[Laser] Daytime dx plus a bunch of questions

[steve kavanagh]

A belated thanks to those answering my question about long distances in the daytime.  It would seem that my 70mm lens receiver (schematic at http://www.qsl.net/ve3sma/LaserSchematics.pdf - the photodiode being a small area one out of a portable CD player) is probably working ok with its daytime dx record of a solid signal at 250-odd metres on a single 20 mA LED transmitter....Let's see...

3 times the receiver aperture => 3 times the distance
3 W/45 mW peak power = 67 times the power => 8 times the distance
1 degree TX beamwidth instead of 15 degrees => 15 times the distance

.....for a total distance multiplier of 360...or 90 km (ignoring atmospheric loss)...which seems to jive with the results others have had.

At night I have just barely heard the 500 Hz square-wave-modulated single LED (15 degree approximate beamwidth) on scintillation peaks at 3.7 km (with minimal audio filtering - just as shown in the schematic)...where it was plainly visible (in the countryside), if not exactly bright.  A bank of 8 such LEDs was comfortable copy.  Is that what would be expected by those of you with more experience (bearing in mind my fairly small receive lens) ?

I got a hankering to try an LED (one of the 20 mA types...so a small area chip) as a photodiode, so I am in the process of building another receiver and trying to get my head around the preamp design.  It's been a few years since I gave this subject much serious thought so I am pretty rusty on the theory.  I'll probably go with more-or-less the same circuit as the referenced schematic unless someone can convince me it's a bad idea.  For now, I am interested in tone modulation in the hundreds of Hz range (morse, maybe some digital modes). I am hoping the narrow band nature of LED detectors may help with daytime operation, too. 

The AC coupling was to eliminate daytime saturation of the preamp....the resulting 66 Megohm bias resistor appears to have a fairly small noise impact if my recent modelling attempts are correct.  If they are, then it appears than the equivalent input noise voltage of the JFET is not overly critical for low capacitance diodes.  The JFET current noise might be if the gate current is anywhere close to the typical 1 nA spec, but it is likely to be much less, it would appear. Which means selection of a JFET may not be very critical to noise figure.  

I have MPF102, J112, 2N4393 (lower gate leakage current and voltage noise spec than most), 2N5458, and various VHF types (J309, J310, 2N4416A) available and some 2SK170's on the way.  Just noticed the annoying upturn at the 1 Megohm end in the noise figure curve in the 2SK170 (Toshiba) datasheet....and I can't explain it theoretically...but it worries me with respect to using those at tens of Megohms impedance, and their low voltage noise may not be useful while the high capacitance won't help.  Has anyone any advice on device selection....or confirmation or disagreement with my tentative conclusions ?

73,
Steve VE3SMA