Why did Leica use a British thread pitch and profile for the L39 mount

[levegh]

Advertisement (gone after registration)

It is difficult to measure the thread angle, but offering small thread cutting tools up to the thread on a lens, I am sure that it is 55 degrees rather than sixty. (That is 'sure' as in 'fairly confident that my eye is not deceiving me.')

Stuart

[Martin K]

This thread is quite amazing.  The more I read about the Leica thread, the more I am convinced that the actual specifications were 39mm 26tpi in spite of the draft DIN standard of 1984, which incidentally is no longer in the DIN catalogue of current standards.  Looking at the spec sheets of enlarger lens manufacturers, e.g. Rodenstock or Schneider Kreuznach, the thread is defined specifically as that, or just as Leica thread.  Maybe the reason Leitz used 26tpi is very mundane, like thread-cutting equipment was set up that way – they were major microscope manufacturers after all where Whitworth threads were the norm then and are still used today for objectives (0,8” 36tpi) for example. Whitworth threads have a 55 degree angle which would confirm Stuart's (levegh) measurements

Which reminds me of the anecdotal story as to how the standard railway gauge of 4ft 8½inch (1435mm) came into being.  This was the gauge Stephenson used for his railway lines.  His rolling stock was built by manufacturers of horse drawn trams that used this gauge.  Their predecessors were horse drawn carriages that operated on the roads and they used this gauge.  On the roads there were wheel ruts emanating from Roman times that were this width, and you don’t want the carriage to ride with one set of wheels in one rut and the other outside the rut, and the left- and right-hand side hopping in and out of the rut all the time.  The average width of the ruts was 4ft 8½inch.  And this was the standardized width between the wheels of the Roman war chariots, selected to ensure the “bums” of two Roman war horses fitted inside the wheels.

[willeica]

55 minutes ago, Martin K said:

This thread is quite amazing.  The more I read about the Leica thread, the more I am convinced that the actual specifications were 39mm 26tpi in spite of the draft DIN standard of 1984, which incidentally is no longer in the DIN catalogue of current standards.  Looking at the spec sheets of enlarger lens manufacturers, e.g. Rodenstock or Schneider Kreuznach, the thread is defined specifically as that, or just as Leica thread.  Maybe the reason Leitz used 26tpi is very mundane, like thread-cutting equipment was set up that way – they were major microscope manufacturers after all where Whitworth threads were the norm then and are still used today for objectives (0,8” 36tpi) for example. Whitworth threads have a 55 degree angle which would confirm Stuart's (levegh) measurements

Which reminds me of the anecdotal story as to how the standard railway gauge of 4ft 8½inch (1435mm) came into being.  This was the gauge Stephenson used for his railway lines.  His rolling stock was built by manufacturers of horse drawn trams that used this gauge.  Their predecessors were horse drawn carriages that operated on the roads and they used this gauge.  On the roads there were wheel ruts emanating from Roman times that were this width, and you don’t want the carriage to ride with one set of wheels in one rut and the other outside the rut, and the left- and right-hand side hopping in and out of the rut all the time.  The average width of the ruts was 4ft 8½inch.  And this was the standardized width between the wheels of the Roman war chariots, selected to ensure the “bums” of two Roman war horses fitted inside the wheels.

You reminded me of this.

 

William

[wlaidlaw]

I have had a response from Jim Lager of the LHSA. He put me in touch with a Mr. Ricardo Davidson, who he regards as an authority on the subject. I have had the following response from Mr. Davidson:

The Leica lens thread is 26 tpi and it always has been. If you consult Roedenstock or Schneider lens catalogues it is thus stated very clearly, as you say a longer thread would bind if it was 1 mm pitch. I actually had this experience some 30 years ago, I cut accurate 1 mm threads in mounting rings and found that the lenses would not thread on properly. As I told Jim, Canon made this mistake in their Leica-inspired cameras, which is why only a worn lens thread would fit a Leica camera. 

You can try digging up what thread pitch was used by the Soviets for their FEDs and Zorkis (and very early Zeniths). If they did a proper copy, they would have used 26 tpi.

