17.5 - 9.5's Big Brother 1


17.5mm - 9.5's BIG BROTHER

Once, many years ago when I was young and foolish, I had a lot of 17.5mm film and several projectors. Moving to London and getting married meant something had to go, so I sold the lot to John Burgoyne-Johnson of the Buckingham Movie Museum in the early 70's. Even then it was obvious that 9.5's big brother had special problems, with many of the films irretrievably damaged and the only generally available projector the probable cause.

However, age does not always bring wisdom, and in middle age I suffered a renewed outbreak of the disease, in a particularly virulent form. It started when I bought a rewinder and a couple of odd reels of film at Pimlico. I was still convinced I could give it up at any time. But at the October 2001 Ealing convention I saw a Home Talkie being bought by a collector and the rot set in. In January 2002 I went to the Argenteuil film fair to look for a projector. I failed, but I was hooked. I found a projector, (a Home Talkie from Ruislip, the film nerd version of going round the world to meet the girl next door), and soon had many reels of film, a tape and several cement splicers and plans to re-convert to 17.5 projectors that had been converted to 16mm.

In researching 17.5, I found four principal published sources:-

The Home Cinema, Gerald McKee

A Guide to 17.5mm Sound, Maurice Trace

Grahame Newnham's web site at

Pathé, Premier Empire du Cinema, Jacques Kermabon (ed)

The latter was published in 1994 by the Pompidou Centre to go with an exhibition on the first 100 years of the cinema. Fortunately, I can read enough French, albeit slowly. 

The first thing to strike me was that there were four 17.5mm sound projectors made by Pathé, whereas I had previously seen only the British Home Talkie and read of the Super Talkie. In France, there was the Natan 175, an obvious but not quite identical precursor to the British Home Talkie. Then there was the Rural Sonore, a sound version of the silent Rural, known (but only very slightly) in Britain as the Rex.I have never seen or heard of one of the "improved" Home Talkies with a blimp. This could perhaps be counted as a fifth machine. Here is the line-up.

kermabonrurala     rex 034a     Rex 035     Natan1     projectors 035b     16RuralOK002     16RuralOK005     Rural Sonore_Seydoux     super rural_combo_005c2

These pix are in a very rough kind of order. First we have Kermabon showing the Rural (and its adaptation to sound), then 2 shots of my Rural (the second one is included because that is the angle Pathé always used). The fourth pic is a Natan, with the island plinth very obvious, but much more interesting in that the location of the photo-cell is presumably a feature I had not previously understood. It is described in adverts for the Improved Home Talkie as  "direct reading of the sound track", but as the improvements also included a blimp, it was impossible to see this feature. Then we have the standard UK Home Talkie. In the early days, Pathé were not very careful and some UK ads showed the French version. Then we have a pair of shots of a Rural Sonore (the sound version of the Rural). This is was converted to 16mm but is in working order and clearly shows how simple the conversion concept was. I have devoted much time and effort to the re-conversion of another 16mm Sonore back to 17.5 but, as yet, without enough success. I have also included a pic stolen from the Fondation Jérome Seydoux website because it is such a beautiful pic as, indeed, are all the site's images. This machine never made it to the UK with one notable exception. During the 1930's, the LNER ran Cinema coaches on some services. For a long time, I had assumed, rather dubiously, that a Home Talkie was used, but then I found this:-

175 en_trainaI think this proves conclusively that one Rural Sonore at least did make it to this country. The final pic is the Super Rural (I am not sure from ads that the French and English versions were identical, tho' I have never come across an avowed British one.

The only UK Rex (sans lamphouse, regrettably) I have ever seen is at Wroughton, near Swindon. This is the large object storage facility for the National Museum of Science and Industry. There is all sorts of stuff here, including most of the contents of the old Buckingham Movie Museum. These went to the Bradford Film and TV Museum, part of the Science Museum. I e-mailed them and arranged to go and see the Rex (Wroughton just keep stuff - the people at Bradford have the responsibility). Wroughton does however have open days from time to time and it is a fascinating place to visit. The website is at www.sciencemuseum.org.uk/wroughton.

Anyway, to return to my Ruislip projector. Obviously, the first job on acquiring any old projector is to take it apart (very nostalgic - doing this with 9.5mm Babies was one of my early experiences and is very therapeutic). Just as well, as mine was not only without any earth wire (normal for its time) but had perished insulation and exposed bare wires on the mains lead inside the casing. After that, job two was to drill a hole and fit a separate lamp switch. As I never throw things away, I had one of the same pattern. Third job was to replace what I assume was originally a rubber pad under the mechanism for reducing noise and vibration. The influence of time, oil and heat had turned it part crumbly, part liquid mess, nasty to remove.

