Hi and Lo mode lm317 driver proposal

Hello you guys,
anyone ever consider making a two (or more) mode, high and low,
variation of the popular lm317 DIY driver?
I thought it would be nice to have on a highpowered pointer, such as
a 445nm diode or an 8-12x Bluray.
You know, one mode to point around with (relatively) safely, and one
mode to BURN!

Say, 300ma & 1200 ma.
or 50ma & 300ma

As a switch, I figured a micro-sized slide Single Pole Double Throw switch on the outside of the host.


What would be REALLY slick is a reed (magnetic) switch, which would be
invisible from the outside, and the pointer would be in "safe"-mode until you
cling your magnet to the pointer!

Eh, Eh? whaddayall think about that!?
If only I could source a reed changeover rated for 1.5A...
EDIT:
hey this one might do the trick...
reed spdt

Yeah, that would work. Best to switch the current when the laser is off though. For the split second when NO resistor is selected, the adjust terminal still sees the voltage of the capacitor, which may cause a spike when the switch makes contact again.

Yes it's been thought of and here is the finished product a 445 powered by two 18650 cells and on high it runs
at 1050mW a very nice burner at that speed and on low it runs at 43mW a very nice pointer at that speed

Wow, that was fast!
Looking quite smooth there, flaming...!
Figured as well someone must've been there before.

Might I ask what kinda switch you used an where you put it?
On a host this big, one might almost get away with a cool-looking toggle-switch.
Did I figure the circuit the same way you built it?
What current did you set to get ~50mw off the 455?
Got a link to the build of this thing?
Can you switch modes when "hot"?
From lo to hi might be OK, but other way round I guess fries the switch
after some time.

Im a really intrigued by the sneakiness of those reed switches though. I guess i'll
have to give them a try. If only they weren't so expensive.

Instead of a switch that moves between paths, use a switch to activate a shunt resistance path.

So that way, you have a path always there, and the parallel path just lowers your resistance. For instance, a 10ohm resistor always in place giving you a 125mA current. Then you close a switch, activating a parallel resistance, say another 10ohm resistor. This gives you 2 resistors in parallel equaling to a 5ohm resistor and 250mA of current. In your diagram, instead of choosing resistor A OR resistor B, you're choosing resistor A for low power OR resistors A & B in parallel for high power.

And you can switch if while running. The 10ohm resistor is always there, the resistance doesn't go through zero or infinity during the switching process. Plus now you just need an on/off switch for the parallel/shunt path, which can be more simple.

Thanks, that's just the kind of suggestion I needed.:beer:
Simplifies things a bit, more robustness, me like.
Makes for a simpler, cheaper switch as well, no more need for a SPDT.
I just whipped up a spreadsheet to comfortably juggle some numbers and came out with:
R1 (permanent, lo) = 15ohm -> Ilo = 83ma
R2 (switched in parallel, hi) = 1.2Ohm
makes Rt = 1.111_Ohm -> Ihi = 1125ma

That should balance it about the way flaming has it on his...

or 18 and 4.7 which makes ~70ma and ~330ma for a BR

Hmm, I like this idea better all the time. I guess by now I am comitted to implement it
in actual circuitry.

BTW flaming, does your 2 mode thinggy perform as expected? Are you happy with it?

EDIT:
dang, that R2 for hi current would have to a beefy one though (5W enough maybe?)

^Good except you need more current on the low setting if you're using one of the common 445nm diodes like the a140. Those don't start lasing until around 200mA. They still turn on at lower power, but they need that much current to turn on. So you need something like 250mA as a low setting, or thereabouts, but I recommend playing around with it. Like using a pot to find the current you like, then get resistors to go in the final build.

OK, dual mode drivers are SOOOOO 2010. *YAWN*
Behold the 16 (sixteen) mode lm317 driver!

Ingredients:
5 resistors: 1,2,5,10 Ohm + 0.2Ohm
1 4Xdipswitch.

My GF said it looks like "barbies circuit breaker panel", before I soldered up the resistors.:crackup:


This is wired up as in the attachment with the 0.2ohm in series and the other ones in parallel
one to each dipswitch. This whole thing now is the Rsense for that lm317 driver.
It lives in my tinkerbox.

I've got all kinds of currents:
120ma- no lase
240ma- WOW single mode from a 445!
354ma- more bars join in and lase
...
...
all the way to a theoretical 1650ma.

In practice though it tops out at 1220ma draw from the battery when all the Rs are switched in parallel. :thinking:
Gotta look into that.

Looks like with that config the switch might have to be able to handle an amp or more. just be careful.

If 1 mode is okay, 2 modes is better and and 16 modes is better yet, then ∞ modes must be AWESOME

Use a <25Ω pot instead. It is more simple, and many more different currents are available.

Cyparagon said:

Use a <25Ω pot instead. It is more simple, and many more different currents are available.

Click to expand...

You mean like this one, which I incidently just purchased?:crackup:
I had to do something until that one arrives.

Trouble with those linear pots is, all the action happens at the last 10~15 degree.