Die4LASER.com
The guide
to soldering grains of sand
To contact
the author you can email me at…
27/August/07 D4Sparks – a ZVS transformer driver for Flyback Transformers and MOTs. D4Sparks page
7/March/07 Die4Thing – a
laser power meter display head and micro-hotplate controller. D4Thing page.
26/December/06 Happy Holidays! –
I’ve been playing with a sealed CO2 laser.
A few pics and some video clips are here.
4th/April/06 – Die4Drive – a 1A laser diode driver
5/Jan/06 –
Built a Laser Projector – pics are here.
21/Nov/2005
– Added the dissection
of a DVD writer.
31/Oct/2005
– Added the Die4DAC
area.
5/Feb/2003
– No time to update at the moment – so just dumping a few photos of interest.
http://www.die4laser.com/thedump.html
First off
you are going to need some specialist equipment:
Stereo Microscope, preferably zoom type. (Very difficult without one
of these)
Hot Plate,
adjustable around 110C
Some method of cooling the hot plate – fan, running water, T.E. etc.
Wooden shafted
Q-tips
High
temperature oxygen free copper slab or bare c-mount
Indalloy 93C melting
solder.
Flux -
liquid Rosin type (I tried the flux that came with the Indalloy
solder from SRO. On bare copper the flux did not work at all.)
If you
cannot bring yourself to spend $30 ish on 0.7 grams
of gold plated copper (c-mount) or your local engineering material store has a
minimum order of 1kg of HTOFC then there is another way. You can reclaim the
heat spreader from large transistors. (Just check that they are cheap silicon
power devices and not Beryllia filled RF power
transistors.)
First take
the donor device
Then distress
it. I used a large pair of pliers as a hammer and a vice body as an anvil.
NB. Eye
protection is mandatory here.
Poor thing. You can see the remains of the transistor die in the
middle.
Heat the
copper block up to 350C and clean off any solder.
Cut to
size and drill mounting holes as required.
A piece of
double sided PCB material is soldered onto the copper slab by heating the whole
assembly with a soldering iron. The PCB material has a gap etched or cut in the
top surface to form two isolated pads. A wire is soldered from the right pad to
the body of the copper slab.
It is
important that the edge of the mount is finished to a sharp square clean edge
where the die will be mounted.
A piece of
0.25mm wire is soldered onto the edge of the PCB material. Bend this wire
vertical so it is out of the way during later operations. If you intend to use
the diode for CW operation only then you can add static protection consisting
of a 10nF X7R 0805 capacitor, 470pf C0G 0805 capacitor and a 100R 0805
resistor. If you want pulsed operation solder a piece of wire between the two
pads to short them together. This shorting link should only be cut when the
diode assembly is mounted and connected to your driver circuit.
This is a
3W 808nm die. The mirror finish and parallel lines indicate that this side is
the anode (+) and the lines run in the same direction as the cavity.
The laser
cavity is on this side of the die and this is the side that must be well
cooled.
The same
die turned over. The slightly dull sputtered finish indicates that this is the
cathode (-) side.
The gauze
behind the die is the top surface of a gell-pack.
There is a clear tacky gell that protrudes slightly
through the gauze. The die is then stuck to this surface. The correct way to
release the die is to connect a vacuum pump to a small hole in the rear of the
plastic frame that holds the gell and gauze. This
causes the gell to be sucked back through the gauze,
hence releasing the die. If you do not have a vacuum pump the die may be
released (at your own risk) by carefully prizing a corner of the die up from
the gell with a suitable probe. *
- Suitable probes include
wooden cocktail sticks, cactus needles, or my personal favorite
the sharpened wooden q-tip. Anything metallic is an absolute no-no.
Wooden q-tips – 2 tools in one. Sharpen one end and it is
perfect for manipulating die, the other is great for moving flux around over
the die mount.
Talking
about the die mount – it must be tinned prior to mounting the die. This
requires a hotplate.
A real
rough and ready set up. A pic16F877 micro, power MOSFET, 3.9Ohm 25W resistor
mounted on a heatsink, precision thermistor
and a fan.
The micro
maintains the heatsink at 110C with the fan stopped
or the micro turns off the heater and turns the fan on to cool the heatsink. The micro board was one I had from a previous
project. The same functionality could be obtained from a potentiometer, thermistor, comparator and MOSFET. There are several
designs on the web for small hotplates so I will refrain from publishing full
details of this one. The heatsink has a couple of
threaded holes added to secure the die mount to whilst soldering.
Now screw
the die mount to the hotplate, heat to 110C and tin the area where the die will
be mounted. Here you will need some flux, the solder and a q-tip.
Liquid
rosin flux – its colour distinguishes it from practically all other fluxes.
A small piece of indalloy 93C melting solder. I prefer to roll
it flat and cut it up into thinner strips lengthways. This makes it easier to
use if you decide to solder the die bonding wire.
Add a
little flux (0.2 of a drop – a syringe and fine nozzle makes dispensing the
flux easier. Then add the solder – build up a little bead – don’t worry at this
stage that it does not flow as quickly as normal solder.
Move the
solder around using a flux soaked q-tip – make sure it tins right up to the
edge.
Apply a
little more flux and wipe all of the solder bead off
the mount. Apply a very small bead (same size as the die) at the edge of the
mount.
If you are
not ready immediately to drop the die on it may be worth letting the mount
cool. The solder seems to oxidize a little if left for more than 5 minutes or
so at 110C. You can always wipe the solder off, apply more flux and replace the
bead if the surface of the solder starts to look a little dull. Now you need
the die.
