July 2015
Eumig PIII

I’ve had a Eumig PIII for ages, but it somehow managed to escape being dismantled, which is the normal fate of machines that enter Cinerdistan. However, recently I saw a clip on YouTube which appeared to show a PIII with the fabled, never-before-seen notching device. For this reason, I bought one I saw on eBay to see what it was like, but no luck – still no notching. Still-frame device yes, stopping on notches, no. However, it is very obvious that there could be notching, as both parts of the gate make provision for it (pic 11 below), tho’ how the rest of the mechanism might work is impossible to guess. I found this out when I dismantled the new acquisition, and then my original a bit, so as to be able to make comparisons.

The two machines had a few differences; the new acquisition carried a serial number in the 14,000 range, my other machine in the 15000 range. I am assuming the earlier number was really earlier in time because it lacked some more sophisticated bits displayed by the second. In the first, the rear gate is screwed directly onto the Bakelite claw/shutter cover at the front, without even any brass inserts to reduce the risk of stripped threads. The second has a substantial steel plate added, behind the claw/shutter, with arms coming forward and around to hold the gate (see pix 4 and5 below). More trivially, the second has more blades to the fan and a neat little cover to the rewind gears on the front spool arm. The second sports threaded screws in one or two places where the first has rivet-like affairs (which I regard as a bit sloppy), but many remain even on the second and make maintenance and repair that bit more difficult.


I have to say that, overall, I was not terribly impressed. A decent standard of engineering in many ways, with nice touches like nearly all the screws are the same and use a common or garden thread size, something Pathé were very bad about, using obscure threads and many size variations. The sprocket assemblies (pic 10) are also very neat and film can easily be pushed in without having to fiddle with knobs etc. But some of the design features are odd, of which more anon. First, the pix.

EumigPIII001  EumigPIII002  EumigPIII003a  EumigPIII004  EumigPIII005  EumigPIII006EumigPIII007

EumigPIII008  EumigPIII009  EumigPIII010  EumigPIII011  EumigPIII012

A basic problem is the fitting of a still picture device, maybe with notching in some machines (often the more sophisticated machines, from whatever maker, never actually reached this country because we were too poor or mean to afford them, so have never been seen here). The blurb also boasts of no diminution of light for still frames, and there is no sign of a heat filter such as on the Bolex DA and PA. What this means is poor light, crippling the machine from the start. There are, unusually, three lenses between the lamp and the film (two are in pic 8; you can just see the second in the hole behind the first), which may be part of an effort to keep heat away from the film; the third element is  immediately behind the gate. Both machines came with a 250w lamp (240v in the first, 110v in the second). There is a shorting plug at the back of the machine, which allows for a supplementary resistance, so that a 110v lamp could be used from a 240v supply; otherwise, the lamp has to be changed to match the mains. Courtesy Tony Reypert, I now have one of these, labelled 110v/300w (pix 1 and 2). It seems doubtful a higher-powered lamp would be feasible as the draught from the fan is not really adequate, particularly on the earlier machine (two of the four fan blades had broken off and I had to rummage around for a replacement). Oddly, and I am not sure how this is achieved, there seems to be no change of engine speed when changing gear – in contrast, the Bolex DA motor speed increases markedly at still frames. Trouble is, I don’t have any PIII instructions to confirm or disprove my speculations. Copies gratefully received.

Being a dual-voltage machine presents certain aspects which need resolving and, in this case, it is done with considerable ingenuity, even tho’ I find it a bit dubious. The motor is presumably wound to tolerate a wide range of voltages, and the rest is achieved by the motor speed resistance (pic3). This has a long range, and different parts of it are used for different voltages. The speed control is a knob passing thru the rear of the bakelite claw/shutter cover. This is journalled in the front plate of the machine, and behind this carries a slotted fork, which in turn engages with a pin on a circular plate on one of the shafts that form part of the mechanism. Even while the shaft carries part of the drive mech, affairs are so arranged as to allow the shaft carrying the pin to rotate independently, as driven by the speed control. At the rear end of the shaft is a wiper that contacts the resistance; the range of travel of this wiper on the resistance is mechanically linked to the 110v – 220v voltage control at the rear of the machine. This is all very ingenious, but the fit of the knob at the front is rather sloppy and using a shaft that has another principal function seems a bit too clever. Moreover, a UK 240v, which is still pretty much what we actually get in practice, doesn’t sit well with a 220v setting and makes the motor run a bit fast. I think the machine is better at 110v.
The very compact design of the machine may be responsible for another oddity. The lamp is mounted at an angle facing towards the rear pic 8); this enables the lamp to sit further forward, to give space to the mech and motor to the rear; a more conventional right-angle mounting might have lengthened the body.

I found the PIII a b*****r to work on. I didn’t find a way of dismantling it that was much short of total. The basic holding together is done by a bracket fore and aft (pic 9), into which pass two screws from each side and from the bottom (pic12) of the machine. The reason there are three brackets in pic 9 is that, while rummaging for something else, I found one in my "possibly useful one day" box of bits of metal. This is curious, as I never had but the one PIII and have never previously taken it apart anyway. All the more curious then, that so unlikely an object should have been acquired somehow from somewhere at some time in the past.

The earlier machine had a mains lead where the insulation had perished, exposing bare wires. It was very fiddly and difficult to re-do the wiring, and I ended up first drilling out the rivets holding the main switch and replacing them with proper nut and bolts, and then cutting the wires and re-joining them using plastic terminals to make it easier to re-assemble without doing them first and then finding them in the way. As ever, mirrors had tended to deteriorate, especially in the earlier machine. I could replace the mirror behind the lamp fairly easily, but the angled mirror behind the gate was thin and small – about 0.5” by 1” – and I had to make do with a strip of polished metal.

A serious defect in operation is the take-up. It is driven from a tiny pulley behind the sprocket (in fact, there is one behind both sprockets, tho’ there doesn't seem to be any provision for one to act as a rewind. Maybe you have to switch belts for silent film to get the opposite direction of turn; when I tried by the traditional belt-crossing method, it tended to foul the spool. Anyway, the diameter of the pulley is so small the belt seems unable to gain any traction unless tension is greatly increased, which tends to undermine the “slipping clutch” effect of a spring belt.

The still picture arrangement is achieved by a push-down thingy rather like the Kid. This one is in RUN mode when the thingy is pushed down; when the button is pushed, a spring pushes a metal strip up to engage with one of the gears and take it out of mesh, (pix 6 and 7) to give a still picture. There must be a risk of the barrel shutter stopping in a position that obscures the image; the only way of “inching” the mech to move the shutter is the main drive pulley at the back.
It seemed to me that the claw was noisy, giving quite a sharp-sounding clatter – this noise of course stops in still mode.