Experiments in Laser Comms DX Notebook

30/12/99: Improved receiver with shielding:

Noting information from other experimenters I decided to mount the photodetector completely in a tin can for shielding.   A handy Christmas tin of chocolate biscuit sticks presented itself.   A hole was drilled into the bottom of the tin and the photodiode was soldered to the outside bottom with the photosensitive area exposed through the hole.

This time an Rf value of only 10 Megohms was used as the first receiver was very microphonic with the seven resistors in mid air.   A small salmon tin was used as a capping shield over the photodiode (although it is not very light-tight at the moment).   The arrangement completely eliminated audible hum when connecting up to the PC desktop Soundblaster card.   However, recording a WAV file and analysing it using my narrowband software (8 megasamples) showed that even with the photodiode unpowered there was still signals coming from (presumably) the PC.   The output of the signal was also too low.  I needed to increase the level into the Soundblaster to mask as much internal noise as possible.

Once again, using information from others, I added a LM386 amplifier running at a gain of 200 inside the salmon tin shield.   This showed a marked improvement and the external noise was reduced to such a low level that it was necessary to switch off all lights and all monitors so that the small amount of external light from these sources would not produce any saturating peaks in the spectrum.   Below is a graph of the noise connected to the desktop PC from the improved receiver.


The spectrum was lowpass filtered at 2000Hz after acquisition.   The low frequency rolloff is produced by the 470n/10K coupling.   Note the small peak on the rising edge of the low frequency side.  This is at 100Hz.

Previous Page  Next Page

Back to Notebook Index