Fourth job, converting to QI, was more complex. The original lamp is 16v 8 amp, with provision to overrun at 17v and underrun at 8v, for reasons which are not clear to me. I decided to go for 15v 150w. I thought if I left the mains input setting at 250v, this should give about 15v since our mains is now about 230v. As for the increased amperage, I just hope the original transformer was over-engineered. The lampholder is mounted to a rod which fits inside the original lampholder without making any permanent alteration.

The new lamp did however mean some form of cooling would be prudent. I wanted to avoid as far as possible any irreversible changes to the machine, so I made an aluminium ring which is a push fit into the lamphouse in place of the plastic end cap, with a slight lip to stop it going in too far. On this is mounted a 40mm, 12v computer fan from Maplin, which can be seen in the pictures. It is set to blow rather than to suck as it says on the packet that the fan is good only up to 70 degrees C. I thought the lamp might deliver more than this if the fan sucked. To provide a DC power supply, I had to build a small unit from a kit (Maplin again) and wire it into the lamp circuit so the fan comes on with the lamp. Power lead goes behind the sound chute and plugs into a socket mounted alongside the lamp and motor connections. (The holes for this and for the lamp switch are the only permanent changes from the original). Future options include a shield to keep the direct airflow away from the lamp (they go black I am told) and another to reduce light spill out through the fan. A higher power lamp (probably fed from an external 24v 250w ex-Elf transformer) is also a possibility, as I find the light is still on the dim side. The power supply for the fan has a variable output so there is some cooling capacity in reserve to cool a brighter lamp.

HmTkMechcubig     HmTkrearcubig     175splrewind

The original amplifier didn't seem to work and I didn't regard them as worth bothering with anyway (times change - I wouldn't say that now [2015]). So I just took out the valves and fitted a solar cell. This is a bit of a clunky affair, being designed for 35mm and in a big casing, but it does the job and precision in mounting is far less critical. The solar cell mount is screwed to a small metal block with a through hole. The original sound mirror can be unscrewed, leaving a rod which goes into the aforementioned hole in the cell mounting block. The wire is led carefully away and under the motor, but left as a flying lead at this stage. It goes into a pre-amp I built years ago for my Super Vox (which never had an amp at all) and thence to a separate main amp. An integral amp is a longer term project. The alu plate above the sprocket is simply to keep the wire from the cell out of the way.

Even before I had a projector I made myself a 17.5 rewind adapter. I have a 35mm rewinder, Phillips type. The spindle size is the same as both 16mm and 17.5mm (and 28mm for that matter) and the drive to the spool fits into the notch in 16mm spool centres. So it makes a good 16mm rewinder with a 100ft plastic spool as a spacer on the outside, and a 17.5mm rewinder with the adapter (made from an old 16mm camera spool) on the inside.

I shall try to include some pictures to make all of this clear (at least, clearer than it is now).

The machine when I got it was in very good original condition and in its original box with matching speaker, and even the separate gate mask for silent film, but with one or two small touches that I had not met before. The mirror was held in place by a bracket, obviously original as it is painted the same colour as the machine, rather than just relying on the friction fit into its ball socket. This is a minor but important point as it is very easy to knock the mirror when threading. The other thing was the provision of two hand rewinders. These are small handles with pins that fit into the holes in the spools around the central spindle. I have previously only seen these for 9.5.

The machine also has all the old problems that I remember, with some pretty fundamental design flaws. First, it has a single sprocket. The amount of contact between the film and the sprocket is therefore very limited. Correct threading is tricky yet absolutely critical. Putting the film through for the take-up easily dislodges it from the feed side. I have seen hundreds of feet of film dented all along the side where it has not been seated on the sprocket pins properly and, of course, many of these dents end up wrecking the sprocket holes. The top loop is lost and the claw then attacks the perforations.

This leads to the second basic flaw. The top loop is not only ridiculously short, but the sprocket is offset in relation to the gate (The 9.5 Vox is like this but with a much longer top loop to give a bit of leeway). As far as I can see, there is absolutely no reason why this offset has to exist. It is all the more odd as the film has to come back across again as it goes around the back of the lamphouse, to meet the take-up side of the same sprocket. These two flaws together mean the machine is very unforgiving; there is no slack in the system and the slightest hitch leads to instant damage. The powerful motor has a fair bit of momentum and can run on causing damage over some length of film. 

