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



Switching driver for 445nm diodes

erdabyz

New member
Joined
Jun 30, 2008
Messages
286
Points
0
Well I've bought a couple of 445nm diodes recently and I've designed my own switching driver to power them.

I've used this integrated circuit:
Linear Technology - LTC3454 - 1A Synchronous Buck-Boost High Current LED Driver

With this coil:
Free sample request

It should give steady 800mA to the diode during the whole discharge cycle of the battery, and 1A until the battery reaches 3.2V or so.
It offers 3 selectable powers via 2 inputs

Here's the schematic:


And i've attached the PCB files (they are in PDF, print without scaling)

You have the PCB top and bottom layers and soldermask top and bottom layers, you'll have to mirror the top layers if you use insolation or toner transfer procedures.

Some notes:
-Both 0805 capacitors are 10uF X5R ones (6.3V min)

-The 1206 capacitor is a 100nF one. I've used 1206 because it's what I have on hands, but you could use 0805 or better 0603. If you want the pcb file to modify it, ask for it.

-See top soldermask and IC's datasheet to locate where the components go.

-The potentiometers I use are these: (again because I have them on hand):
Trimmers | Digi-Key

- The 4 pads in the top-left corner of the PCB are for power selection. Read the integrated circuit datasheet. You can either make a solder bridge or solder wires to a daughter PCB with switches and that...

- The two holes that connect the coil to de IC are vias and you must use a solid connection between layers.

- The L+ and L- connection are possitive and ground for the diode. You should solder wires from there to the diode.

- This is how the driver is designed to be mounted:

The six holes that are placed in a "circular fashion" would be used for connection bridges between the driver PCB and the positive contact PCB.
If you see the soldermask, there are two pads, one in the top and one in the bottom, in the right side of the PCB. You should solder a piece of copper wick between them, and that should be the negative side contact point to the heatsink. You also have other holes in the grould plane for other grounding cofiugurations, like using the case pin of the diode, and there's also a bigger hole for positive contact to solder a thick wire in there if you don't wanna use the two boards configuration.

You should cut the PCB to a 2cm diameter circle, you have the outline for that in the soldermask.


I still haven't got the coils, but they should arrive this week. When they do, i'll mount the driver and set up some testing to confirm it works properly. It does in the simulator.

Read the LTC3454 datasheet to know how to adjust the driver, what are the potentiometers for and all those things i don't wanna explain here.


Hope this information helps someone. If you wanna give me a thank you for this, donate something to me or whatever just send a PM. I just post this info here because making these drivers is a hobby for me and I don't have time to manufacture them in order to sell them or something so i guess it's better to just post all the info so people can take advantage of it.

Any questions, doubts, criticism, just post and i'll try to answer.
 

Attachments




Rafa

New member
Joined
Feb 4, 2010
Messages
437
Points
0
Oh, those are some awesome drivers. Hope I get soon tools needed to make them.
 

mojo_1234

Active member
Joined
May 1, 2009
Messages
187
Points
28
Nice driver.

Do you or anyone else know about an "easy" to get buck converter IC (step down) which is adjustable just by a reference voltage?

Thanks,
mojo
 

ArtDen

New member
Joined
Oct 28, 2009
Messages
53
Points
0
I tested LTC3454 at 1A output. LTC3454 has low efficiency at 1A and produces big amound of heat :( It works about 30 seconds until overheating and then switches off due to overheat protection. But at 0.7A (and less) it works fine.
 

rev0

New member
Joined
May 5, 2010
Messages
110
Points
0
Have you tested the output of this thing on an oscilloscope? I've had some pretty bad issues with a similar project, see here: http://laserpointerforums.com/f42/pwm-drivers-lasers-why-why-not-52477-3.html#post748475
Switching converters have inductive spikes that could potentially kill or damage laser diodes. It would be interesting if you could get an oscilloscope screen or two of your driver, I know mine has some pretty poor switching noise and I don't feel like risking my 445nm diode.
 

ArtDen

New member
Joined
Oct 28, 2009
Messages
53
Points
0
Switching converters have inductive spikes that could potentially kill or damage laser diodes
Height of inductive spikes depends of type and capacity of output capacitor. 10 uF ceramic capacitor will smooth spikes very well. You can find oscilloscope screen images at fifth page of datasheet. Value of spikes doesn't exceed 20 millivolts with 0.5A current at the load.

