DVD drive laser for CNC etching?

http://planetstephanie.net/hi-tech-fun/lasers/dvd-laser/

Be *very* careful when doing metal, 'cos it's reflective and you need to make sure you don't accidently blind anyone.

http://planetstephanie.net/hi-tech-fun/lasers/dvd-laser/

Be *very* careful when doing metal, 'cos it's reflective and you need to make sure you don't accidently blind anyone.

Dave,

Thanks for the link, this is exactly what I need to try it. And I didn't think of being blinded, I will add shielding when I try this.

Also, welcome to E14, and thanks for your first post being a reply for me.

E

from a Physicist who has used semiconductor lasers (Bath, 1994.. PhD, 20Watts from a heavily overdriven GaAs chip) and other industrial Argon-Ion lasers (4 Watts blue and 0.2 Watts hard UV at 244nm (Victoria Uni, and Swinburne, Melbourne Australia) my first advice is to send your friend to the engineering and physics sections of your nearest university library and read everything which you can find on any unfamiliar words which follow.

ablation, welding, absorption, reflection

diode laser arrays, Peltier thermoelectric cooling, water cooling

single mode, M-squared, tapered, high power,

threshold current, light current curves, wallplug efficiency, facet coatings, facet passivation

catastrophic optical damage, dark line defects, electromigration

That last line is what you'll get if you build a ten Amp nanosecond Blumlein impulse generator with five ohms output impedance and hammer the DVD diode with it.

Trust me.  It will be too small to fail spectacularly.  For spectacular you'd need a medical eye surgery laser or preferably a heavy industrial welding laser (these do exist, German I think) and those are not suitable without professional training because there are many ways to die when you have a kW being fed into a box and no idea where it will come out.

You can have lots of fun overdriving any laser which you manage to pick out of the DVD, though definately not enough to cut down swathes of trees or set fire to your sister, or anything fun like that involving cutting metal.  Rigging three car batteries in series with two feet of !hot! coathanger wire wound on a brick and in series with a 2000 turn heavy coil might suit your aims though.  That combination should melt anything of grounded metal which is prodded with it and might weld if the anode is a suitable welding stick and all surfaces have been suitably prepared with a grinder.  !!!!!!!!!!!!!!!!!!!!!!!!!!!! elf and safety warning; at your own risk DO NOT TRY THIS WITHIN SIGHT OF SAFETY INSPECTORS OR WITHIN EARSHOT OF YOUR LOCAL POLICE CONSTABLE !!!!!!!!!!!!!!!!!!!  recommended: dig a six foot by two foot trench in your back garden to save your relatives some expenses and explanations which cause non-techies to glaze over when told and sort of shuffle away toward someone who is more inclined to talk about football or girls or some nonsense like that.

Back to DVD lasers.

The DVD laser is designed to modify very extra tiny spots at its focus, for example melting a 0.4 x 1.0 x 0.2 micrometer zone on a fast moving polymer target and not vaporising it.  DVD's can be scorched if slowed down sufficiently, and it is imaginable to modify a DVD writer to print onto DVD-R disks like a spirograph or possibly more complex marking.  All of the changes would be to override control of the motors.

Procedure to max up the output power of a small diode laser beings with finding its normal operating current with some fairly dull multimeter measurements, and then lower the current to find out when it goes out.  This is called the threshold current and is really crucial because it tells you how much current the laser might handle.  DVD's might have a threshold of 20mA and a normal operating current of ? 50mA ? and the rest of this is based on these figures.  If the normal operating current is less than 10mA then you probably have a VCSEL and those are smaller and even less suited to raising their power.  You probably want to also measure the voltage at turn on and at twice threshold.  From these a straight line extrapolation tells you roughly whether you want 0.2 Amps or 0.4 Amps to acheive 8x theshold current, which is about where catastrophic optical damage can be expected.  Next build or borrow a pulse generator which must work at that desired current.  Pulse width less than 100ns, repetition rate 1kHz is feasible on home-made etched PCB.  Going pulsed lowers the average power.  If you don't run it pulsed you'll get catastropic thermal damage before you've reached the maximum light output, usually failing to open circuit when the wirebonds melt or the contacts lift off like thermostat metal.  irrecoverable !!! tssssss !!!!

Once you've found the maximum pulsed current that it can take, look up all the tricks used by overclockers to cool other PC components, particularly small ones.  If you can obtain a ten Watt watercooled heat-sink and plumb it in to a water cooled PC then you've a good platform for tweaking laser diodes.  This is not safe because anything above 0.5mW can wreck your eysight, so bolted down and pointing away from everyone is strictly necessary, with black curtains and cardboard boxes the minimum beam confinement.  Another less expensive way to get some temporary cooling is to drill out a block of solid Aluminium to set the diode case in, leave the whole unit in the freezer for a while, power it for up to a minute at a time.  If you see the diode output colour shift orange-red to red, or blue to greener, that is really really bad and indicates more than 50C change at the diode junction.  That redshift on heating can be measured and write to systems253@zerotechnology.co.uk if you'd like to know how.  That is what one needs to maximise the pulse rate having found the highest pulse current.  This WILL NOT CUT METAL but might make a spectacular disco light, so long as it is kept on the top shelf pointing upwards.  Modulate the pulse rate proportionately to, say, an audio signal, which can be done with substantial electronics work and a £1 PWM chip from Fairchild, using a diode and 2.2 nF capacitor to get mushy spikes of 10,000ns from the edges at repetition rates as high as your thermal limits can handle.  Sharpen those up by feeding to a Schmitt trigger, and drive a fast £1 power transistor with the output.

If you are really ambitious and have megabucks, then 'mode-locking', 'femtosecond pulse generation' and 'ablation' become relevant, but there is almost no way to set up for that in your garage and no way to observe the ultrashort pulses without either buying $200k worth of test equipment or building an auto-correllator.

Have fun, and read everything which you can find.  Most of it does not help me to earn money but it does make the world seem more interesting.