Welcome to Laser Pointer Forums - discuss green laser pointers, blue laser pointers, and all types of lasers

LPF Donation via Stripe | LPF Donation - Other Methods

Links below open in new window

ArcticMyst Security by Avery

1500mA 445nm diode... use an LM317t or LM338t

Joined
Jun 12, 2010
Messages
462
Points
28
I'm going to be building a labby style laser with a pot for variable output power.
Active cooling, switchs, dials, outlet power... the works :D
I'll be supplying 12V DC from a wall outlet or PC power supply and would like to deliver 10mA-1500mA.

1500mA is right on the upper limit of the lm317 and stepping down from 12V it will get very hot.
Am I better off going with the LM338 and a heatsink? Is there any downside to the LM338?
Will I suffer poor performance at lower currents from the LM338?


Reading around, I gather that a 125ohm pot with 0.8ohm resistor in series should net me 10-1500 mA for an LM317t.
Would I use the same values for a LM338t or is there a different equation for this part?
I also plan on using reverse diode protection, and a capacitor filtering (47uF)across the LD outputs.
 





ped

0
Joined
Nov 25, 2008
Messages
4,889
Points
113
LM338k should be fine if the min voltage is still 1.2v?

Or use a series pass transistor like a 2N3055 or TO220 equiv?
 
Joined
Dec 23, 2007
Messages
2,494
Points
0
The resistor values are the same for the 317 and 338.
Here is a handy calculator, you want the current regulation one at the bottom

LM317 / LM338 / LM350 Voltage and Current Regulator Calculators

A very large heatsink would work fine with a 317, look inside broken electronics for some form fitted heatsinks.

Edit: You should also use a pot that can handle the wattage the calculator gives you.
 
Joined
Jun 12, 2010
Messages
462
Points
28
The resistor values are the same for the 317 and 338.
Here is a handy calculator, you want the current regulation one at the bottom

LM317 / LM338 / LM350 Voltage and Current Regulator Calculators

A very large heatsink would work fine with a 317, look inside broken electronics for some form fitted heatsinks.

Edit: You should also use a pot that can handle the wattage the calculator gives you.

Thank you for this information!
I think I'll go with the LM338t then. More safety factor on the max current.

A 5ohm 5W single turn potiometer in series with a 0.8 ohm 2W resistor should give me decent control from 216mA - 1563mA... Perfect for a 445 diode. :D Too bad I can't find any logarithmic pots in the <100ohm region. That'd give me even better control.

Potentiometer: Digi-Key - CT2150-ND (Manufacturer - 026TB32R5AOB1A1)
1.6 Resistor: Digi-Key - P1.6W-2BK-ND (Manufacturer - ERX-2SJ1R6)
LM338: Digi-Key - LM338T-ND (Manufacturer - LM338T/NOPB)
LD Bridge Capacitor: Digi-Key - 493-3275-ND (Manufacturer - UTT1E470MDD)
10k LD Bridge Resistor: Digi-Key - 10KYCT-ND (Manufacturer - FMP200FRF52-10K)
Reverse polarity Diode Protection: Digi-Key - 1N4001FSCT-ND (Manufacturer - 1N4001)
5.5mm Jacks: Female / Male
12v DC Outlet Power: http://www.mini-box.com/24w-12v-2A-AC-DC-Power-Adapter
Power & active cooling switchs: DPST / SPST
 
Last edited:
Joined
Dec 23, 2007
Messages
2,494
Points
0
Thank you for this information!
I think I'll go with the LM338t then. More safety factor on the max current.

A 5ohm 5W single turn potiometer in series with a 0.8 ohm 2W resistor should give me decent control from 216mA - 1563mA... Perfect for a 445 diode. :D Too bad I can't find any logarithmic pots in the <100ohm region. That'd give me even better control.

Potentiometer: Digi-Key - CT2150-ND (Manufacturer - 026TB32R5AOB1A1)
1.6 Resistor: Digi-Key - P1.6W-2BK-ND (Manufacturer - ERX-2SJ1R6)
LM338: Digi-Key - LM338T-ND (Manufacturer - LM338T/NOPB)
LD Bridge Capacitor: Digi-Key - 493-3275-ND (Manufacturer - UTT1E470MDD)
10k LD Bridge Resistor: Digi-Key - 10KYCT-ND (Manufacturer - FMP200FRF52-10K)
Reverse polarity Diode Protection: Digi-Key - 1N4001FSCT-ND (Manufacturer - 1N4001)

You've intrigued me now, I think I might build an adjustable labby as well. You've already done all the work for me :D
 
Joined
Jun 12, 2010
Messages
462
Points
28
You've intrigued me now, I think I might build an adjustable labby as well. You've already done all the work for me :D

Just keep in mind the link above is for a 1.6ohm 2W resistor.
I'm going to run 2X of them in parallel for 4W total capacity and circuit redundancy (continuity remains if one fails).
Plus putting two resistors inline should help average out any errors in tolerance on the resistors.

