making a double pole single throw switch

[Brianc]

Scroll down for message...

[Brianc]

I've noticed that there are quite a few members who have an interest in metalwork so I thought that I would recount (retrospectively) how I solved a problem set me by a friend, Tony, who restores player pianos and specialises in reproducing pianos, especially of the Duo-Art type. These reproducing pianos are surprisingly sophisticated when it comes to control. They are “powered” by vacuum or should I say “lack of air” as the motive power to strike the strings, operate the pedals and transport the music roll is supplied by causing Pneumatic motors, usually comprising two rectangular leaves of wood hinged at one end with a flexible air-proof cloth glued round the edges with the leaves opened to about 20°. When the air is extracted from the volume between the leaves, the pneumatic closes thus providing near-linear motive power. These come in all sizes from 1” x 2” to pedal operating pneumatics which can be up to 9” x 3” (say).

The human interface to these pianos is achieved using switches of varying complexity from simple “let air in” to multi-function devices which control several functions from the same switch. These switches are, in the later instruments (1910ish onwards) rotary devices made from brass and often nickel plated.

The switches Tony wanted me to make were of the double pole single throw type meaning that the switch prevents or allows the flow of air through two separate circuits and are connected into the mechanism by rubber tubing pushed onto brass elbows soldered onto the switch.

                             

It was then over to my Bridgeport mill equipped with a dividing head (I bought the mill for scrap value when a toolmakers’ shop was being closed and it is fantastically useful. If you have the space, keep your ears and eyes open as the older mills are not required in the modern factory!). This is a device used often for cutting wheels for a clock as the 360degrees of a circle can be divided, extremely accurately, into as many parts as is required, for example 48 parts in order to make a 48 toothed wheel. However, I didn't need to use that facility but used it to rotate the part being machined by a given angle read of the calibrated ring shown in the picture.

The stem of the body was mounted in the chuck and the dividing head set to 0°. I have a “clock” mounted on the quill of the mill which enables me to align a workpiece parallel to or at  90° to the bed and the picture shows how this is used.

           

The workpiece is rotated in the chuck (which is locked in the zero position) until there is no movement of the dial indicator and the chuck tightened. Please note these pictures were taken after completion! The quill is then positioned exactly over the centre 4mm hole (in this case I used a pointed rod as absolute accuracy was unnecessary otherwise I would have used the optical centre finder I made a number of years ago). The table was then moved 6.43mm south and the dividing head rotated through 31° and, using a 3mm slot drill, a hole 4.27mm was bored. This was repeated at 149° 211° and 329° to produce the 4 holes required for the two poles. The fixing and stop pin holes were drilled again using co-ordinates. The Tufnel under the workpiece is called sacrificial because it allows the tool to pass right through the work without hitting the chuck jaws and damaging them (or both).

The dividing head is then turned through 90° so that the edge of the body can be worked.

   

The face is then set to zero so that the holes in both edges can be bored using co-ordinates. First the 4mm bore was machined in both positions and then the 3mm hole was bored until it met the 3mm bore in the face of the body. Thus there was then a connection between the holes in the face and the positions for the elbows in the edge. The dividing head was then rotated through 180° and the other “pole” completed.

The last bits were the disks. A piece of 50mm brass bar was chucked in the lathe and a portion turned down to 40mm and centre drilled 4mm. Unfortunately, I didn’t have any 40mm stock. This was then transferred to the chuck in the dividing head in its vertical mode (had eccentricity been of importance, I would have transferred the chuck from lathe to dividing head). The hole was centred on the quill and a 3mm slot drill fitted. The table was moved once again 6.43mm south and the dividing head set at 86°. The curved airway was started by lowering the cutter 2mm into the workpiece and leaving the cutter rotating, the dividing head was rotated to 148° machining a 3mm wide curved slot. Remembering to extract the tool, the dividing head was rotated to 266° and another slot machined to 328°. To ensure that the tongue used to limit the rotation of the disk would be in the right place, a notch was machined at 0° at the edge of the disk. The disk was then parted off the bar using the lathe and then returned to the mill to machine the 1.5mm slot for locating the operating shaft. A brass tongue was soft soldered onto the disk later using the notch as a locating device. The picture shows one of the supplied disks being machined as I took no pictures of the “real thing”!

   

[Refugee]

Those switches must have been very expensive 100 years ago.
The instruments were exclusively for the rich in there day.

[AlanBeckett]

Splendid Brian
At last there's someone else who understands AutoCAD. Post the .dwgs and bugger the others
Seriously, I think I might have packed up the Disc in your last Picture. I think I'd have been a bit nervous about just having it in the Chuck.
Anyway, a splendid Job.

Alan

[Alan Douglas]

Aeolian must have been really upscale: those were made of pot metal in an Ampico.

[Murphyv310]

Hi Brian.
Excellent engineering, with a write up to match. Quality work like this is sadly lacking in many these days and it's a pleasure to see people still using their skills  

[Brianc]

I agree Alan. I should have packed up the disk but as I could tighten the church well without worrying too much about damage and the cut was only 2mm x 3mm, I didn't bother. Of course I set the piece in the chuck using a parallel between it and the chuck.
I used to use ACAD almost every day at work but I never mastered 3D. Did you?

Thanks for your kind comments, Trevor et al.

[Radio Fixer]

Great piece of work Brian and of course I have seen in the flesh, so to speak, some of your first class constructions.

I have never used Auto CAD but can recommend CorelDraw which has a most helpful user group with almost instant replies to problems. There is a Student edition' available cheaply' which is pretty much the full package as far as I can make out. I have never found a limitation for me ... Only one is user ability.

http://www.amazon.co.uk/CorelDRAW-Home-Student-Suite-X7/dp/B00M6M6R0M/ref=sr_1_1?s=software&ie=UTF8&qid=1437984771&sr=1-1&keywords=coreldraw

Gary

[Nick]

Ok - I have remove all the spurious formatting - this included odd characters still remaining, e.g. non-standard apostrophes and huge numbers of "font"-related commands. Also removed the original top post and placed a re-direct there.

Generally, it's not a great idea to compose in Word and then cut-n-paste into a forum editor as the MyBB parser makes assumptions which differ to those of Word etc. - results are then somewhat unpredictable.

A better approach, if you wish to compose off-line, is to use notepad or a similar plain text editor, then cut-n-paste that content.

Cheers

[Brianc]

Thanks for the tip, Nick. I will do just that next time.