How laser diode drivers work; An Explanatory Thread

[benmwv]

I have an lt3083 board designed, and i have blank pcbs for it in a box somewhere. I never got around to testing it because they are big, inefficient, and expensive. They have a cool adjustment mechanism, but they aren't really good for a laser driver unless you are planning a labby.

If you are interested in the lt3083 I could try to find the pcbs and mail you one, or at least email you the design so you can get pcbs made.

 

[tony1]

rhd,
you have a good point there, I think it needs further discussion and verified .

 

[Pman]

Realize this is old but nice write-up. I'm in Rochester, N.Y. by the way

Pete

 

[DalDames]

Hi everyone..i am new here. As per my knowledge LM317 driver, the voltage just follows the IV curve of the diode, and will end up being whatever that particular diode drops at the current you've set it to. So if you set your current on a test load with the wrong voltage drop, it shouldn't matter.However, one scenario in which it WOULD change, would be if your driver didn't actually have enough input voltage to handle the laser diode's Vf + its own dropout, but DID have enough input voltage to handle the test load's drop + its own dropout. In this scenario, you'd find that the actual current once connected to a real laser diode, would be lower.

pcb assembly services

 

[Andratos95]

Hey all! This is written for any new people out there who wish to design and build their own drivers. Or, for those who are just curious!

There are roughly 4 types of drivers out there:

Buck

Boost

Buck-Boost

Linear

Each of these designs have their own pros and cons. They are cost, reliability, and efficiency. (the rating system is by +'s, there are a max of 4. a - is half a +)

All regulators can be made to be constant current; it's just how feedback is changed.

But, before we go over the types of regulators, we must go over the feedback method used!

CURRENT SENSE RESISTOR

Current sense resistors are typically low valued resistors put in series with the load (usually a laser diode). For any given current, there is a voltage drop across the resistor. For most regulators, the feedback voltage reference is around 1.25 volts.

This is rather inefficient, as the resistor must have 1.25 volts across it to maintain regulation. For example, lets say, you want 250 mA. The resistance can be found with R = E/I. So, we would need a 5 ohm resistor!

However, when we calculate the power lost, (with P = I * E) it's approximately 310 mW of power is lost in the resistor. 310 mW is a lot of power lost in something that's just supposed to be sensing correct? How can we fix this?

Well, some regulators do you a favor and instead have a much lower internal voltage reference, somewhere in the 200 mV range. Using the same values above, we can find that the resistor would need be around 0.8 ohms. With 250 mA of current flowing through it, it would lose only 50 mW of power! That's a pretty decent increase in efficiency over a 1.25 volt reference!


For regulators that DON'T do you this favor, you can correct it very easily yourself. If you take an op amp, and use a ....0.1 ohm resistor, say. The voltage across the resistor is around 25 mV.


You take the op amp, and use it to amplify the voltage to 1.25, so the regulator *thinks* that it's 1.25 volts, and will regulate the current as necessary. It would keep it at 250 mA, but there's a LOT less power dropped (only 6 mW!) across the resistor as a result. It makes it MUCH more efficient.

This method can be applied to any regulator, switching, or linear. You can make the famous LM317 work a lot better using this... instead of dropping 3 volts by default, it'll only drop around 1.5-2 volts instead! (basically, you won't need as much "default" voltage to get it regulating right)


LINEAR REGULATORS

Linear regulators are usually the most simple. The LM317 driver is the most common you'll ever hear about. They work with two internal parts; an op amp, and a transistor.

The op amp inside has a 1.25 volt reference on it's + input, and the output is tied to the transistor's base. The - input of the op amp is tied to the "ADJUST" pin, where you usually have it measure voltage across the current sense resistor. The op amp will turn the transistor into its linear region; (hence the name linear regulator) where the transistor basically acts like a resistor. It will turn it on more or off more to maintain 1.25 volts at the - input of the op amp, and will regulate current as necessary this way.

They are only good if your laser diode's forward voltage is LOWER than your supply voltage. However, the more your supply voltage is over your laser diode's forward voltage, the more inefficient it becomes.


Cost wise, they are very very cheap. At max, they require maybe 3 external components.

Cost: ++++

Reliability is basically 100%; these almost never fail.

Reliability: ++++

Efficiency is where these really lose it. A linear regulator will drop the excess voltage as HEAT, which becomes wasted. However, if the input voltage is close to the output voltage, then efficiency is improved.

Efficiency: ++




Buck Regulator

A buck regulator is a switching step DOWN regulator. Most of these for our purposes are monolithic; they contain all needed parts inside of the IC package.

They step down voltage by using a switch, typically a MOSFET, a type of transistor. By rapidly turning the switch off and on, (as fast as 2 mhz!) they chop the DC supply up into something called a square wave.