The third flaw is the sound smoothing system. It uses a chute (a similar arrangement was later used on the 9.5 Pax, although the 17.5 one has one side hinged to allow the film to be inserted). There is no additional smoothing between the sound reader and the sprocket. And the wrap around the flywheel is too short. The chute is a particular problem. The instruction book specifies that the film must be pinched together from the sides to take up a bowed profile. With side sprockets this is asking for trouble at the slightest weakness in perforations . Any misalignment of splices is likely to be a problem too. 

The final major flaw in my eyes is the lamphouse. Because it is enclosed by the film path, it cannot be enlarged and has minimal cooling from a crude fan on the motor shaft. The same lamp also serves for the sound.

All of this means that film has to be in apple pie order to project at all, and then only under conditions of great vigilance. Which is why I went to the perhaps rather extreme lengths of having a basic form of tape splicer made. I am increasingly concerned to do everything possible to preserve film and have come to regard cement splicing as a form of vandalism. You can't take cement splices apart (they all too often do this for themselves anyway) and lose two frames if you have to remake. Tape splices do everything cement ones do, only better and with less damage. A much wider range of repairs is possible too, which is critical with sound film where the loss of even one frame is so much more noticeable than with silent. And I have the complete answer to anyone who suggests that tape splices do not last. One of the 17.5 films I have acquired recently is one I knew in the old days. It has some splicing tape reinforcement that I know was applied at least 30 years ago. No drying out, no discolouration. Q.E.D.

Anyway, my tape splicer is simply a lump of metal with a channel in it of 17.5mm width. The film is held by pins in the right place to cut with the built-in blade and to bring the two cut ends together. Cutting the splicing tape once applied is by blade down a gap in the side of the channel. A separate punch then re-makes the perforations. Its not as good as a boughten one of course and is more laborious to use, but even this basic version was expensive as a one-off. The main additional requirement is the patience of Job, as so many of the films are so damaged. I did acquire a converted cement splicer with the machine, but I see no point and there would be no way to make some of the films projectable with this. 

Now to some more details. Kermabon gives the image size (all in millimetres) for a number of different film gauges:-

35mm 18 x 24. Projector aperture 15.2 x 20.95 according to another book I have

28mm 14 x 19

22mm triple band (1912), each image 4.2mm x 5.6mm

17.5mm 9 x 12 (silent)

17.5mm 9 x 11.3 (sound)

16mm 7.5 x 10.4. Projector aperture 7.231 x 9.652

9.5mm 6.5 x 8.8 (silent)

Oddly, the 17 .5 silent gate mask seems to have an aperture of 13.1mm x 9.7mm. The sound mask is 11.3 x 8.7. This whole area of camera, printer and projector apertures, mixed up with film shrinkage during processing and afterwards in storage seems to be an absolute minefield.

I have figured out or measured a few more dimensions for 17.5. Kermabon refers to the perforations as square, but some at least seem to be slightly rectangular, at 1.6mm wide by 1.5 mm high. The frame height may be 9mm but the overall sprocket pitch including the frame line is 9.5mm, which is as you would expect twice the pitch of a 35mm sprocket hole (think about it). I make the gap between the perforations and the edge of the film a mere 1.2mm. I have also tried measuring the sound track and its position but this seems widely variable. I have French film with tracks from 2 to 2.5mm wide, with the picture 3.1 to 3.6mm from the edge of the film. A British one shows 2.8mm of track in a margin 4mm wide from edge to picture.

(I later found a proper drawing of 17.5 dimensions; this does not mean they were necessarily followed!)


I havent reached a conclusion on sound separation yet. 35mm is 20 frames , 15 inches, 16mm is 26 frames, 7.8 inches. 17 .5 may even vary between France and UK.

John Cunningham kindly gave me (Imperial) dimensions for 17.5mm sprockets of varying sizes as follows:- 

8 tooth: Drum 0.940"-0.945", Overall 1.016" 

12 tooth: Drum 1.414", Overall 1.49" 

16 tooth: Drum 1.8945", Overall 1.97" 

He also gave pitch as 0.372", which I believe corresponds in metric to 9.5mm (see above).

This is all probably a lot more than you ever wanted to know about 17.5mm projection, but I have had a great deal of fun with it.

Click below for Part 2 of the Big Brother saga.

 Big Brother 2  17.5