Most all lightwave receiver designs I have seen in electronic hobbyist books and Amateur Radio publications are truly pathetic. The best range I've seen claimed is a couple thousand feet, more often just a few hundred or even tens of feet. In my opinion anything under several miles does not merit the time of a Radio Amateur worth his or her salt. Bear in mind that for the ARRL VUCC laser award you have to do five two-way QSOs, one of those all the way across a grid square - something like 60 miles or more! A well designed receiver is essential. Photomultiplier receivers are well capable of this performance, but I believe that a properly designed photodiode receiver will also do. The photodiode receiver design I am presenting here works as well or slightly better than the photomultiplier receivers I used to set the current 57.7 mile HeNe laser DX record. This highly optimized design when bench tested side-by-side with one of my PMT receivers worked better. (See figure 2.)
Pd, the photodiode, is a EG&G Vactec VTP1188S. It is available as stock number 95F9029 from Newark Electronics for $2.36 (good photodetectors don't have to be expensive!) Its plastic package also serves as a lens. The package is about 0.3" in diameter and 0.3" tall. The typical sensitivity is 0.55 A/W (Amps output for Watts of light input). This does not mean that the device will put out 0.55 Amps under a 1 Watt light bulb - photodiodes don't generate that much current. 0.55 microamps for 1 microwatt is more realistic. The 0.55 A/W figure is typical for any silicon photodiode. The sensitivity of silicon photodiodes does not vary much from some standard figures due to the laws of physics. The sensitive area of the actual VTP1188S photodiode chip is 11mm2. Much smaller area photodiodes are available and result in higher speeds due to less capacitance. But the smaller the active area size is, the more critical the optics are. You have to get as much signal to efficiently illuminate the photodiode as possible, which would be more difficult if the chip is very tiny. Much larger area photodiodes are also available. They also offer us no advantage. They cost a lot more, don't put out any more Amps/Watt, and generate more noise. The VTP1188S is by no means the only photodiode that would work for our application, but it is an excellent performer, readily available, and inexpensive.
Rl, the load resistor, serves mainly to give capacitor Cc a discharge path to ground. Its value is not critical, but best performance was found at 50 megohms, plus or minus 10 megohms. This is a much higher resistance than most Amateurs, or even electronic professionals are used to dealing with. You will not find a resistor value above 10 megohms from any of the normal parts sources, such as Newark or Digi-Key. Specialty resistor manufacturers such as Victoreen make the high resistance, low capacitance resistors commonly used in photodiode transimpedance amplifiers. Values are available in the Gigohms! Unfortunately, these resistors are expensive ($5 to $15) and high minimum orders ($100) are required. Fortunately, the circuit shown here is not too critical and a string of 10 megohm resistors will work just fine. Cheap 10 megohm resistors are available from Digi-Key as their part number 10ME-ND, five for 28 cents, or 100 for $4.60. Strings of these resistors should be used for Rl and Rf. I have tested this circuit with both Victoreen resistors and strings of the Digi-Key resistors and saw no difference in the resulting signal-to-noise ratio.
Cc, the coupling capacitor, serves to let AC signals from the photodiode through to the amplifier while blocking DC signals. This is a very important feature of our circuit. Capacitive coupling is very unusual for transimpedance photodiode amplifiers - they are normally DC coupled. After working for over 2 1/2 years for one of the top photodiode manufacturers I had never seen this done! If you don't capacitively couple, the DC signals generated by sunlight, moonlight, incandescent light, etc., will totally overload the amplifier. The value of Cc is not at all critical. 0.1 uF works fine, but 0.01 uF and 1.0 uF work fine also. It should be a reasonably good capacitor, that is, not leaky.
Rf, the feedback resistance, is another high value resistor. 50 megohms in the form of a string of 10 megohm resistors works fine. Again, this value is not critical. This is a very important part of a transimpedance amplifier. The output voltage of the amplifier is determined by multiplying the input current times the value of this feedback resistor. This amplifier will put out 0.5 Volts (a very large signal) with an input of only 10 nanoamps (10-8 Amp)! The photodiode will generate 10 nanoamps with a light signal of only 18 nanowatts (0.000000018 Watt)! The "power" of a transimpedance amplifier is incredible.
Cf is a very small capacitor, in the order of 1 to 10 picofarads. Without it a phenomenon known as "gain peaking" will occur. At some relatively high frequency the gain of the amplifier will peak above what it is supposed to be as determined by the feedback resistor. Cf also is used to reduce the normal amplifier gain at high frequencies. Gain outside of the desired bandwidth, which is 300 to 3000 Hertz, the standard communications audio bandwidth, just contributes to unwanted noise. This capacitor can be adjusted for the best sounding signal when receiving a very weak signal. A better way to adjust it is to connect a sensitive AC voltmeter to the output of the amplifier. Make a signal source by powering a LED from a audio signal generator set to 1000 or 2000 Hertz. Shine the signal into the photodiode and adjust the physical arrangement and the signal generator output voltage to get a very weak signal ( a few mV). Turn the signal generator on and off while adjusting the capacitor for the largest ratio of signal-plus-noise (signal generator on) to noise (signal generator off.)
