Need a high-speed TTL driver

Hey guys, I am new here with a quest of information about building a driver.

I have an 808nm diode that has these specifications that I measured while it was in operation with a constant-current driver

Diode Voltage Drop: 1.9V
Diode Current Consumption: 374mA
Calculated Diode output power: 721mW

its probably a 750mw, because my supply was at 3.2v instead of 3.7v which is what the driver wants. it was from a green laser, but the pump crystal is missing however I dont need the crystal.

I want to be able to modulate this diode to send on-time pulses as low as 40ns, to maybe as wide as 150ns.

The repetition time will be somewhere around maybe 3 to 6 pulses per second.

is there a schematic out there that could do this? I've stumbled across a couple but they are slow, in khz range.

Bump. Nobody?

TK ---
I think everyone here is running the math on
your numbers. According to your data, this
808 diode is over 100% efficient - Nice!!
I WANT ONE TOO☺
Replys to data like this can take a while
to be fully recalculated.
HMike

As Mike said, the diode has 750mW power consumption NOT light output. it probably has between 300mW and 500mW of light output.

Sure, there is nothing hard about generating 40nS (about 25MHz) pulses, but what do you want it to do? Are you looking at modulating information onto it using pulse length or just using it for measurement? A simple TTL oscillator block will give you pulses at that speed and you could just switch it off a few times a second.

I am using it for measurement. The numbers I mentioned are power consumption of the CW Diode I currently have. I incorrectly stated output power, Sorry. Not too sure of the output power, but that isnt important at this point, but I would like this adjustable in case I want to try another CW diode down the line. I have looked at Q-Switched and pulse diodes but I cant justify the cost, plus they are many times faster and more elaborate than what I need. I cant really debrief too many details because not even I know exactly what the end result will be. Experiment type thing for now.

I need single shot pulses. Meaning, the laser is off (biased maybe). the whole time until I trigger a single event pulse, in which case would be laser output for 40ns. Then off again until I retrigger from a microcontroller, such as a PIC or AVR.

so a 40ns pulse from the microcontroller would give me 40ns of light output. (again, im assuming 100% efficiency here, even though its not, so of course I am not counting capacitance and propagation delay.)

As far as the pulse width goes, I will make the adjustment in the microcontroller on-time of the trigger pin, or maybe external gates. (as AVR/PIC wont toggle a pin that fast, or maybe a retriggerable adjustable one-shot circuit, more viable option, but ill figure that part out. Just need the driver that can handle it.)

I don't think a PIC or an AVR will run that fast, you'll need one running at several hundred MHz to get pulse widths that short.

You could use a PIC to drive a Pulse Generator IC. A quick search of the net found IC's for driving laser diodes that are able to generate a rise time of 1ns or so here.

There is a thread here that touched on it

Hope this helps.

Thanks.

I am aware the PIC/AVR isnt fast enough, thats why I mentioned in my post that i probably could use a one-shot or some other timing method. Anyway...

I took a look at that IC, and the max current is only 400ma. If i needed to go to a higher laser diode power, that IC is ruled out. wonder if a discrete output follower could be added to it?

Edit: Nevermind, this IC might be out of the question. cannot find a suitable supplier (digikey,mouser) and no datasheet/downloads.

I uploaded a laser driver circuit that I found inside a laser printer manual. Could this be modified to work somehow? I am good at digital circuitry, but my analog circuitry is a little rough, thats why I ask.

Might start looking at printer optical engines for parts.

Yes I re-read your post and missed that you knew a PIC wouldn't be fast enough. There is some practical information here although it looks like he is getting around 1ms rise times Pulse Circuits for Infrared LEDs and Visible Diode Lasers

You may be lucky using a high speed VHF or UHF MOSFET for switching and a simple resistor would be ok for current limiting as the diode should be able to handle a lot more current for short pulses as long as you have a 1:20 or 1:50 onff ratio.

Ya, well at this point I am not sure if I need to exceed the currently recorded CW calculations.

But ill keep that in mind. Need to find a suitable transistor.

I just had a thought on this- You may be able to get an interrupter from ebay second hand like this one- they will open for as little as 10ns.

I had a look at that datasheet, it does as low as 10ns on time, but the problem is the rise/fall time of the shutter is nearly 3ms. too slow, because by the time I issue the command to open the shutter, it takes 3ms before it finally opens, my measurement window has passed 5 times before that happens.

Anyway, I am probably going to start my experiment off like this:

Probably going to use a 2N3866 RF power output transistor in a shunt configuration. Use a B-C shotkey to prevent full saturation, a speed-up capacitor network on the base, and then a resistor on the collector for full diode current, and a parallel resistor for the LD bias.

I guess i could do it both ways, a diode shunt so when the transistor is on, it shunts across the diode through the bias resistor, so the LD is in bias mode until the transistor shuts off, then its fully on.

Other way would be a sink driver. LD is biased partially on when the transistor is off. Transistor comes on, full current hits the diode. But from my findings on google the shunt method is faster.

Might start here and see where that goes. the transistor drive will probably come from a logic adjustable one-shot circuit, or even a small xlinx FPGA.

2pcs adjustable current laser diode driver 80-500mA w/TTL | eBay
10kHz max switching potential.
I just ordered 2, idk if they any good yet tho.