Don't forget the power supply and switches I added.
Wait until you see the water-jacketed diode pocket I'm machining for the liquid cooling loop. :evil:
(lol - this is what happens when engineers get hooked on a new hobby... :()
 
Joined
Dec 23, 2007
Messages
2,494
Points
0
Don't forget the power supply and switches I added.
Wait until you see the water-jacketed diode pocket I'm machining for the liquid cooling loop. :evil:
(lol - this is what happens when engineers get hooked on a new hobby... :()

:wtf: For real? That is going to be epic!

Jeez, I was just going to use the lowly TEC from the casio projector :D
 
Joined
Jun 12, 2010
Messages
462
Points
28
Liquid cooling circuit will run about $150 new. Taht's more than the water cooling in my PC :(
80mm radiator, 80mm fan, small pump, fittings... the big cost is the radiator and pump.
Maybe I'll play it patient and hawk Ebay/craigslist for used parts.

The wife's anniversary is coming up so I have to get her something. Plus my lathe's toolpost post clamp striped out and cracked so I have make/buy a new one before I can make most of the water jacket. It'll be my long term project but it should be sweet once i can buy all the parts. A140 diodes can take 1500mA pretty easy so the whole labby is more of fun/cool setup than functional. It should be safe to assume it will have 24/7 duty cycle though. My goal is a self-contained laser in a 3.5"x3.5"x10". Lay the componets out on a plate, mount the switches to a bracket, and bend some expanded metal for the encasing. If I have money left over I'll try for corrective optics.

I'll probably have to sell the damn thing as soon as it's built because it'll cost so much just to make - lol.
 
Last edited:
Joined
Apr 9, 2008
Messages
374
Points
28
^^ was thinkin JUST the same :)

Me too!! :D

LazyBeam,

Is this a correct diagram of the driver that you are building:

3491-lm338-constant-current-aka-ddl-circuit.jpg


Sorry, I am not good a drawing :yabbem:

Thanks!
 
Last edited:
Joined
Jun 12, 2010
Messages
462
Points
28
Me too!! :D

LazyBeam,

Is this a correct diagram of the driver that you are building:

3480-lm318-constant-current-aka-ddl-circuit.jpg


Sorry, I am not good a drawing :yabbem:

Thanks!

That is acceptable and preferable. The resistor, cap, and diode on the left are solely for the protection of the laser diode. They protect against latent static charge, power spikes, and reverse polarity. Obviously, the pot adjusts your voltage and the two resistors in series with the pot limit your max driver output when the pot is @ zero ohms.

I am unsure of what the differences are between an LM338 and the LM318 you have posted.
For our purposes the differences are likely negligable.

Unfortunately my project got sidelined. Days after I got my wife a $200 anniversary gift, her car took a shit while we out on the road which cost another $300. I hate mechanics... they charge too much. :scowl:
 
Joined
Apr 9, 2008
Messages
374
Points
28
Thanks LazyBeam.

Yes, I too prefer those extra components to protect the laser diode :D

About the LM318, oops! That is supposed to be LM338, here is the correct schematic:

3491-lm338-constant-current-aka-ddl-circuit.jpg


Sorry to hear that, I know how it feels... I just "invested" $350 with my mechanic :yabbmad:
 
Joined
Apr 9, 2008
Messages
374
Points
28
With a few changes to the schematic, I changed the LM338T to a LM350T the 1N4001 to a 1N5002 to handle higher currents :D

3492-lm350-constant-current-aka-ddl-circuit.jpg
 
Joined
Jun 13, 2010
Messages
102
Points
0
Wouldn't a pelter cooler work better in this application then water cooling? Your only dissipating a few watts of heat. A decent sized heatsink will work just as well. And my pc is water cooled as well, 130W+ of heat is a little more worthy, lol.
 
Joined
Jun 12, 2010
Messages
462
Points
28
well, because you can is a perfectly fine reason for anything. :)
But you're right, peltiers would be better for a few reasons.
1) Liquid cooling uses amient temp to cool... TECs are heat pumps and can drop temps below ambient.
2) Cost & simplicity.
3) When a peltier fails, you don't drown your electronics in water.

I did end up changing my design to a tri-peltier configuration with forced convection aluminum heatsinks due to cost. The TECs will surround the LD mount. Basically, I'm driving each ~5W peltier with it's own LM317 and driving the diode with an LM338... cooling fans can run directly off 12V.

The main reason was cost and my being broke. Small TECs can be had for <$10... watercooling requires ~$175+ (pump, radiator, fittings, hose...). I'd machine/build somebody a water-cooled setup FREE if they covered the parts/shipping. It'd be so neat that I'd just be thrilled to build it. I have all the parts selected... I'm just too broke to buy them. Anyways, for anyone reading this... the offer is out there.
 
Joined
Feb 1, 2008
Messages
2,894
Points
0
Question:

I set one LM317 to regulate at 1A.

I set one LM338 to regulate at 1A.

Which runs hotter? Will either chip require more/less heat sinking than the other?
 




Top