IC1 is an op amp. The "input noise current" rating of this op amp must be very low for this kind of circuit. Fortunately, there are a number of good op amps that aren't too expensive. I began by testing nine op amps ranging in price from $1 to $20. They included the Burr-Brown OPA627 and OPA111 - premier op amps for photodiode transimpedance amplifiers, the PMI OP07, the Analog Devices AD743, Linear Technology's LT1028, LT1037, and LT1055, the new National LMC6001, and even the lowly 741. The input noise current for some of the better op amps tested was rated under one femtoamp, that's 0.000000000000001 Amp! This rating is important as the current from the photodiode is going to be amplified 50 million times by our circuit. In some tests op amp A did better than op amp B, in other tests just the opposite. A transimpedance amplifier circuit is deceptively simple, there actually are dozens of nuances and a lot of interplay. The testing was done by measuring the signal- plus-noise to noise ratio in a 300 to 3000 Hertz bandwidth while receiving a weak signal from a 1 KHz square wave modulated red LED. The results in a nutshell: the OP07, OPA111, OPA627, AD743, LT1055, and LMC6001 all performed well. The losers were the LT1028 and LT1037 - presumably their extremely high bandwidths contributed to the noise. The 741 did surprisingly well, but there are much better choices for just a few dollars more. I have settled on the AD743JN because of its excellent performance and its price and availability ($5.53 from Newark). The Burr-Brown parts are top performers but are very hard to get and are expensive. An excellent alternate to the AD743JN would be the LT1055CN8 at $3.04 from Digi-Key.
When designing a circuit for absolute lowest noise even the op amp supply voltage needs to be taken into consideration. The high quality op amps we are dealing with are normally spec'd by the manufacturer for operation at +/- 15 Volts. Voltage causes current, current causes heating, heat causes noise. I measured lower noise at +/- 9 Volts, which is also convenient for battery operation.
The above covers all of the critical components in the laser receiver with the exception of circuit layout and shielding. Those used to dealing with RF and microwaves may think they have seen it all, and that there can't be much to a circuit that just amplifies audio. Nothing could be farther from the truth when you are dealing with gains in the tens of millions and currents in the picoamps! It is important to use a printed circuit board with plenty of ground plane. Keep the input well isolated from the output. Keep leads short. Keep the board and components clean and de-fluxed else picoamps will leak across the insulation. This much gain at audio frequencies will result in 60 Hertz pickup, so shielding is needed. I built the transimpedance amp alone on a 4" diameter round circuit board with the photodiode in the center. On the back of the circuit board I soldered a 3.4" diameter Coffee-mate(r) jar steel lid. I did the same on the front of the circuit board using a lid with a 0.5" hole in the center for the photodiode to look through. The finished round board fits into a 4" diameter plastic pipe along with a 4" diameter lens. The filter and speaker amp are on a separate round circuit board.
The filter is a TOKO THB111A 300-3000 Hertz active bandpass filter. It is available from Digi-Key for $21.39 (their part number TK5425-ND). It is 30 dB down at 100 Hertz and 7.9 KHz. Its main purpose is to attenuate 60 and 120 Hz optical interference street lights, room lights, etc. It also attenuates non-optical pickup of 60 Hz electric fields. And it also limits the upper bandwidth to reduce noise, which would be heard as a hiss in the speaker.
IC2 is the renowned National LM386 audio power amplifier. The capacitor across pins 1 and 8 sets the chip to its maximum gain of 200. This chip is available in three versions. The LM386-1 is optimized for 6 Volt operation at which it will put out 325 mW of audio power into an 8 ohm speaker. The LM386-3 is optimized for 9 Volt operation at which it will put out 700 mW. The LM386-4 is for 16 Volt supplies and will drive a 32 ohm speaker to 1 Watt. The -1 and -3 versions can be run as high as 15 Volts. Only the -4 requires the 0.05 uF capacitor and 10 ohm resistor on pin 5. These components can be left off circuits using the -1 and -3 chips. I strongly recommend not running the LM386 off of the same 9 Volt battery used to power the transimpedance amp. The LM386 draws significant current which may disturb the incredibly sensitive photodiode amplifier. It would be ideal to use a LM386-3 and power it from a separate 12 Volt NiCad or Gel Cell. Alternatively, a separate 9V alkaline transistor radio battery would do.
The Edmund Scientific folding stand magnifier lens is ideal for the receiver antenna (their model G38,599 at $9.50). The 4.3" diameter glass lens with a 8.5" focal length can be popped out of the plastic stand. Its diameter is perfect for mounting inside of a length of 4" ABS "DWV" pipe. This pipe can be purchased from a hardware store. It is black ABS plastic with a 4" inside diameter and a 4.5" outside diameter. The lens can be pinned between the end of the pipe and a plastic pipe coupling, also available from the hardware store. The same mounting method is used for the detector/transimpedance amp circuit board and the filter/speaker amp circuit board. The speaker and on-off switch can be mounted on a matching pipe cap. The entire laser receiver, including batteries, will fit in a neat and inexpensive 4.5" diameter by 2 foot long housing. A second pipe cap can be used to cover the lens end of the receiver during storage and transport. (See figure 3.)
