Dividing this would have been an interesting exercise !!

[Michael GilliganParticipant @michaelgilligan61133]

This exhibit is described as a Wire Diffraction Grating : **LINK**

http://waywiser.fas.harvard.edu/objects/261/wire-diffraction-grating

The description is brief; but of course the fundamental requirement would be to have the wires parallel and equispaced. … So, I wonder how much effort must have gone into placing the holes neatly on three concentric circles ?

Unfortunately, the photo is not at high resolution; but it looks to be a fine piece of Scientific Instrument Making !

How would you have done it, with the tools of the day ?

Note that the centres of the ‘tuning pegs’ would be offset to account for their own diameters, and that of the wire !

MichaelG.

.

Edit: __ for comparison, here is a description of some ‘rectangular’ versions:

https://digital.kenyon.edu/cgi/viewcontent.cgi?referer=https://www.google.co.uk/&httpsredir=1&article=1109&context=physics_publications

Edited By Michael Gilligan on 15/08/2021 08:53:54

[Michael GilliganParticipant @michaelgilligan61133]

[Dave SParticipant @daves59043]

16 lines per inch, there are at most 72 wires.

Think they are 1/16" apart – sounds like rule and dividers layout. : )

Dave

[Michael GilliganParticipant @michaelgilligan61133]

I would respectfully suggest that you may not have considered some subtleties of the problem, Dave

MichaelG.

[pgk pgkParticipant @pgkpgk17461]

Without a DRO and with my limited skills, I would go about it as follows:

Make up/order/buy a strip of metal 1/4in + 4 wire diameters wide and drill/ream holes to match my punch's diameter and spaced and offset for the middle and outer pin holes and in it’s middle for the inner holes, noting the 'handedness' of the pins is different left than right. I'd also create a fixture ring as a stop for my marking bar.
At that point I'd mark/punch and drill the first pair of inner holes and place the pins then lay my marking bar against them for the next set of holes to be marked and drilled, placing only the inner pins and repeat.
Yes I accept that there is a tiny angular difference between inner pins since the wire run starts from an outer but depending on the purpose of this fixture it may not matter and it'd be a heck of a lot more accurate than me trying to work and keep track of all the dial numbers…

pgk

[John HaineParticipant @johnhaine32865]

Never mind how they made it, with a spacing of ~1/16" what was it used for?

[pgk pgkParticipant @pgkpgk17461]

Posted by John Haine on 15/08/2021 09:34:21:

Never mind how they made it, with a spacing of ~1/16" what was it used for?

Apparently it was attached to a telescope, presumably a large objective lens and if so then I’d guess used for plate photography?
Either that of the astronomers were fond of egg sandwiches…?

pgk

[Dave SParticipant @daves59043]

They have to be parallel and equally spaced.

The divider gives me a set of circles, the 1/16" scribe ensures I don't make a gross error.

The (assumed) milling machine /jig bore table takes care of the parallelism – drill one side then move in Y to the other side and drill. Move a precise 1/16 in X, set up over the scribed circle, drill the next one, move in Y.

Repeat until done.

That's how I would do it.

The item is a diffraction grating, used for looking at spectrums of stars I think

Dave

Edited to add: the circular pin layout is purely aesthetic, and it isnt that old -1920 – that it would preclude the use of an accurate leadscrew driven xy machine

Edited By Dave S on 15/08/2021 10:05:15

[Michael GilliganParticipant @michaelgilligan61133]

Posted by John Haine on 15/08/2021 09:34:21:

Never mind how they made it, with a spacing of ~1/16" what was it used for?

.

The listing suggests Microwave or Infra-Red [click the + signs]

… but there’s not much detail there.

MichaelG.

[bernard towersParticipant @bernardtowers37738]

Is it possibly a bahtinov mask

[JasonBModerator @jasonb]

I would have laid it out say 10 times bigger and use some form of pantograph machine to place the holes, would also have made it relatively quick assuming these were made in batches rather than a one off.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Dave S on 15/08/2021 10:01:22:

[…]

Edited to add: the circular pin layout is purely aesthetic, […]

.
… and that’s what piqued my interest : The craftsmanship that appears to have gone into that elegant but unnecessary layout.

MichaelG.

[Mick B1Participant @mickb1]

Posted by Michael Gilligan on 15/08/2021 10:20:22:

Posted by Dave S on 15/08/2021 10:01:22:

[…]

Edited to add: the circular pin layout is purely aesthetic, […]

.
… and that’s what piqued my interest : The craftsmanship that appears to have gone into that elegant but unnecessary layout.

MichaelG.

Yes, but AFAICS a rectangular layout would have to be bigger. It may be the whole thing had to fit inside a tube of particular diameter to be at the right focal distance or suchlike.

[John HaineParticipant @johnhaine32865]

Posted by Michael Gilligan on 15/08/2021 10:20:22:

Posted by Dave S on 15/08/2021 10:01:22:

[…]

Edited to add: the circular pin layout is purely aesthetic, […]

.
… and that’s what piqued my interest : The craftsmanship that appears to have gone into that elegant but unnecessary layout.

MichaelG.

