Whilst I do have a 6m long 3" diameter light pipe fitted up in the attic that I can do receiver sensitivity testing with, it is inconvenient to use (lugging test gear gear up there, and sitting in the dark during measurements, particularly during periods when there's an active wasp nest up there too). Recently, an old Wickes aluminium lined toolbox became surplus to requirements, and it was noticed that the lid closed in a way that produced quite a good light seal - the two halves coming together in a tongue-and-groove fashion. Could this be used as a suitable test jig housing?

The box has timber walls, covered in a reasonably strong, castellated plastic foam, as are the dividers, which split the case into six usefully sized compartments. A perceived problem was that the light source would be very close to the receiver module under test, and with such a high input impedance, was likely to pick up the drive to the light-source LED. This was solved by having a remotely sited LED feeding a length of 2mm light-pipe which was threaded through several holes in a few of the divider walls, as above. To obtain a more even illumination of the test receiver, two pieces of optical diffuser sheet* were fitted in front of the open end of the light pipe (one each side of a large hole machined into the relevant side wall. This proved very effective...

* a scrapped PC monitor is likely to contain a few layers of suitable diffuser sheet

Driving the light-pipe from a standard domed 10mm LED turned out to be quite easy to do. A hole was drilled dead centre down through the top of the LED package, stopping well short of the LED element itself. This arrangement is inherently reliable, though a spot of super-glue was used to keep the light-pipe from being able to be pulled out from the (tight fit) hole. Very little drive power is required to light up the LED adequately, and I tend to just use the AF output from an elderly HP8640 signal generator. No series resistor is fitted (but perhaps it should be, to avoid the LED turn-on knee being as sensitive to drive voltage as it currently is). Since the LED is only just being pushed into conduction, the light output is rich in harmonics, as can be seen opposite. The LED, with it's first 10mm or so of light-pipe were then resin-potted into a miniature plastic box. This gave strength to the assembly and its connections, and also absorbed all of the escaped light. Where a selection of emitter wavelengths are required, the light-pipe can be cut at some point along its length and a joint created to allow several potted assemblies (red, IR and UV, say) to be made up and fitted as necessary. The large (2mm dia) light-pipe was found to be much more efficient than conventional light fiber, though it is obviously less flexible. Ground bonding It was found that despite the all-metal hinges, electrical contact between the lid and the main box was not reliable, and a braid strap was needed to bond the two parts together.. Also, the box and lid need to be connected to the ground side of the receiver under test, to ensure that there is no electrostatic pick up.

Tidy ups

The flying lead connections to the outside world are a nuisance when storing the box, so it would be better to fit a connector panel instead.

Conclusions

The test enclosure has been a real success. It has proved completely reliable and is quick to set up. All my measurements are relative (against a 'gold standard' unit) and provides a simple way of ensuring that new receivers (I must have built 30 or so of the Finningley transceivers) are within a couple of dB of having the same sensitivity.