The square wave is then fed into an inductor, which smooths it out into an average voltage of the square wave. It is then fed through a smoothing capacitor, which then smooths out the voltage even further, to essentially pure DC. Once it is fed through the capacitor, it goes through the current sense resistor, to the feedback op amp.

Feedback in a buck regulator works by modifying the duty cycle of the MOSFET. This means, that when the voltage across the current sense resistor is too high, (AKA too much current) the regulator will turn off for a little while to drop the output voltage a bit. If the voltage then becomes too low, the regulator will turn back on again, causing the voltage to then rise. This repeats again and again thousands of times a second to maintain regulation.

Buck regulators are only good if your laser diodes forward voltage is LOWER than your supply voltage. They can supply additional current with more supply voltage though!


Cost is generally higher than linear regulators, though some can be relatively cheap.

Cost: +++

Reliability is pretty good in buck regulators. However, they can be somewhat harder to troubleshoot due to their increased complexity in respect to linear regulators.

Reliability: +++

Efficiency is very good with buck regulators. The higher your supply voltage is the more output current your regulator can supply. There is usually very little heat lost in these types of regulators.

Efficiency: ++++



Boost Regulators

Boost regulators are rather common in single cell (one lithium ion battery)builds. They step UP voltage at the expense of increased current draw from the supply. Again, they are usually monolithic, with almost all of the parts needed included within the IC's package.

Boost regulators work again using a MOSFET typically. There is an inductor in series with a mosfet, and a diode that feeds away from the MOSFET's drain. An inductor is a coil of wire. When an electric current flows through it, it creates a magnetic field. When current stops flowing through it, the magnetic field dumps back into the wire preventing the current from stopping instantly. A boost converter takes advantage of this property; As the MOSFET is in its on stage current flows through the inductor. All is happy in the world, but then the mosfet turns off! Oh noes, where's all that magnetic field current going to go?

Well, the current is dumped into voltage. The more current, the higher the voltage is boosted (up to a point, anyway, there's something called saturation that makes really high voltages impossible). So, the voltage is looking for a way out, and see's the diode right next to the mosfet's drain. It goes through that, and is fed into a smoothing capacitor, so that the output is DC. Then through a current sense resistor, and then to your laser diode.

Feedback is essentially the same as a buck regulator. It will alter the duty cycle as necessary to keep a certain voltage across the current sense resistor, and thus keep a certain current flowing through your laser diode.

Boost regulators are great if your laser diode's forward voltage is ABOVE your supply voltage. They increase voltage at the expense of more current drawn from your battery.


Boost regulators are getting very cost friendly. They can be had for very cheap nowadays; (as shown here http://laserpointerforums.com/f67/free-diy-open-source-boost-driver-tested-working-71433.html)

Cost: +++-

Reliability CAN be lacking in some cheaper drivers. They can be somewhat difficult to troubleshoot, as well. Because they boost voltage, they can kill your laser diode if it is connected up backwards. (the regulator will try to boost the voltage more and more until your laser diode dies from reverse overvoltage!) Some drivers have protection against this though.

Reliability: ++-

Efficiency is very good with any switching regulator; however, boost regulators typically have a slightly lower efficiency than most bucking regulators that I've personally seen.

Efficiency: +++



Buck Boost Regulator

Buck boost regulators are a combination of the above two switching regulators (there's also SEPIC, but that is effectively the same thing as buck/boost, just the component layout is a bit different). This topology maintains regulation by boosting or bucking the voltage output as necessary. Basically, it can step down, or step up the voltage as it sees fit, to run your laser diode at a specific current. You can use this in a single cell, or multi-cell build for your laser.

Operation is a combination of boost and buck regulation. Feedback is achieved the same exact way with all switching regulators, by modifying the switch's duty cycle.

They are great for ALL builds. Single or multiple cell, this driver type really just doesn't care!


Be prepared to pay more money (typically) for buck boost drivers, as they have much more utility than basic linear, bucks, or boost drivers alone.

Cost: ++

Reliability is decent with most.

Reliability: ++-

Efficiency depends on what mode it's in; typically more efficient in bucking mode.

Efficiency: +++



With any laser diode driver, you must find one that fits what your looking for. Linear drivers are the most cost effective. Bucking regulators are great if you want more run time than linear drivers, but typically cost more. Boost regulators are perfect if you want to use just one lithium cell for your build, while Buck/boost drivers are perfect for just about anything; the only compromise is expense.


I hope you enjoyed my explanation of laser diode drivers! :thanks:

Thanks for reading!

Hey, I know this is 10 years old, but I'm losing my mind... Can you elaborate further on how to create a LM317 linear regulator with the op amp trick to reduce the power dissipated on the resistor? I keep searching online but I can't figure it out. It would be extremely useful if you could supply some circuit diagram and what components to actually use for this... Thank you!