Professor Mark Csele's Homebuilt Lasers Page

DPSS Lasers
While a DPSS appears to be a relatively simple device, building one that is truly stable requires a bit of
engineering. This page describes the conversion of a simple, inexpensive green laser pointer into a lab-grade
laser.

Abstract
The heart of this DPSS laser is an inexpensive green laser pointer. Such pointers are
actually diode-pumped solid-state (DPSS) lasers consisting of a small vanadate
(Nd:YVO4) laser (operating at 1064nm) pumped by an 808nm diode laser and frequency
doubled to produce 532nm output. The actual vanadate crystal and KTP frequency
doubling crystal, along with the HR and OC mirrors, are built as a small one-piece
composite unit under 5mm in length. Radiation from the pump diode (at 808nm) passes
through the HR to pump the vanadate (the HR is a wavelength-selective coating which
transmits 808nm while reflecting 1064nm). [1]

A close-up of a green laser pointer DPSS as seen through a microscope. The

vanadate and KTP crystals are constructed as a single unit with the HR and OC
reflectors deposited directly on the surfaces of the optical components. The
assembly is pumped directly from a diode (usually around 200mW output at 808nm)
which passes through the HR.

Introduction
There are two basic problems with a simple laser such as this: the wavelength of the
pump diode varies with temperature, and the phase-matching characteristics of the KTP
doubling crystal also varies with temperature. In the case of the pump diode, all diodes
behave in a similar manner with wavelength shifting towards longer IR wavelengths as
temperature increases (of course, as drive current increases, temperature does as well).
Unfortunately, vanadate (as well as YAG, and other neodymium laser host glasses) has
sharp absorption peaks as so allowing the pump wavelength to vary over a span of a few
nanometers can move the wavelength from one of great absorption to one of less
absorption resulting in a decrease in output power. In decent commercial lasers, diodes
are provided with a separate Pelltier thermolelectric cooler (TC) and temperature sensor
(usually a thermistor) to keep temperature constant regardless of drive current or ambient
temperature.

Phase-matching of frequency-doubling crystals is, similarly, temperature sensitive. Again,

high-quality commercial lasers use a separate thermoelectric cooler to stabilize the
temperature of the crystal since a variation of only 0.2C can result in a decrease in
efficiency of over 60% (at an anomaly point ... normally variations with temperature are
not quite this large).

A commercial DPSS laser, a Crystal Laser, is seen here with the cover removed to reveal
key elements. The pump diode is mounted directly on a TC with a thermistor attached.
Pump radiation then passes through lenses and prisms to shape the beam and pumps a
vanadate/KTP laser mounted in a copper block with a separate TC. The output beam is
then sampled by reflecting a small portion from a piece of glass onto a photodiode.
Ultimately, output power controls pump diode current resulting in stable optical output
power. TC's ensure diode wavelength is stable and the KTP material remains phase-
matched for efficient second-harmonic production.

Components of a typical commercial DPSS system (as explained in the text).

Building The Device

Lacking the ability to add two TC's to the unit, the easiest solution is to use electronic
feedback such that diode current is adjusted to maintain constant output. To allow this
scheme to work (a) the entire laser must be heatsunk quite well to prevent heat buildup
and (b) the laser must be run at well below the maximum power output allowing suitable
range for adjustment. With respect to heatsinking, one prototype laser literally cooked
when operated at rated current for over five minutes (the casing was only about 45C at
the time). The new laser was mounted into a large, solid, block of aluminum. The
aluminum was drilled to a diameter of 9/16" allowing the brass cylinder containing the
laser (stripped of the black housing), to fit snugly with thin shimstock metal and thermally
conductive grease inserted to fill the remaining gap. A set screw was tapped at right
angles to secure the laser into the block.

As for output control, a basic feedback circuit was employed almost identical to that used
with basic laser diodes however with the feedback taken from the output beam itself. The
circuit itself is seen in the figure below:
Schematic diagram of the DPSS driver. It is almost identical to the simple driver used with a laser
diode alone, in which the laser diode and photodiode are mounted in the same package (the
photodiode receives a small amoung of radiation emitted from the HR of the diode). In this case,
feedback is obtained from the actual output beam and so represents actual laser output (i.e. after
frequency doubling in the KTP crystal).

The incoming supply is regulated to 2.90V which sets the maximum output (as well as
maximum current for the laser diode) - the laser assembly itself contains an integral driver
designed for maximum output (of about 4mW of 532nm green radiation) at at voltage of
3.00V (and a curresponding current of 375mA) ... the chosen value of 2.90V results in a
maximu current of about 350mA. Light from the laser assembly is first filtered to remove
residual 808nm and 1064nm IR (of which there was over 12mW measured) using a
narrow-band dielectric filter. The resulting beam then passes through a piece of quartz
with the small reflection incident on an FPT-100 phototransistor. The prototypical optics
are seen
here:
The prototype of the DPSS system. The laser was mounted in a portable vice, and
optical components on "third hand" alligator clips normally used while soldering. The
beam path is outlined in green showing how a portion of the beam is split off for
sampling by the phototransistor which is mounted on the breadboard along with
components for the driver circuit. A commercial power meter is seen in the left
corner, used when calibrating the circuit.

When first turned on, Q3 is not conducting (since no light is produced yet) and so current
flows through R2 & R3 causing Q2 to conduct. Collector current from Q2 flows into the
base of Q1 causing it to conduct turning on laser diode D1. As the laser operates, light is
produced and falls on Q3 causing it to conduct. Base current through Q2 is hence
reduced (since it now flows through Q3) hence reducing base current through Q1 as well.
In this manner, the system reaches an equilibrium value of optical output power. It must
be set to a relatively low value to allow a decent control range and so R3 is set for an
optical output power of 400μW. If a higher power is chosen, current will increase until the
maximum value of 350mA resulting in a decrease of optical output power below the set
value.

As with any system, a transfer function exists and gain may be set, crudely, by setting the
amount of light from the laser which is fed-back to the phototransistor. On the prototype, it
was found that insertion of a neutral filter between the quartz plate and the phototransistor
reduces the overall feedback of the system. The values given were found to be adequate
for small optical output powers however at large powers the system tends to oscillate and
so a filter must be inserted immediately before the phototransistor as per in the prototype.
The finished version was designed for power levels of under 1mW - underrating the laser
this much allows for a large operating range (ambient temperatures may reach up to 30C
and still allow operation).

The finished DPSS is seen here mounted in the aluminum block. The beam path was
drilled through the block first and a large hole in one end for the laser. In the middle of the
block two areas were milled to hold the green filter (first) and the quartz beam pick-off
(with FPT-100 phototransistor). Two holes on the bottom were tapped to hold the laser to
an optical bench.

The completed DPSS system. The laser is entirely enclosed in the large aluminum
block used as a heatsink with only the set screw visible on top. Machined portals
hold the dielectric filter, beamsplitter, and phototransistor. Normally, a cover conceals
these elements. The entire laser is normally mounted vertically via two tapped holed
on the bottom.

References:
[1] Fundamentals of Light Sources and Lasers
Mark Csele

John Wiley
and Sons, 2004 (ISBN 0-471-47660-9)
 [2] Crystal Lasers 532nm DPSS laser

You are visitor # to this page

Diagrams (except those as noted), text, and photographs Copyright © 2012 by M. Csele 2004. Portions are excerpted from Fundamentals of Light Sources
and Lasers (ISBN 0-471-47660-9) and are copyright © 2004 by John Wiley & Sons.
Reproduction forbidden unless written permission granted by John Wiley & Sons.