A VCSEL-PIN Transmitter

Wes Hayward, w7zoi, 5Oct2017

Here's the other side of the June, 2017, QSL card from Bob, N7FKI.    Not many cards will include the complete schematic of the transmitter, but this one does.   Read on, for it's an interesting transmitter.

n7fki qsl back

An initial look at this schematic would seem senseless.     The circuit seems to contain no more than a couple of diodes.   But is this not the most rudimentary model of a bipolar transistor?   The physics differs in this version of the bipolar transistor, but many circuit properties are identical.     The OPC332 is a Vertical Cavity Surface Emitting Laser diode.   The optical output of this diode is at a wavelength 850nM.   Laser diodes of this sort are more efficient than a LED and also have a very wide input bandwidth.   This part is manufactured by TT Electronics, Optek Technology, Inc.

The other diode, the OP905, also from Optek, is a PIN diode.   PIN is an acronym for a PN diode structure with an intrinsic layer between the positive and negative doped regions.   The OP905 is optimized for optical detection.   Notice in this circuit that the diode is reverse biased, just as the collector-emitter port of an NPN transistor is reverse biased.    (Incidentally, the Laser diode, the OPC332, is also a PIN.)

The base of the synthesized NPN transistor is biased with a DC base current of about 4 mA.   The measured equivalent beta is 0.1, so the DC current in the collector PIN diode is about 0.4 mA.   A beta less than 1 does not mean that oscillation is not possible.    The spec sheets for the two diodes show wide bandwidth, suggesting that an experimenter could achieve oscillation at much higher frequency.   The circuit shown is a 7 MHz tuned collector circuit with a crystal in the feedback path.    The tuned circuit presents a suitable, high impedance to the output PIN diode.   

Bob's transmitter generated an output of one tenth of a milliwatt.   While hardly a DX machine, the signal from this transmitter is easily copied at my place.   Bob is about a mile from me, over a ridge.   He uses a back yard vertical and I used a dipole.    We have made CW contacts with as little as a microwatt.

Notice the construction detail in Bob's schematic.   The two diodes are mounted in a hole in a block of hardwood, with the diode ends touching each other.   This structure provides tight optical coupling between diodes while also guaranteeing that the relatively intense output of the Laser diode cannot be "seen" by the user.   850 nM is just outside the human visible wavelength range.     (See the Wikipedia discussion of Laser diodes.)   The figure below is found on the Optek data sheet for the OPC332 diode.   
Warning  It is important to observe safety precautions when using these parts.

If any of you should duplicate Bob's circuit, we would love to hear about your results.    The diodes are not expensive.   Most of the parts, including even the crystal, are available from Mouser with a total cost of around $10.