Bellpush

Not much progress on the bellpush over the past week because most of my attention has been focussed on another non-concertina-related project. Yesterday was a day off though, so in the evening I managed to stop thinking about the other project for long enough to tackle the problem of the spring that provides the button resistance.

My original plan had been to use a single strip of brass as both the button spring and one of the switch conductors. This turned out to be impractical though because the button has a vertical travel of 5mm and there isn’t enough room inside the box (once you’ve allowed room for mounting screws and electrical screw terminals and things) for a really long spring strip. A short strip probably wouldn’t be able to flex by 5mm without permanently deforming and even if it could, it would likely result in fatigue failure before long. I experimented with folding a brass strip back and forth in a zig zag to increase its effective length, but it looked like it would take up far too much space and probably still wouldn’t work very well.

So the solution I eventually came up with was to separate the two functions. Make a concertina-style spring from coiled phosphor bronze wire with 5mm travel, and a separate brass strip switch that only has to move by 1mm or less at the bottom of the button travel. I haven’t yet made the switch but I have made the spring. It took me four attempts before I developed a shape I was happy with, using a hole in a block of scrap wood as a test rig. It has more turns than a standard concertina spring because of the large amount of movement relative to its length.

Here it is installed in a recess chiselled into the back box. It’s in a slightly weird location because of where I intend to put the mounting screws and switch contacts:

I won’t bother posting a photo with the top installed because it doesn’t look any different on the outside! The action feels pretty nice though.

Following my earlier, unsatisfactory, attempt at making a pair of captive nut plates for the bell push by soldering stainless steel nuts onto thin brass sheet, I bought a set of M2.5 taps and a piece of much thicker (3mm) brass sheet for my second try.

I stained the brass with a permanent marker to make the lines show up more clearly, and marked it out using a steel ruler and scriber under magnification:

I also upgraded from a cheap, dull wood-cutting countersink bit to a small good quality (Dormer) HSS three flute metal-cutting countersink. The resulting countersunk holes for the wood screws that will hold the plates to the back box are way smoother, cleaner and more accurate as a result. I tapped the M2.5 holes by hand using Trefolex tapping compound (I bought that small pot at a Model Engineering exhibition about 20 years ago and it’s still half full!).

A brief aside about the heads of the visible machine screws/end bolts that hold the top on. This is what the brass bolt heads look like on my antique Lachenal:

It turns out that head style is called a slotted fillister (I’d not heard that word before I started searching through screw catalogues, and it isn’t in my computer’s spell checking dictionary). I couldn’t find any suitable ready-made M2.5 stainless steel slotted fillister head screws for sale (I much prefer the appearance of slotted heads for this application), so I instead got some cheese-head screws and domed and polished them in the lathe. The next photo shows them before and after modification. Not quite right but I think they look reasonably good.

The final photo shows the two finished nut plates. I’m far happier with how these turned out than I was with my first attempt. The brass wood screws are bigger than necessary, but I happened to have a packet of them sitting around and there is enough room to use them. I don’t think there’s much risk of them pulling out of the oak!

Tonight I had my first experience of turning acetal (often known by the trademarked name Delrin). It’s an engineering thermoplastic that is relatively expensive but is designed to machine well. I’ve turned other kinds of plastic in the past and usually struggled to get a good accurate smooth finish – they are usually soft and gummy and don’t cut cleanly. Acetal in comparison was a joy to use. With a sharp HSS tool and a high spindle speed, it cuts almost like it’s not there and leaves a lovely smooth finish straight off the tool. You can even take really fine cuts without it rubbing and melting. Lovely!

The part I was making was the sleeve for the bottom part of the bell push button. It will act as an electrical insulator to isolate the switch contacts from the metal part of the button. It also needs to slide smoothly into a hole bored in the wooden backing box underneath the contacts.

Because I have read that most types of glue don’t stick well to acetal, I designed the interface between the two parts of the component in a slightly unusual way. The stainless steel pin was slightly flared towards the end (actually caused by deflection from cutting forces when I turned it, but I expected this to happen and deliberately didn’t do anything to prevent or correct it). Also, the hole in the acetal was drilled 0.5mm larger than the diameter of the pin for roughly the bottom 90% of its length. The combination of the two produced a small tapered gap between the two which, when filled with epoxy resin, should act as a mechanical fixing that will prevent them separating even if the glue doesn’t bond to the plastic at all.

The finished two-part button: