody!

If you are going to make a linear driver, there are several options

1. paralleling a few LM317, with a balancing resistor on each's output.
2. Just use a higher AMP capable IC like you suggested, or an 1085, or ....

Boosting the voltage to burn it off as heat on the next step sounds very silly.

You might get away with the 4V of a single LiIon, if you choose something low dropout
like the 1085 over the LM317.
Just check if you get the 3.5A from a single LiIon, if not, add another LiIon in series.

Even the sturdy 18650 won't last long under 3.5A draw....
So put another one or two in parallel for 2x or 3x times the runtime.
So yeah you see you weren't far off with your 2x2 battery idea.

Or just plug it into the wall, or is this going to be a pointer?

You trying to put this thing in a laser pointer or is it okay for a wall output?

If wall output, my personal fav is the LM338. 5-7A regulator.

LM338 - 5 Amp Adjustable Regulator

A good heatsink for it(same kinda as the laserbee and radiant alpha heatsinks) is this

Digi-Key - 345-1053-ND (Manufacturer - 641K)

That is assuming you buy the TO-3 package LM338.

Its an expensive IC though.

LM338K STMicroelectronics Linear Regulators - Standard

anselm said:

If you are going to make a linear driver, there are several options

1. paralleling a few LM317, with a balancing resistor on each's output.
2. Just use a higher AMP capable IC like you suggested, or an 1085, or ....

Boosting the voltage to burn it off as heat on the next step sounds very silly.

You might get away with the 4V of a single LiIon, if you choose something low dropout
like the 1085 over the LM317.
Just check if you get the 3.5A from a single LiIon, if not, add another LiIon in series.

Even the sturdy 18650 won't last long under 3.5A draw....
So put another one or two in parallel for 2x or 3x times the runtime.
So yeah you see you weren't far off with your 2x2 battery idea.

Or just plug it into the wall, or is this going to be a pointer?

Click to expand...

I admit, I played with the thought of a 3watt pointer... and I may still do it if I can build the driver small enough. In which case I MAY actually hook up a 30mm fan to use the excess voltage if I use the 18650s in series (and parallel unless I"m good with 20-30 minute usage!)

The reason I am looking at the LM350 is because it only costs $.50 as opposed to $10+. I can swing another battery to save that much $. I think the host would look better longer anyways... I'm thinking of a maglite... but I haven't tried fitting a 2x2 or even 2x3 config of batts in yet.

I did some testing last night straight from the batteries but could reach a max of only 2.5A. I may have the test load set up wrong though. Should I still be measuring across a 1ohm? I've seen use of .5 ohm on 445nm test loads? Not sure what I'd need for 808nm. Would I need to have 3 diodes in series to imitate the voltage drop only? Or do I need more to imitate the amperage demand? I am using N5404s
1N5404 Diode Rectifier 3A 400V DO-201

If all else fails with the batteries, I do have a 14V 25A PS I can use for a bench unit. If I set it up as a bench unit, I have a CPU heatsink+fan I would use for heatsinking.

Thoughts on the test load?

Thanks guys