I'm still waiting for some feedback from the microscope side, I'm fairly sure that Oskar Barnack did not actively select that pitch, he just ran a common pitch he had in his machine tools (36 tpi would have been another, and possibly 40 tpi). Pitches in inches were common in the microscope industry at the time, as were British and American machine tools, metrication was not as fully achieved by then as it would seem. A lathe can have a metric lead screw and cut a t.p.i. thread with appropriate gearing, and vice-versa. The only correct gearing must include a 127 tooth wheel, all other combinations are approximations which may do for most work but are not exact. If you have an inches lead screw on your medium or large lathe, it makes much better sense to work with that than to buy in more wheels just "because". The diameter would have been worked out on some other premise, again why not 40 mm, or 38 mm?

All this gels with Stuart's measurement of the thread at a 55º subtended angle, correct for a Whitworth form thread. 

Wilson

[TomB_tx]

In “The Soviet Copies of the Leica II” Dieter Walzholz also notes that the proper Leica lens mount is 39mm x26 tpi, while the early Soviet copies used a “purely metric” diameter, pitch and thread form. (M39)

So the Soviet copies were like the early Canon mounts - but likely not as precise.

The FED I copy was introduce in 1934. Canon’s first prototype (the Kwanon) was made in 1933, while refined versions were shown in Japanese photography magazines in 1934; so both appeared about the same time. I believe FED achieved reasonable production faster. 

Edited February 8, 2019 by TomB_tx

[jaapv]

On 1/25/2019 at 1:34 PM, wlaidlaw said:

Paul, 

I believe at some point in the 1960's or 70's the tripod bush was changed from 1/4" and 3/8" Whitworth to 1/4" and 3/8" UNC. These threads are slightly different with Whitworth using a 55º included angle, against UNC using a 60º. If the male to female mating is not very tight, you can often persuade a Whit nut to screw onto a UNC thread but not always. This is particularly relevant when working on older British cars, which use British Standard Fine (BSF) and Whitworth coarse for everything and as you say, it is getting difficult to find new nuts and bolts. I always try never to reuse nuts and bolts where possible. A prime cause of breakdowns on older cars. The motor racing industry world wide generally used UNC and UNF threads for everything except the engines, which are normally metric. This aways meant I had to carry around two complete sets of spanners, socket sets and hex keys when going to some event. The reason for this is that they use US aircraft specification bolts and nuts, which are all non-metric. My modern Morgan three wheeler is the other way round with UNC/UNF bolts and nuts on the American S&S engine but metric on all the chassis, body and transmission (Mazda MX5), so I still have to have two sets of spanners and hex keys. 

Wilson

I use Metrinch tools, as my car has collected an eclectic mix of metric and UNF. Having just one spanner set makes life easy.

[alan mcfall]

Advertisement (gone after registration)

The RMS thread (.8 inch nominal by 36 TPI) was reportedly adopted in 1858. Minor revisons in tolerance were agreed to in 1896, 1915 and lastly in 1936; and remains still, as per ISO standards of 2013 (Wiki). It was always Whitworth 55 degree, although in later times the M25, M27 and M32 were included in the standard. Small changes were a normal evolution as materials and cutting methods were also evolving. Heat treating effects, switching from raw round stock to tublar stock and maybe even plating thicknesses. Wear on both cutting devices and gauging sets bring other concerns. Today, CNC machines with computer programs would be used.

One can imagine the evolution at Leitz as new products were introduced.  Nickle plate, then chrome plate, brass and then some accessories in aluminum and other alloys.. An RMS .8 inch by 36tpi tap can be purchased commercially, but is expensive at about $100. A speciality grinder could make a 39 by 26 tpi. I use a 39 by 1mm die and it seems to do fine with the short thread length of lenses, just to clean them up when necessary.

Certainly, Leitz had over 50 years of experience with inch and Whitworth threads before the Leica appeared. Why 39mm, was probably driven by engineering as stated in one of the above posts. With a 39mm flange. starting with the Ic camera, the top plate required a cut-out for mounting. Model 1a cameras retured for updating had the metal removed at the top front of the top plate, and the body shell hole enlarged, and the transition from 3 lens holding screws to 4 flange holding screws.  I don't think Leitz would have introduced this new complexity (removing metal from the top plate) unless it was necessary optically or mechanically.

