Here is the text form this section on Sam's Laser FAQ...
"
Sam's High Power Laser Diode Driver 1 (SG-DH1)
This isn't exactly an entire design but one that uses a common logic power supply in an unconventional way. It may be possible to use a high current switchmode power supply as a variable current laser diode driver as long as it has remote sensing capability. The remote sensing feedback loop maintains a constant voltage (the spec'd supply voltage) between RS+ and RS-. Normally, this is used to compensate for the voltage drop in the wiring harness. By applying a variable control voltage between RS+ and V+, the power supply can be fooled into producing any output voltage from near 0 to its maximum rating as long as its minimum load requirement is satisfied. With a small resistor in series with the laser diode (or for those willing to take risks, the resistance of the laser diode), this results in a variable current to the laser diode. The only limit on output current is the maximum rating of the power supply. These types of power supplies, capable of 50 A, 100 A, or even higher current, are readily available on the surplus market. However, this scheme may only work with certain models, those which power their control circuitry separately from the main output and don't go into some sort of undervoltage shutdown if the output voltage goes too low. I don't know how to determine which models satisfy this requirement.
Vicor has application notes on doing this (among other things) with some of their Flatpac (among other) models. Search for "Programmable Current Source". The power supplies shown have an additional input called "Trim" which makes the modification particularly easy.
I have not yet attempted to close the loop and provide actual current control but have opted for voltage control for now at least. The unit I've been using for these tests is a Shindengen PS5V100A, a fully enclosed fan cooled switchmode power supply that's about 15 years old. This unit is also nice in that it regulates well with no load. All that was needed was to remove the shorting link between V+ and RS+ and install a 20 ohm, 2 W resistor in its place. Then applying 0 to +15 VDC current limited by a 47 ohm, 5 W resistor across RS+ (+) and V+ (-), the output voltage would vary from near 0 to 5 VDC.
Code:
RS- <------ Remote Sense -------> RS+
o o
| V- Vout V+ |
| o o |
| | R0 | R1 | R2
| | 250 | 20 2W | 47 5W Vcontrol
+-----+---/\/\---+-----+---/\/\---+---/\/\---o + 0 to 15 VDC - o---+
| | |
+---|<|---+---/\/\---+-+-------------------------------------------+
LD1 | R3 |
Laser | .05 500W | Adjusting Vcontrol from 0 to 15 V varies
Diode o o Vout from 5 V to 0 V.
VS- Vsense VS+
(R0 is internal to this particular power supply.)
R3 can be constructed from a length of building wire. For example, 20 feet of #14 copper wire has a resistance of 0.05 ohms but water cooling would be needed if run near full current. I'm actually only using a head lamp load for testing and it works fine.
The same scheme using RS- did not have enough range, probably due to the internal circuit design. This is too bad because the op-amp circuitry to drive it might have been simpler, or at least more intuitive to design.
(I did try a test of the same approach with a Pioneer Magnetics dual output power supply (5 VDC at 59 A, 12 VDC at 67 A). While control was possible, it didn't behave nearly as perfectly as the Shindengen supply. More than 1/2 A of control current was required to change the 5 V output to 4 V. And while the 12 VDC output could be reduced to near 0 V, the cooling fans cut out at about 8 VDC so they would need to be powered separately for continuous operation at high current. But this might be nice for driving series connected laser diode bars.)
The challenge is to convert this to a user friendly form that is safe for the laser diode. I am designing a control panel which incorporates what I hope will be fail-safe circuits to minimize the chance of excessive current either from power cycling or by user error. It will use closed loop feedback so the actual current can be set (rather than voltage) and includes a multifunction panel meter (set current, actual current, diode voltage). It will enable diode current only if all power supplies are stable and correct, the 10 turn current adjust pot is at 0, and with the press of a green button.
However, initially, I'm using a 10 turn pot to control the current with a digital panel meter monitoring current via a 0.025 ohm sense resistor. Current is limited to between 50 A by a 0.06 ohm power resistor. Believe it or not, even 50 A is way below the limit for the diodes I need to test! See the section: Characteristics of Some Really High Power IR Diode Lasers.
The schematic in Sam's High Power Laser Diode Driver 1 includes the control panel, connections to the 100 A power supply, and laser diode wiring.
The basic control panel includes an Enable switch (eventually to be replaced with a keylock switch), Diode On and Off buttons, the 10 turn pot and DPM which reads 0 to 100 A. A differential amplifier converts the voltage across the current sense resistor into a DC voltage for the DPM. Without the differential amplifier, the control current was seriously affecting the readings as 1 A is only 2.5 mV. It's not possible (or at least not convenient) to separate the power and signal wiring to provide a proper single point ground.
Both the sense and current limiting resistors are simply lengths of #14 copper wire with forced air cooling. This works very well with the diode's output digging pits in my brick beam stop. However, for continuous operation, it may be necessary to replace the #14 with #8 because even the modest heating of the copper changes its resistance enough to noticeably affect current.
With minor changes in part values for the current limiting resistors, and the set-point for the power supply output voltage, it should be possible to drive a pair of laser diodes in series as long as they can be isolated from the common point. (The positive connection to a high power laser diode is usually the mounting block of the diode but it may not be connected to the external case itself.) However, one risk with this setup is that if one of the laser diodes fails shorted, it will likely take the other one as well since the current will spike to a very high level.
The setup is shown in Photo of Sam's High Power Laser Diode Driver In Action. The water-cooled laser diode in the aluminum box is capable of 35 W output at around 55 to 60 A. The power supply is at the upper left with the control panel in front of it showing 40 A. Behind the power supply is the coil of white wire acting as a current limiting resistor next to its cooling fan. The current sense resistor is the 12 inches of so of red wire running from the power supply to the terminal strip. The blue-white glow is my digital camera's response to intense IR. The camera is really confused. When viewed through IR blocking laser goggles, a line on the brick starts glowing at a current of around 35 A and is white-hot at 45 A, where the current limit of the power supply is presently set (via the current limiting resistor and wiring resistance with the power supply adjusted for a maximum output of 5 VDC). The old darkroom enlarger timer in the upper right is used to turn the driver on for exactly the 20 seconds needed for my "meat thermometer" type power meter to take its reading, which would show about 23 W at 40 A for the diode in the photo. The reading at 45 A is about 27 W."