I have
persuaded a die from the gel pack whilst it was over an anti-static bag *. I
was tempted to make this picture part of a spot the die competition ;)
- The way to make sure you
do not kill your die due to static discharge is to make sure that two
surfaces are at the same potential before transferring the die (or any
sensitive device from one surface to the other. This need not include
hundreds of $$ of anti-static mats, humidity control and operator
grounding wrist-straps although if you have them to hand it helps. If you
do not have all the anti-static gear the human body can be the perfect
shorting link to prevent discharge or the worst charged body to supply a
discharge from. The only difference is the order in which you touch
things.
- Example method for moving
device from conductive surface A to conductive surface B. Touch surface A.
Whilst holding surface A pick up device. Let go
of surface A. Touch surface B. Whilst holding
surface B place device on surface B. Let go of device. Let go of surface
B.
Now the tricky bit. How go you get the die onto the mount?
Non-mar tweezers ? …. NO …. The die WILL slip and get launched over
to the far side of the lab.
Vacuum
pick up tool?.. Fine if you have one (and a vacuum
head small enough) $$$ if you need to buy one
Or you
could try flux !!!
To a die
flux is the second stickiest thing on the planet. A sharpened wooden q-tip
dipped in a little liquid rosin flux will cause the die to stick to it. Gently
touch the cathode side of the die – you can then inspect the die under a
microscope to identify the cavity direction and output coupler side. The die in
this picture is a very small 1.5W 808nm multimode.
That blue/
purple reflection (808nm die only) indicates that this side is the output
coupler. The die supplier will be able to help you with the appearance of the
o/c coating for other wavelengths.
Remember I
said flux was the second stickiest thing on the planet,
well the surface tension on molten solder is the only thing with more grab.
Gently touch the anode side to the top of the solder bead and the die will be
grabbed by the bead and will remain there when the q-tip is removed. Make sure
the o/c is facing out from the edge of the mount.
The die is
down. Position it such that the o/c edge of the die is coincident with the edge
of the mount. A pair of sharpened q-tips make
positioning the die easier. A colleague or a footswitch to enable the mount
cooling cycle on your hotplate comes in handy here. Now is where the rosin flux
excels – it is highly soluble in IPA (iso-propyl
alcohol) or in any residue free flux cleaner (Electrolube
FLU Fluxclene is my favourite – great citrus smell
too.) Drop 20 drops of fluxclene onto the die. This
will remove most of the flux residue. This is merely to enable inspection of
the solder joint and clean the cathode side of the die.
Now form
the die bonding wire
The wire
is cut 2mm longer than the edge of the mount and the end is then formed into a hook
so that it is a rounded edge that will touch the cathode side of the die
Carefully
bend the wire down towards the die. By holding the wire away from the die at
the die end with a q-tip and bending it down towards the die with a pair of
tweezers at the PCB end it is possible to cause the wire to gently spring
against the cathode side when released. Practice this several times on mount
with no die attached. Now you have 3 choices for making the cathode connection.
1)
Conductive silver ink/paint may be transferred to the die/bonding wire junction
on the end of a sharpened q-tip. This works well and mistakes are recoverable.
If you miss and end up shorting anode to cathode by bridging the die the ink
can be removed by running the die under drips of fluxclene
before the ink dries. The finished joint is however brittle and may fail if the
mount is subjected to thermal cycling.
2)
Conductive silver epoxy. I have not tried this. Mistakes are probably not
recoverable. The epoxy will by more viscous so may be
harder to transfer onto the die/wire joint – but the finished result would be
more rugged.
3) Solder.
If you have managed to stress the bonding wire correctly and it is gently pressing
down on the cathode side of the die – you can solder the bonding wire without
disturbing the anode joint. Simply heat to 110C on the
hotplate. Apply a little flux near to the die – the flux will boil
initially and spread over the die. Touch the bonding wire with a thin sliver of
solder. Surface tension will cause a bridge of solder between the bonding wire and die. If the die moves it may be repositioned before
cooling using the q-tips as mentioned earlier.
End view
showing the o/c and bonding wire to cathode solder joint. (the
die is a 1.5W 808nm, the wire is 0.25mm diameter)
Now it
just remains to clean up boil in IPA / acetone for a couple of minutes –
although I have found dripping 100 drops of fluxclene
onto the o/c is adequate (rosin flux is highly soluble and visual signs of it
disappear after the first 5 drops. It is worthwhile checking the o/c under the
microscope to check for any signs of contamination.
It only
remains now to test the beast.
The
hotplate with no power connected makes a reasonable heatsink
for brief operation. The PSU is a standard adjustable lab one. It has been
checked to see that it behaves itself – no surprises due to dirty
potentiometers. The PSU is used in Voltage mode. There are two 1ohm resistors
in series with the die. Initially the die is shorted out. The PSU is enabled
and the o/p voltage set to 0v. The shorting link is removed and the voltage is
slowly turned up. The current was monitored and the voltage set to give a
current of 1A. This corresponds to an output power of just under
1watt. (I didn’t want to run the diode too hard because there was no airflow
across the vanes of the heatsink.
Now – one
of those Casix hybrid DPSS crystals later (plus a
homebuilt smps laser diode driver and an old
telescope as a collimator) and hey-presto – the laser pointer from hell ;)
The beam
terminates on low cloud.
Thanks and
appreciation to:
Bob from
the alt-lasers newsgroup – for moral support, pearls of wisdom and most of all the laser die.
Sam for
the excellent LASERFAQ
Usual
disclaimer – If you blind yourself it’s not my fault – use appropriate
protective eyewear – if in doubt don’t do it.
Best
Regards,
Robin