Continuing the restoration of the Edison-Bell Type B crystal wireless set, a next step was to find replacements for the missing fittings. A British product of nearly a century ago, it naturally does not use any metric standard-size fasteners or fittings. This made it a bit of a puzzle to determine the right sizes - and then a puzzle how to get these today.
From playing with 1920s vintage Meccano (the Electrical Outfit), had learned that BA-threads were common in electrical equipment of the era. Checking with a Meccano part 304 (6BAx1/2" screw) confirmed that e.g. the mounting holes for the lamp in the panel were indeed 6BA. Measuring the remaining fittings on diameter and threads-per-inch confirmed all fittings to be 4, 5, and 6BA thread. (To measure t-p-i, firmly roll the screw in a bit of paper or cardboard to make clear marks and then measure ~10 marks to get the value.)
Again the internet-age allowed ordering of individual BA-sized brass screws and nuts. These duly arrived from Britain (despite current conditions!) and were quickly nickel-plated like the originals. One 4BA screw first had the cheesehead filed down to a flat top - from the tool-marks on the original contact-studs, this was exactly what the Edison-Bell must have done in 1923.
Now all fittings cleaned and missing fittings replaced by new - the panel looks very presentable again:
No more unsightly empty holes. Also underneath, the missing screws and nuts are added - more complete, but still without wiring. (The variometer is still assembled wrong in the below image - that one turned out to be quite the puzzle.)
The battery-studs and lamp-holder-screws are 6BA, the variometer fittings and mounting screws are all 5BA, as is the detector mounting screw. All terminals and studs are 4BA. (In case anyone has one of these radios and needs to replace a fitting.)
The variometer (or variator?) is a rather clever set of parts by Edison Bell - they also sold it separately for building into other sets and it can be assembled in several ways. The mounting brackets can e.g. be mounted any side - also it took me a while to work out the correct, intended assembly of the circuit for the coils.
Both the outer and the inner sphere have 48 windings, at ~65 and ~70mm diameter. These should be in series, allowing the centre sphere to rotate and work with or against the overall inductance. A surprising element (to me, at least) is how the outer shell coil-halves are connected. This is simply a pressing of the bare wire agains the small nut in the other half that clamps that other half's wire.
The two pivot bolts that hold the inner sphere are the leads into and out of its coil. The brackets on those bolts allow routing the circuit to the outer coil terminals. Took a bit of puzzling, this.
One other item was the dial scratching the panel. It turned out to not be warping of the dial (as originally suspected), but a manufacturing defect in the top knob. The middle insert-nut must have been placed slightly askew in the mould, before over-moulding. This was mitigated by shimming (waxed paper) between the top-knob and dial.
The markings in the dial were then re-done with white wax - following again the advice from the excellent book on typewriter repair by Teege.
Radio repair with thanks to the Typosphere and Georg Sommeregger :-)