I have an old Volvo with metric inside the engine, and mostly inch outside it.  My BSA(A70) motocycle had me keep a small set of English tap and dies on hand.

[pgk]

Ifyou are unaware of it this site can be very useful http://www.macrolenses.de/start.php?lang

Looking down the last column in this table reveals a number of treads used for macro lenses http://www.macrolenses.de/objektive.php?lang including RMS and intriguingly M39 x 1/26". Leitz seem to have used quite a variety of threads for macro applications (I have a 12cm Summar with a thread of M36 x 0.75mm and its adapter of M40 x 0.75mm). One Minolta lens is shown as M39 x 1mm so there is clearly some known variance on the M39.

I am aware that in the 19th Century the RPS had some dealings with 'standardisation' and were responsible for tripod thread adoptions. I will email the expert and see if he is aware of any involvement with  M39 x 1/26" tpi.

[pgk]

A bit more info and links:

https://en.wikipedia.org/wiki/M39_lens_mount - suggests the Leitz were already tooled up to use  Whitworth form threads because of microscope production (RMS threads)

And from http://www.earlyphotography.co.uk/site/lens_mounts.html regarding earlier flanges"

"The usual method of attaching the [barrel type] lens to the camera was a screw thread, standard thread sizes were proposed by the Royal Photographic Society (RPS) which were adopted by some manufacturers including Taylor, Taylor & Hobson (T.T.H.) and, from around 1890, Ross. The RPS sizes, first proposed in 1881 and later modified, used a Whitworth Angular thread, for small diameters 24 threads per inch was used for sizes above 3" 12 tpi."

Given the history of using Whitworth threads in both microscopy and photography it would make obviously sense to continue to do so for a 'miniature' camera simply because they were 'tried and tested' and more importantly, they worked well.

This makes for interesting reading https://www.sizes.com/library/technology/RMS_thread.htm Conrad Beck was another British photographic lens maker (Beck) so it appears that the Whitworth thread for was well proven in similar applications already and the 'fit' was well understood.

[maximilianm3]

i don't know, but lot's of German companies used british/whitworth thread until the late 30's when they 'upgraded' to the metric system.

[luigi bertolotti]

On ‎2‎/‎8‎/‎2019 at 12:05 AM, wlaidlaw said:

I have had a response from Jim Lager of the LHSA. He put me in touch with a Mr. Ricardo Davidson, who he regards as an authority on the subject. I have had the following response from Mr. Davidson:

The Leica lens thread is 26 tpi and it always has been. If you consult Roedenstock or Schneider lens catalogues it is thus stated very clearly, as you say a longer thread would bind if it was 1 mm pitch. I actually had this experience some 30 years ago, I cut accurate 1 mm threads in mounting rings and found that the lenses would not thread on properly. As I told Jim, Canon made this mistake in their Leica-inspired cameras, which is why only a worn lens thread would fit a Leica camera. 

You can try digging up what thread pitch was used by the Soviets for their FEDs and Zorkis (and very early Zeniths). If they did a proper copy, they would have used 26 tpi.

I'm still waiting for some feedback from the microscope side, I'm fairly sure that Oskar Barnack did not actively select that pitch, he just ran a common pitch he had in his machine tools (36 tpi would have been another, and possibly 40 tpi). Pitches in inches were common in the microscope industry at the time, as were British and American machine tools, metrication was not as fully achieved by then as it would seem. A lathe can have a metric lead screw and cut a t.p.i. thread with appropriate gearing, and vice-versa. The only correct gearing must include a 127 tooth wheel, all other combinations are approximations which may do for most work but are not exact. If you have an inches lead screw on your medium or large lathe, it makes much better sense to work with that than to buy in more wheels just "because". The diameter would have been worked out on some other premise, again why not 40 mm, or 38 mm?

All this gels with Stuart's measurement of the thread at a 55º subtended angle, correct for a Whitworth form thread. 

Wilson

Good response from a source I personally  respect a lot : and the statement I evidence above is a perfect no-nonsense one : diameter, as I speculated in my post of some days ago, was an egineering-driven decision, the pitch was a manufacturing-driven one : I have, and personally live,  dozens of examples of this behavior still today.