17.5mm Rex






Pathé brought 17.5mm to the market in 1926 or even 1927, a full three years or more after Kodak introduced 16mm. It was almost certainly this delay that meant 17.5, a gauge obviously technically superior to 16, eventually lost out. Their first machine was the Rural, glimpsed occasionally in adverts in the UK as the Rex. They are very rare, I suspect because most of them were converted to sound - see Rural Sonore. I found one at the Argenteuil film fair near Paris in 2008.

I have just spent a lot of time fettling the Rex up for use, as I do have some silent film and have recently acquired some more. The problems I encountered are typical of those that beset the restorer of machines that are ever older; it is perfectly possible that this machine has run little or no film since before WWII. This has not stopped it fighting me every inch of the way. This is the machine from the angle usually shown in publicity.

rex001     Rex002 

The first problem was that it had no lamp transformer, having obviously been cannibalised at some point. (I have since realised that the lamp is indeed designed to run from a resistance, not a transformer, tho' I have never had the nerve to actually try it.) However, I had already established that the Rex and the later sound version used a lamp with the same fitting as the Vox and, I am guessing, actually used the same or similar lamp and tranny. I had a loose Vox tranny which I decided to fit. As ever, the problem is trying to fit something without altering the projector, which you will know is anathema to me in most instances. There were only two mounting holes available, positioned centrally; the Vox tranny had one per corner, all therefore in the wrong place. I took a 1cm square iron bar and drilled a central hole for a bolt to fit one of the available holes in the projector. Threaded holes at the ends of this bar provided mounting holes for one side of the Vox tranny. Obviously, with just a single point fixing, the tranny wobbled, so I made up a long rod to provide support from the other original mounting hole. A flat plate covers the gap from this rod to the two remaining holes in the Vox tranny. This plate also gives a convenient mount for a terminal strip. It's probably all a lot clearer from the pic; you can just see one end of the 1cm square bar at the side of the tranny above the green earth wire (I try not to operate without an earth having a strange desire to go on living.).

 Rex 014

This leads me on to the next point. The machine as designed has a single switch to operate both motor and lamp, which I don't like. To fit a separate lamp switch without altering the machine was a simple matter of extending the two bolts at the bottom of the switch panel to provide fixings for a plate on which I could mount a switch (and earth connection), with just enough clearance to fit below the projector base. This is the view from the outside; you can see the rest in the pic above. You can clearly see the bolts and nuts and the spacers to give clearance for the switch in the top pic. (I have since covered all the switch terminals with heat-shrink insulation to obviate any risk of shock). All terribly simple. Except....... The two bolts screw into threaded holes in the projector base, but I could not find any bolts in my stock to fit. This, I discovered, was because although the bolts were a standard M3 size, they were not the standard 0.5 pitch, but instead 0.6. So I had to send off for a die, and a tap (because I would need to make the nuts as well). Even then, the first die for some reason gave off-size results and I had to ask for a new one.



Anyway, in the end it got done. here is a close up of the switch panel, with resistances for both motor and lamp. The ammeter (the two terminals at the very top of the first pic above) is scaled to show a red line at 2 amps for the lamp. 200 watts at 110v - sounds about right, if a bit low. The problem is, with no tranny I had little idea of what the wiring should be. It looked a bit as if both the transformer and the lamp were wired as one circuit. Anyway, I had to change things for the separate lamp switch. I originally planned to eliminate the lamp resistance, but in the end kept it. I tested the motor direct from 110v, and it was fine. Wire it into the circuit and zilch. In the end I found the motor resistance mat had a tiny spot where there was rust, and the wires had simply ceased to be for a few thousandths of an inch. Fortunately, this was very near one end, and I only had to remove about 2 turns to get a new end.

The lamp shows well under two amps on the meter even with the resistance turned up as high as it will go (Mind you, the ammeter may not be very accurate - it had a small pool of dried-up oil in the bottom of the dial). The tranny reads about 14.7 volts with no load. Have not yet measured under load. There are two inputs to the tranny, for 110v and 130v; the second gives about 18v. I have tested with a Vox lamp and (because Vox lamps are hard to come by) a 15v 150w halogen. More on this later. 


rex005     rex006     rex007

The machine came with spool arms which take only 500' spools. Fortunately, I had some ready-made "Super Arms" in the shape of the arms from one of my sound (16mm) Rurals. Yet another snag here - the brake mechanism for the feed spool was busted - Mazac had distorted and enlarged and was solid. I had to mill it out in the end. Then, of course, I had to make a new part to replace the Mazac one. The bright aluminium is the new part. The two black sectors are brake shoes, pushed against the rim by springs running thru the alu bit. The centre screw is the end of the spool spindle, and takes a dished spring washer and nut to complete the assembly. On the subject of 500' spools, I was initially highly sceptical and could not imagine this being workable without two machines because of the frequent breaks involved. It was, however, pointed out to me by Mikael Barnard and the PPT that in fact such breaks often happened in the early days of cinema; the parallel with modern TV adverts is also relevant. I later spoke with one or two knowledgeable French cinéastes who confirmed that breaks after each reel were quite accepted and were used to order more drinks, chat with friends and neighbours etc. Once sound came along, shorter running time per reel meant spool capacity needed to increase. At least some amplifiers for the Rural Sonore and Super Rural were supplied with provision for changeover operation.

Having cleaned and oiled and generally checked all was well, I finally felt in a position to run some film, so got out the first two reels (of 8!) of my tinted copy of Christus - yes, the one that's on 9.5, but even longer. I got about two thirds of the way thru. I noted along the way that the framing was off - and there's no adjustment on these machines. Nor is there enough slack in the screws that secure the gate to make any significant difference. At this point we may as well look at the lamphouse and gate.


rex008     rex009

I had been concerned at the small size of the lamphouse given the amount of heat an incandescent lamp would generate. However, there seems to be a pretty good draught thru to the lamphouse and, when running, the machine makes quite a howl. An interesting feature is that some of the stream of air is diverted to the gate, which even has little "wings" to channel the air. You can see in the first pic, at the bottom, where the air goes to the main lamphouse. Do you by any chance recognise the lamphouse locking mechanism in pic 2? The machine was built by the same concern as the Pathé Baby and also has a cam which is just a scaled-up version of the one in the Baby.

To try to get just a little bit of adjustment for the gate plate to deal with the framing problem, I have made replacement securing screws with the head and the shank reduced from the original. However, I can't yet see if this works, as I shall explain.



I said I got about two thirds of the way thru my reel of Christus. Then I lost the bottom loop, the cause being the claw failing to go back in after the downstroke, and simply moving the film back up again; the claw had broken.


rex011     rex012         rex014

In pic 1 you can see the giant Baby cam behind the up/down cam. You can also see the two claw pins and, at the back of the brass claw carrier, a little sticky-out bit which actually goes over the giant Baby cam. There is another bit you can't see which sits on the near side of the same cam, and these two together provide the in and out claw movement. Pic 2 is the actual claw bit removed from the carrier; you may be able to see that one of the two sticky-out bits has fallen off. To me, it also looks like it's failed before, as there are signs of a brazed repair. This is not something I think I can fix myself so, for the time being, I'm stalled. I have made a drawing of the claw shuttle, with a view to trying to get one made - I think it is some special or toughened steel. In fact, I ended up making one myself. It's not quite right, but I know what is needed now and when I can summon up the energy I can hopefully make a really good one. The main problem is that in trying to get it to run freely, I took a bit too much off the sides, so it has too much lateral movement and enters the film at an angle. My new one is finally OK, and I have made an attempt at case-hardening it to help limit wear.  

rex015     rex016

A peculiarity of this machine that I have not previously encountered on 17.5 is that there is a moving plate inside the lamphouse, which either cuts off the light altogether or acts a a heat shield for a still picture - I think. The lever on the front has two clear indented "stops" which as far as I can tell correspond to the two possible positions of the plate noted above. It links to the spring-loaded plate in the lamphouse by a wire in a cable like for bicycle brakes. The lamphouse end was adrift and I had to find a good way to re-attach it. It all looks a fairly professional job, so may be original. There is no provision for stopping the projector mech, tho', so I assume the function was to enable the machine to be started with the light blanked until the leader was thru and the film proper began. I overcame this same problem by adding a separate lamp switch. (On later reflection, I was not at all convinced there wasn't some sort of clutch arrangement to stop the mech and show a still.)



I have for a long time wanted to make a tungsten halogen lamp converter for the Vox that does not involve changing the projector in any way. Past conversions I have done or seen mount the lamp in the side of the inner lamphouse, burning sideways. I reckon if a halogen lamp can burn sideways, it might just as well burn upside down. This is bound to shorten its life, but the lamps are a bit more readily available than Vox lamps. I have tried in the past to make up a ring with pins like the Vox lamp pins, but have always found this too tricky for my capabilities. I have therefore adopted a slightly modified approach as shown in the pic. An aluminium disc about 41mm diameter and 2 or 3mm thick sits neatly just above where the Vox lamp would sit, where there is a convenient (insulated) "step" in the lamphouse. The lamp is suspended from this by two support bolts. These in turn attach to a separate little unit made up from a washer drilled for the lampholder and the support bolts, with spacers to get the lamp down as far as is needed. There is fine adjustment of the lamp position heightwise by moving the nuts on the support bolts (either side of the washer) up or down. Electrical connection is made first by a bit of bent brass strip screwed to the alu disc and bent down thru a slot in the side of the disc. This sits into one of the holes the Vox lamp pins use and its fixing bolt also secures the first lamp lead. The second connection is thru the insulated rod, which is threaded at one end and so bolted to the alu disc. The other end is drilled for a short brass rod, with a threaded hole part way down matching a hole on the side of the insulated rod. A screw connection can then be made to the second lamp lead. The brass rod protrudes thru the top of the insulated rod to connect to the contact in the lamp cap.

It all sounds OK, but this job nearly drove me crazy. The main reason is that whatever idiots designed the lamp base set the lamp pin sockets at a weird angle to the holes I use for screw mounting - you can see this in the separate lamp base I have put in the picture above. My first approach was to use brass rods with a screw thread at each end, directly bolted to the lamp base. This ought to work, but..... The filament in the Vox lamp is suspended from two rods which are bent to move the filament closer to the condenser lens. I found I couldn't get away with straight rods moved further across as they then fouled the bore of the top section of the lamphouse and also the heat shutter mechanism. So I tried using bent rods. However, because the fixing holes and the lamp pin sockets are at that weird angle, one rod has to go behind the other where they attach to the disc in order to get the filament square to the condenser. This seems to mean that the rods cannot be identical, 'cos I tried and failed. I could not figure out how to make the second rod the right size. In the end, I came up with the design I have described, which seems to overcome most of the problems, but it took me a lot of trial and far more error to find exactly where the holes for the support rods needed to be in the alu disc. I was homicidal by the time I finished, but at least I now have something I can simply copy.

It took hours and innumerable dismantlings and re-assemblings of the claw and gate to get it right. The in-out movement of the claw is very short, so the pins have to be exactly the right length to pull the film down but retract out of the way on the upstroke. This was all complicated by the gate binding and failing to provide enough pressure to hold the film. I was trying all sorts of things, including putting a shim behind the front, fixed gate, which in turn affected the length the claw pins needed to be. The gate binding was a real nightmare. The gate is pretty much completely enclosed in a channel, so you can't see exactly where it is binding. I filed a bit here, polished a bit there, experimented with various thicknesses of spacing washers on the lamphouse pivot; each time I tried anything, it was take apart then put back together over and over again. What I had thought would be a reasonable job took many hours of work over several weeks.

It will run film now; the framing problem is almost fixed but is still just visible on titles and some shots, so sooner or later (probably later!) I shall have to try something else, even if only to appease my own perfectionism. There is also occasional framing problem on the left of the screen - I think the silent aperture plate was cut a bit too big. I am currently trying a secondary mask behind the fixed aperture plate. Another little problem was that the top sprocket retaining roller was not quite right and led to the top loop shortening. The locating pin did not quite fit in its hole and there was no spring to help hold the roller in place. I enlarged the hole a bit and fitted a spring. Finally, I spotted a small hole in the bottom of the lamphouse, for which I could see no use. However, on projection you get a shaft of light onto the ammeter!  

I went back to Argenteuil the next year and bought another. As usual, there were problems. Inevitably, one buys largely on sight (or someone else gets it) and takes the chance. I could see it had no lens or lens barrel, but I was sure that was readily fixable. What I had not bargained for was finding the motor was dead to the world, despite looking in very good condition. This is not necessarily the end of the world; it is possible to have motors rewound, tho' very costly. Or one could remove the innards and fit a different motor within the original case. I am actually going to try out yet a third approach. I have a total of 4 Rurals Sonores, all converted to 16mm. One I have restored to working order as a 16mm machine, one is a long-term conversion-back-to- 17.5 project. The other two are in very poor condition, and would be a major project to restore even as 16mm machines. I have therefore pinched the motor out of one of them to fit to my silent machine. There are snags; obviously the sound-projector motor has a condenser, and runs at a fixed speed so as to deliver 24fps. What I have done therefore is to make a new pulley, rather smaller but not, I think, quite down to two-thirds the size of its original. Hopefully this will deliver a speed somewhere around 18fps. Virtually all the paint had peeled off this motor, so I have done a re-spray, with lots of masking off the bits that need protecting from the paint (Hammmerite satin black). Never easy, eh? I've also re-sprayed the rear cooling fan cover, after filling a couple of holes where the Mazac had split off. I will keep you posted.


Been plugging away at the second silent Rural. I have repainted the donor motor, tho' inevitably it immediately starts to acquire a new set of scratches and knocks. Then, of course, the problems started. First off, the motor in the Rurals is held in place by a single knob that screws into the side of the motor casing. The motor housing is shaped in part to match the curve of the motor and so help to hold it in place. Well. First off, the way the blower fan fits to the motor is different - the silent machine has a LH thread on the motor shaft and in the bore of the fan. So I had to take the fan from the donor machine as well. Then I find that in fact there is a subtle difference in the motor, which means it sits at a slight angle which, in turn, means the fan is at a slight angle, so messing up the very tight clearances and causing the fan to foul the motor housing. (it tends to make a very bell-like ringing). The only thing I could think of was to make a "shoe" to provide some support at the rear of the motor to straighten it up, using the motor thru bolt as a fixing point. See pic. Another difference is in the oil delivery system; the original was a robust piece of copper tube which would by no means work with the donor motor. The little white tube you see by the top motor fixing bolt is part of a cotton bud, which is hollow and has just enough flexibility to ease around the bolt and get oil to the right place. Even now, it's not quite right, and the fan keeps starting to ring if I don't get it set up just right. I am therefore still pursuing plan A and finding out about a possible rewind of the original motor.

Anyway, I got to the point where I could seriously test the machine by running it. I had been concerned from the start that there seemed to be a bit of binding in the mech, which appeared to arise from the cam being too tight at some parts of its rotation and too loose elsewhere. It would run with the motor, but was stiff and sometimes wouldn't start, leaving the belt slipping. I was afraid that I might have to actually make a replacement for one of the brass surfaces the cam rides on, not a task I would relish as it requires engineering of considerable precision and I'm not sure I can do it. I took the claw/cam mech apart and removed one of the brass surfaces - it just looks like a small bit of brass angle strip. There is already a pic of the claw/cam from my other machine which shows you what I mean. I puzzled over this for a day or two, and measured up the brass angle in case I needed to make one. Then I put the machine back together and, blow me if it didn't work fine! All the stiffness gone, the only problem being I don't really know what I did to make a difference.

Another thing I have done is to make the lens tube and lens mount.

rex019     rex020     rex021

Pic one is the view from the back of the machine - the lens sits in a long tube which reaches back from the front of the projector to close to the plane of the shutter - I'm guessing this is to help protect from oil splashes. I didn't fancy buying a bar of expensive aluminium only to turn most of it into swarf (which anyway is the principal product of a home workshop) by making a huge hole thru it and cutting away the outside as well to form a lip. (Top thing in pic 3 is the lens mount and lens from the other machine). So I managed to find a bit of thick-walled alu tube; this wasn't quite big enough to make the full lip you can see on the front end of the original, so I made a ring (steel) to fit over the alu and hold it in place (pic 2). I also had to make the screws as, although they are 3mm, are not the standard thread and so cannot be bought off the shelf. The lens is 40mm from a Bolex; the rear portion of the barrel was about the right size and I bored out my tube to match it. The front portion I cut off and fixed to the front of my new lens tube - the intermediate section is a bit of old standard copper water pipe. I haven't tried the lens yet - I suspect I may have to shorten the copper tube a bit, as I am not at all sure I won't get the edges of the picture cut off - the original lens is a 55mm. It is very difficult to find a lens that will do the job - the absolute maximum possible size to fit within the lens mounting tube is 30mm diameter and, of course, it needs to have a wide rear element and be able to fit down at the bottom of that tube; in the original, the rear element of the lens is only a few mm from the end of the tube.

The little brass bit at the bottom of pic 3 is a replacement retaining nut I had to make for the blower fan. there are lugs on the back which engage with the shaft, but because of the slight differences with the donor motor, the lugs on the original were too short for it to work quite right - not all that clear as to why.

All this work - and I bet I end up with the old motor back. And still a lot more to do before it can leave the garage and join the good, working projectors in the loft. And I still have not resolved the issue of the lamp - I am told that a slightly larger version of the familiar sort of resistance mat serves to take the lamp power supply from 110v down to 15v 200w. My first Rural I didn't even know of this and fitted a tranny; I'm still dubious and highly reluctant.

There is good news; it seems I shall after all be able to get the motor for my Rural re-wound at respectable cost, tho' I shall now have to unpick some of what I have done - ain't it always the way?


rex022     rex022a     rex023     rex024     rex025     rex026     rex027

The first of these pix shows the donor motor, with the new pulley wot I made. I have fitted two x 2mm plug/socket thingies in existing holes in the top of the motor housing. These are from Maplin and I find them useful; be ready for them to cease being stocked almost at once therefore. These take the capacitor wires (remember, the sound machines used capacitor-run motors) down thru a hole I punched in the cardboard (or "shoddy", ie old-style suitcase material) sheet that sits between the motor housing/mech and the base. (The capacitor itself is in pic 3; note the sophisticated mounting using a bent strip of metal. I have fitted a tranny, but have yet to wire it up.) Pic 2 shows the simplicity of the silent to sound conversion of the Rural. This is a sound base, but only because it has the long rods. The silent version has short rods and bolts directly onto the base. The sound unit is about 2" thick and with holes thru which these longer bolts pass, with the mech thern bolted on top.

The remaining pix are basically workshop shots (sorry about the backgrounds) to help me should I ever decide to restore the original set-up - please note that, as (nearly) ever, I have made no permanent changes that cannot be reversed (bar that little hole in the cardboard). I have also left myself a little file note about the wiring, 'cos I know I won't remember it tomorrow let alone years from now.

The more I work on the silent Rural, the more I am impressed by its design and suitability for purpose. As the above pix show, all the electrics are in a single module, which removes with three screws and the disconnection of two wires. These are the two you can see flapping loose - they run between one leg of the power supply, via the ammeter, to the lamp socket. The motor and lamp connections are external. The drive from the motor to the mech is external too, so belt replacement is a cinch. The base is a very solid casting, maybe even iron. This is simplicity and design at its best - if anything goes wrong, it needs little skill or time to replace a module, be it electrics, motor, lamphouse or mech. The entire thing is robust, and came in a solid wooden crate for transport. In short, ideal for lugging round small villages and odd venues. And if one accepts, as I think I must, that the lamp is indeed controlled through the resistance mat, not a tranny, we have no AC/DC worries, and tolerance of a wide range of voltage, too. I am, however, not persuaded to remove the tranny.

On top of this, the design proved readily adaptable to sound with remarkably few changes. A design classic, I think.