PS: I will use LTC3454 for my first driver for my future 445 (at 0.7A) :)
 
Last edited:

rev0

New member
Joined
May 5, 2010
Messages
110
Points
0
Height of inductive spikes depends of type and capacity of output capacitor. 10 uF ceramic capacitor will smooth spikes very well. You can find oscilloscope screen images at fifth page of datasheet. Value of spikes doesn't exceed 20 millivolts with 0.5A current at the load.
Voltage may be much smoother than current.

Here is an oscilloscope shot of my driver, channel 2 (blue) is current (1mV = 1mA), channel 1 (yellow) is voltage:



I am using an 18uF ceramic capacitor as suggested using National's simulation tools.
 

ArtDen

New member
Joined
Oct 28, 2009
Messages
53
Points
0
I don't have oscilloscope here so I can't show my diagram. But I trust the information from datasheet :)
 

erdabyz

New member
Joined
Jun 30, 2008
Messages
286
Points
0
I have access to an oscilloscope so I could test the voltage, current or whatever. By the way, it's not the first switching driver I design and test, and I can tell you that if you properly design the PCB (most issues come from there) they give a really stable output, with ripples in the tens millivolts range at most.

LTC3454 should be pretty efficient if designed properly at 1A, datasheet says no less than 80% and simulator confirms that. Saying that a 445 diode works at around 4.3V 1A, that means 4.3 electrical watts delievered to it, which is 5.37 watt drawn from the battery at 80% efficiency, which is 1Watt dissipated in the circuit. I designed the PCB so the circuit would have a big copper area surrounding it to heatsink, and with some via holes to spread the heat to the even bigger copper area in the back... i hope that's enough. 80% is not at all "low efficciency", what happens is that you're dealing with serious amounts of power whose 20% is still a serious amount of power.

Issues might also come from the inductor. I use a massive one, a little more massive than the recomended by linear for 1A applications.

As I said inductors should arrive this week so i could build and test the driver.

EDIT: just in case, I feel safer powering my diode with a switching driver than doing it with a linear one.
 
Last edited:

Rafa

New member
Joined
Feb 4, 2010
Messages
437
Points
0
EDIT: just in case, I feel safer powering my diode with a switching driver than doing it with a linear one.

^Are you serious? lol
 

erdabyz

New member
Joined
Jun 30, 2008
Messages
286
Points
0
My first 445 is working at 0.75A driven by LTC3454 now :)

Good to know! My inductors should have already arrived so I guess they'll arrive early next week.
I'll post testing results. I have some 0.6mm thick copper foil and I could use it as driver heatsink to go up to 1A.

Could you post your PCB files to see how your driver is?
 
Last edited:

ArtDen

New member
Joined
Oct 28, 2009
Messages
53
Points
0
Could you post your PCB files to see how your driver is?
No. I have only photo of my driver :)
3454.jpg
(inductor is on other side of the board)

It is too small and overheats very quickly at 1A :(
 
Last edited:

erdabyz

New member
Joined
Jun 30, 2008
Messages
286
Points
0
Okay, my driver is done and working. Stable 800mA for now (cound't turn the potentiometers higher before the test load diodes were almost fried and the potentiometers I use are a pain in the ass to adjust).

Running for like 50 seconds and NO heat in the driver. Well, to be sincere, around 40ºC but barely noticeable. I knew my thermal design was "right" because when I was soldering the through-hole wires I had to increase my soldering iron's power to get the solder attached to the copper pads, and when that happened almost all the PCB was at 180ºC or so...

Gonna make more testing later.

BTW ArtDen, good job!
 
Last edited:

ArtDen

New member
Joined
Oct 28, 2009
Messages
53
Points
0
BTW ArtDen, good job!
I planned to use this driver with bright LED not LD. Now I plan to use more power chips for "driving" blue laser. Even 1A is not enough for me. I need minimum 1.3A :)
 

erdabyz

New member
Joined
Jun 30, 2008
Messages
286
Points
0
I planned to use this driver with bright LED not LD. Now I plan to use more power chips for "driving" blue laser. Even 1A is not enough for me. I need minimum 1.3A :)

We all need moar power I guess xDD I'm also trying to find the perfect IC for drivers. LTC3454 has just EVERYTHING someone would want for a 445... except for the current.... It's buck-boost, it has selectable power, it has soft start, it has thermal, short and open circuit protection, it doesn't waste power on external resistor and uses reasonable sized inductors.

BTW just an advice, try to use wider traces on your next PCB's. I mean, expand them as much as you can as soon as you can after they exit the IC surroundings and try not using straight angles. They can be problematic at high frequencies.
 




Top