Well no, surely? If the rows were straight either the aperture would have to be smaller or the OD larger which might not fit. If it was microwave it wouldn't be used on an optical telescope, and the spacing is too large even for infra red.

I think Bernard is on the right lines, it generates a diffraction pattern in the telescope focal plane when focused on a bright star, which is characteristic of the optical performance. Ideally if the spatial frequency of the wires is constant across the array there should just be a single line, but that would be smeared out by aberrations. Maybe Neil could cast some (coherent) light as our resident astronomer?

[SillyOldDufferModerator @sillyoldduffer]

Posted by Michael Gilligan on 15/08/2021 08:26:35:

…

The description is brief; but of course the fundamental requirement would be to have the wires parallel and equispaced. … So, I wonder how much effort must have gone into placing the holes neatly on three concentric circles ?

Unfortunately, the photo is not at high resolution; but it looks to be a fine piece of Scientific Instrument Making !

How would you have done it, with the tools of the day ?

… for comparison, here is a description of some ‘rectangular’ versions:

**LINK**

Dave's method would work, and is the only way I can think of producing a round diffraction grating to fit in the end of a telescope. (Rectangular gratings are rather easier to make once the kenyon paper has been read.)

The hard part is the accuracy required. Not difficult to make a rough grating if the purpose is only to demonstrate diffraction. Much harder to make a good grating for accurate measurements. Fine results demand the wires be accurately equidistant and accurately parallel, and I don't think an ordinary mill and DRO would do the job particularly well. I think the wires need to be much the same diameter too – accurarely made, and not stretched randomly when the grating is strung.

Really annoying, I can't give it a go. My workshop is full of boxes at the moment pending daughter leaving home, and I can't get at my tools or bench.

This paper has more on the theory. Went clean over my head at school! Don't bother reading it unless diffraction gratings are your thing.

Dave

[Ady1Participant @ady1]

Its a 1920s veg-o-matic insert

[Michael GilliganParticipant @michaelgilligan61133]

Posted by John Haine on 15/08/2021 10:31:58:

Posted by Michael Gilligan on 15/08/2021 10:20:22:

Posted by Dave S on 15/08/2021 10:01:22:

[…]

Edited to add: the circular pin layout is purely aesthetic, […]

.
… and that’s what piqued my interest : The craftsmanship that appears to have gone into that elegant but unnecessary layout.

MichaelG.

Well no, surely? If the rows were straight either the aperture would have to be smaller or the OD larger which might not fit.

.

There would be no need to change the size or shape of the item, John

There is already an infinite number of possible locations for each peg.

The maker chose the elegant layout.

MichaelG.

[SamsarandaParticipant @samsaranda]

Michael’s link at the top of the page lists one of the users as Oak Ridge from 1931 to 2005. Oak Ridge was the premier nuclear facility for preparing enriched uranium for the Manhattan project and the USA’s subsequent nuclear Arsenal, could there be a use for this item that we are not considering? Dave W

[John HaineParticipant @johnhaine32865]

Different Oak Ridge – this one is in Massachusetts, the "nuclear" one in Tennessee.

[Peter Cook 6Participant @petercook6]

I would have thought dividing to that precision would be well within the capabilities of the clockmakers of the time.

This is mid 19th century work.

The dial is 230mm across, the third wheel (centred in the dial) is about 75mm diameter and less than 1mm thick. I haven't counted the teeth. And I don't think I want to try and make it!

[Michael GilliganParticipant @michaelgilligan61133]

Angular dividing is simple, Peter … But this device also needs to include all the offsets that relate to the use of those ‘tuning pegs’ and I think it would be a little trickier than dividing a clock wheel.

MichaelG.

[Jeff DaymanParticipant @jeffdayman43397]

This job would be very simple on a large rigid jig borer (Loewe, Brown and Sharpe etc). Any watch, clock, gauge or instrument factory in the US in the 1920's / 1930's making their own precision dies for stamped parts would have such a jig borer. Builders of the telescope discussed would likely know of, or be able to find a shop to have the part made. Hole positioning accuracy on the Loewe jig borers was +/-.0002" or better, to XY coordinates, not rotational dividing. (position tolerance was often closer than .0002" on one in a factory toolroom I worked in)

[Pete RimmerParticipant @peterimmer30576]

I don't see it as a circular dividing problem at all, in fact you could not do it by dividing since the pins are not radially equi-distant as dividing would give.

The only consideration is that the wires are equi-distant linearly – the pins are only off-set radially due to space constraints. A jig bore would make easy meat of it, or a mill and modern DRO.

EDIT: as a matter of fact the easiest way to set the wires at equal pitches is to use a slitting saw of the same thickness as the wire diameter and cut alignment grooves in the raised annulus. The pins are then inconsequential.

Edited By Pete Rimmer on 15/08/2021 15:29:09

[old martParticipant @oldmart]

As Pete Rimmer says, it is the parallel spacing that matters, the radial aspect could have been taken care of simply with a scratch made with dividers at two radii.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Pete Rimmer on 15/08/2021 15:24:47:

I don't see it as a circular dividing problem at all […]

.

I think that’s pretty-much where we came in, Pete

MichaelG.

[Author]

Posts