4 MHz driver for 50 mW 405 nm laser driver

Hello,

I am trying to modulate at least a 1 Mhz digital signal into the laser beam of the 50 mW 405 nm laser diode SLD3232VF. The signal comes from MCU @ 3.3V logic level - however it can be shifted as needed.

I started with the simplest setup - LM317 with BC337 transistor which can pull the adjust pin of the regulator low (which shuts down the laser) or can act as a switch (I tried both variants). After some troubles with the saturiation of the transistor I managed to switch the the transistor at 1 Mhz. In this setup with connected resistor instead of the laser, I could saw a nice 1 Mhz rectangle signal in the current on the oscilloscope. However when I switched back to laser, everything works up 500 kHz - than the laser beam dissapears. However the scope shows nice rectangle signal at the laser input..

My question is - how to switch laser on these (even better higher - I am aiming for 4-5 Mhz) frequencies? What am I doing wrong with my setup? How could it be improved? How could it be implemented in another way?

Is there a complete driver module, which operate on these frequencies (at reasonable price)?

PS: I'd like to also switch to 500 mW in the future instead of 50 mW.

I've been playing with high speed in the past, and you could use a circuit like this:

Merghart.com - 2 Transistor current source

It is very simple, reasonably accurate, but most importantly: can be made to switch -really- fast. Some modification is required: You could take out the 10k resistor to positive, and run a 1k resistor into Q2's base. I'm not sure it would run from 3.3 volt logic, 5v works fine.

I haven't really tested the absolute maximum speed of this circuit, but given the transistion frequency of the BD139 (190 MHz) and the current gain it needs to provide to drive a 50 mW laser diode (10 times perhaps) i see no reason 4 Mhz is a problem. Do add some proper decoupling caps to the power rail though

If you need more speed you could use faster transistors - you'd probably be looking at ones intended for RF instead of general purpose: Since you laser is low power it doesnt require much current and you could use something like a BRF96 transistor for both. You will need to construct the circuit very well for really fast (say > 10 MHz) operation, and watch out for radio interference too!

You could try driving the ADJ pin with dual Op-Amps instead of a transistor. I can dig out a schematic at some point for you if you can't figure one out!

LM317 should have no problem with those frequencies, so that's not the issue. Unless your diode can't keep up, perhaps the pulse is too short to get stimulated emission going?

Diode lasers have no problem in that regard - just imagine how fast a diode has to switch on and off in a dvd writer to keep up with the data rate!

DPSS lasers will give you problems when trying to switch them at high speeds. Projectors that mix diode and dpss (e.g. red direct mixed with green dpss) have problems with blanking timing that need to be corrected in software to get the desired output.