High-Speed Laser Driver with Arbitrary Waveforms

Thread Starter

AlexX 52

Joined Mar 15, 2018
14
Hello everyone,

I have a red laser diode (this one) which I want to use in an application.
It should be driven with currents up to 300mA (resolution of ~0.1mA) at pretty high speeds (up to 10MHz, 100MHz would be even better) with arbitrary waveforms.
These waveforms are generated by a microcontroller connected to a DAC.

What I understood so far is that a constant current driver is required since voltage regulation is rather unprecise. Laser driver ICs on the market seem to be used only for digital communication, so they are out of question. I thought about connecting the DAC output to a high-speed Opamp acting as a current source driving the laser through a BJT (because MOSFETs at such speeds have capacitance issues).

Are these requirements even feasible? And is my approach ok or far off? I am relatively new to LED driving, so any help is appreciated.

Thanks in advance.
 

Thread Starter

AlexX 52

Joined Mar 15, 2018
14
The application is a device that mixes three lasers to produce different colors. Since a lot of laser and human parameters are nonlinear in respect to perceived color, the uC's job is to calculate the correct output voltage colors. The arbitrary waveforms can be controlled by the user in form of either standard waveforms such as sine, cosine, square, etc. and waveforms generated by code procedures running on the uC. The device should also be able to communicate with each other, so that's why such speeds are necessary.

About specific parts... well, everything is still in early planning phase, but a few things are known:
The DAC needs to operate on these speeds (depending on what is doable) and offer a precision of at least 12bit.
The microcontroller will most likely be the STM32H743ZI since it satifies those requirements.
 

Papabravo

Joined Feb 24, 2006
22,082
Does your DAC have a voltage output or a current output?
I'm thinking you want something like a push-pull output stage. Don't know what opamps you may be looking at. Don't know of any DACs that work at 100 MHz. either. My experience is in process control and industrial automation.
 

Thread Starter

AlexX 52

Joined Mar 15, 2018
14
Current output DACs don't seem to offer the current requirements and especially not the high-speed ones.
Do push-pull stages produce a constant current from a variable voltage? How would it look like to drive a LED with it?
 

Thread Starter

AlexX 52

Joined Mar 15, 2018
14
Regarding opamps: I read an article from Analog Devices (here) that describes how to design a high-speed current source. Don't know if this would work with laser driving.
 

Papabravo

Joined Feb 24, 2006
22,082
Current output DACs don't seem to offer the current requirements and especially not the high-speed ones.
Do push-pull stages produce a constant current from a variable voltage? How would it look like to drive a LED with it?
Not exactly. BJT devices are basically base current it: collector current out. Variations in base current are often derived from a voltage source. That was not the reason I mentioned it. With a single transistor driving stage you often have an active driving direction and a passive recovery. A push-pull stage, like the output of a logic gate can source as much current as it can sink. It seems to me this symmetry is important in achieving high speed operation, especially if the diode looks like a capacitive load.

BTW constant current sources use a resistor to set the current level. You won't be able to do that.
 

crutschow

Joined Mar 14, 2008
38,506
The application is a device that mixes three lasers to produce different colors.
Since the eye response to a change in perceived color is many milliseconds, why do you need sub-microsecond response to change the color? :confused:
The device should also be able to communicate with each other, so that's why such speeds are necessary.
What devices?

You need to better explain exactly what you are trying to do, so we can offer better solutions to your problem, not just a solution to your proposed solution.
 
Last edited:
Top