A way to fasten the round pipe to a tripod is needed. The receiver can be mounted to a metal plate with a 1/4-20 threaded hole in it. 1/4-20 is the thread on a camera tripod screw. "Minerallac Straps" can be used to mount the pipe to the metal plate. Minerallac straps are "U" shaped metal clamps that snap around pipe and allow it to be fastened down by a single bolt. Electrical supply houses or very well stocked building supply stores will have these. Two straps can be used to fasten the receiver to the metal plate.
Obviously, the atmosphere is a good medium for transmitting visible light. Atmospheric transmittance can be considered excellent from 500 nanometers to about 950 nanometers. Meteorologically speaking, on an "exceptionally clear" day, visibility is considered to be 50 to 150 kilometers. It is not quite as simple to predict just how far you can communicate on a beam of light. There is a formula though:
Range = meters.
Po = The power output of the laser in watts.
Ar = The area of the receiver lens or mirror in square meters.
To = The transmissivity of the receiver optics.
Ta = The transmissivity of the atmosphere.
Pt = The threshold power of the receiver system in watts.
D = The laser beam divergence in radians.
Lets try some practical figures.
Po = 0.002 watt (2 milliwatt), typical for a small HeNe or Diode pen pointer laser.
Ar = 0.008 meter (4 inch diameter lens).
To = 0.84 (84%), no filters, two typical 8% reflections off of the lens surfaces.
Ta = 0.9 (considered a clear day).
Pt = 5 x 10-12. Approximate Noise Equivalent Power of the EG&G Vactec
VTP1188S photodiode in a 300-3000 Hz bandwidth.
D = 0.0015, (1.5 milliradians) for a typical Helium Neon laser or Diode laser pen pointer.
Range = 1,017 kilometers, or 633 miles!
Is this figure realistic? Not taken into account is the noise that the receiver circuitry adds to the signal. If we make Pt ten times worse we get a range of 328 kilometers, or 204 miles. Probably more realistic, but still fantastic. Atmospheric distortion and the problem of finding two points this far apart and still line-of-sight would likely be the real limiting factors.
Note the importance in this range formula of keeping the laser beam divergence small. Halving the divergence doubles the range. Small divergence is a quality to look for in a laser. Note too the importance of antenna area. A 2" diameter lens gives half the range than a 4" diameter lens.
I was a member of the Los Padres Microwave Group in the late
70's and early 80's. This was a West Coast VHF/UHF/Microwave
contest group founded by W6OAL, WB9KMO, and myself. My job
was to provide and operate the microwave-and-higher contest
gear. After two years of one-way testing we completed a two
way HeNe laser QSO with the K6MEP contest group in the June
1979 ARRL VHF QSO Party. We were located on Frasier
Mountain, Ventura County, Santa Barbara section. The K6MEP
group was on Reyes Peak, 15 miles away in the same county and
section. The equipment at K6MEP was a 3 mW Spectra Physics
HeNe laser owned by W6OAL. I modified the laser by adding a
chopper modulator and a tripod mount. I built the receiver
around a 931A PMT with a 3" lens. The gear the Los Padres
group used was a 4 mW chopper modulated HeNe laser I built
around a surplus plasma tube. The receiver was a 931A PMT
with a 10 inch Fresnel lens. The choppers were made from
small DC motors and tin can lids. The chop rate provided 1
KHz modulation. The resulting audio modulated beams were
Morse code CW modulated by interrupting by hand. This proved
to be one of the more difficult aspects of the contact as the
hand interruption had to be made in inverse Morse code. It
took a little getting used to! By far the most difficult
problem was aiming the lasers. First, finding the other
party is difficult at 15 miles. Second, keeping the laser
aimed is equally hard. With a one milliradian beam
divergence, discounting any path distortion, the beam at 15
miles away would be 79 feet in diameter. Moving the laser by
just one degree would move the 79 foot diameter beam at the
far end by a whopping 1382 feet! Contact was established by
slowly sweeping the lasers back and forth across the
suspected target point. Two meter liaison was essential.
The person at the receiving end would signal on 2M when he
saw a flash of red light. This was not sufficient to stop
the beam in time but it did give the people at the
transmitter end a good idea of where to sweep at a slower
rate. Finally the beam would be somewhat steady at the
receiving end. At night it was very bright red. Tests
during the day were also successful, but the beam did not
appear as bright and looked somewhat pink due to the
competing sunlight. All equipment was mounted on standard
SLR camera-type tripods. Heavier tripods, such as used for
video cameras, would be highly desirable. Laser two way
contacts were achieved by the Los Padres Microwave group with
the same equipment a half dozen times since the 1979 contact,
although it wasn't until the June 1982 ARRL VHF contest that
the ARRL would grant us contest credit.