Diode-Testing Lab Power Supply Question

Hi,

I want to use some type of ac-dc ps to test diodes. And I was wondering if what I was planning should do the trick.

I am about to convert an old atx computer ps into a ~lab ps with 0V,-5V,5V,-12V and 12V ouputs.
(using this method: http://www.wikihow.com/Convert-a-Computer-ATX-Power-Supply-to-a-Lab-Power-Supply)

I plan to add a variable voltage output using an lm338 as descrtibed here:
http://www.wikihow.com/Add-Variable-Voltage-to-Your-ATX-Based-Bench-Power-Supply

I would plan to set the variable voltage output to about diode voltage drop (~3V for 20x or ~5V for phr) + 3V and take this output into my DDL driver. I might even add the DDL into the lab ps for real easy diode testing. This way I could test variable voltage output for correct volts, set ddl output for correct current using dummy load, power down and drain capacitors, then connect to breadboard attached to laser diode and power up ps.

This seems like it should work great to me, anybody have any thoughts otherwise? Iwould make something more simple, but I have other uses for the many voltage outputs from the ATX ps.

Thanks for your advice!

-Chop

Thats what i've done

Mine was a tiny psu from a small form facter computer (can't remember the maker) but it's rated at only 180 watts. It's got the ground, 3.3v, 5v, 12v and -12v which, with the 10 watt resistor providing a constant load, i can get mix and match any of the pins to get any output voltage required for most things. e.g. 3.3v + 5v would get me 1.7v
It's also got an extra variable output with an LM317 for when i need a bit more acuracy.

It's one of those things you realise you need until you have it, then you wonder how you managed without ;D

I don't know why I never thought of it before. Someone mention the idea in the geek forum and I did a quick search. Now I've got this decent DC power supply made out of something I just had sitting around. I too have one of the small 180W power supplies, but the case looked too small for me to work in especially if I add the voltage regulation circuit. I'm using my old 300W ps which has much more room and has two cooling fans. With the extra power I'll also be able to use it for a cool setup my school's society of physics student can use to show off to kids. I have an old 40W car audio amp and an old 12" sub woofer that I will use it to power. Throw some plastic wrap over the speaker, put a non-Newtonian fluid (corn starch and water) on it and run a sine wave from my computer through the amp and watch the fluid grow where the maximums of the standing waves are! It's real cool.

Update: It works great! I just used it to test my two PHR's that I got in the mail from Drew. There's pics too

http://www.laserpointerforums.com/forums/YaBB.pl?num=1225074433

I added the voltage regulation circuit, but I haven't mounted it yet. Before I made a mistake I realized that the voltage on the heat sink is different than the voltage of the power supply case (ground) since the voltage regulator is connected to +12 and -12V to get the total potential difference if 24V. Right now it's just hanging out of the side of the power supply where the cables used to come out... :-/

It works good though, I have -12V, -5V, 0V, +5V, +12V and variable 1.25V to ~22V. It also works as you might think for any combination (i.e. (-) connected to -5V and (+) connected to +12V yields a potential difference of 17V...

Be careful of voltage stored in capacitors when connecting/disconnecting laser diodes as the sudden discharge of energy from an intentional or intermittent connection can be enough to damage a laser diode. Often, with the linear regulators, there can still be residual charge in the caps, so discharging them before any connection or disconnection to a laser diode is a good idea.

Also, you can put a high value resistor (like 2k or more) in parallel with the laser diode output (from positive to negative output terminal), to help discharge any output cap before it can damage a diode, though it may not be fast enough depending on your capacitance and how much time is allowed for discharge (for example, an intermittent connection may be for a few milliseconds, possibly not enough time to discharge the cap enough to prevent damage).

So, make sure you have good connections to the diode while operating the test supply and also have some way to prevent spiking discharge from the circuit to the LD.

Thanks for the advice. I've been killing the mains power an letting the 10W internal resistor dissipate the caps in the main ps. It usually takes 5-10 seconds for the mains and power LED to dim all the way. I have a 660ohm resistor in parallel with the LD, maybe I should put a higher value to waste less power. Even with the resistor I've been sure to short the 47uF cap on the output before I connect the LD.

I mainly wanted an easy way to test my diodes, which seemed to work. Your drivers should arrive today so I'll be able to make a permanant build as soon as my hosts arrive!

My next project in planning is to make a voltage divider box to allow me to use my 24bit 96kHz sound card as a oscilloscope. Anybody ever done that? I just don't have the money to buy a real scope.