The Talking Machine Forum — For All Antique Phonographs & Recordings

Herewith are some pictures of a small EWC (Excelsiorwerke Cöln) phonograph, modelled on the Columbia Q in its later form, which I have recently bought. I do not usually accumulate machines, but the price was moderate and I was charmed by the red and gold decoration which is still bright and almost complete.

I believe this machine is not far from being fit for use. It should have a wooden base or perhaps a box with a rotating lid, but this does not affect it mechanically. The reproducer from my Q seems to be compatible, as does the winding-key. (I already have an EWC recorder.) The reproducer carriage is of pot-metal and has swollen and distorted in the usual way, but it still slides freely on the guide-rod, and the cam which acts on the reproducer and the half-nut is functional despite having been broken and repaired. The mainspring is undamaged, the drive-belt (fabric rather than leather) is serviceable – thought I have enough material to replace it if necessary – and the motor components, including the governor, appear sound. However, a problem appears when I try to wind the machine: at first all appears normal, but after two or three half-turns the key tears itself out of my hand and whizzes backwards at a fearsome pace. My first idea was that the mainspring was broken near its outer end, so that it took the strain until its tension was strong enough to overcome the friction of the overlapping ends; however, as the pictures show, it is unbroken. I now believe the fault lies in the ratchet-and-pawl assembly; I say this because I once encountered identical symptoms in an old American shelf clock when the pawl itself fractured. Here the components – the ratchet incorporated with the spring cage and the pawl mounted on the inner wall of the motor housing – seem undamaged, and the pawl pivots freely on the shank of its retaining-screw. It strikes me as unusual, however, that the pawl is mounted upside-down, with its pointed ends upwards, whereas usually this part sits above the ratchet-wheel so that its own weight helps to keep one end securely between the teeth of the wheel; inverting this arrangement, as here, would seem to mean that the force of gravity becomes an enemy instead of a friend. Is it possible that there is a part missing, namely some kind of light spring which should bear on the tail of the pawl and hold it against the wheel? (Clock movements usually have such a spring, although I must admit that I have not seen the like in a phonograph or gramophone motor.) I can see no hole or stud which might have located such a spring, but it could have been simply wedged into a corner of the casing.

Is anyone familiar with these German machines? Should there be a spring as I have tried to describe, or am I missing some other cause of the problem?

Oliver Mundy.

Your problem most likely is the ratchet pawl. It might have been reversed by some previous owner. The pointed end should engage the coarse winding gear when tension is placed on it from winding to keep it from spinning backward. Try flipping it over to change that and see if it works... I have an Excelsior machine and can check it, but hate to take it apart. This might be why you got a good deal on it... someone screwed up.

Thank you, Curt! This seems entirely believable and can be checked in a couple of minutes.

Oliver Mundy.

Hi Oliver
Just checked some internal photos of my EWC and the part is the same way around.
Maybe not 'catching' on each wind. Can you check it partially assembled by winding the spring with the case lid off.
Brian

Oliver
Another photo of the spring housing and ratchet? Mine appears more pointed than yours. Maybe there is more wear on your ratchet and allowing the spring to slip. Just a thought.
Brian

Oliver
Check the gears out. Look at my last photo. Your gears are not touching each other.
Brian

Brian, thank you for your contribution. Having tried reversing the pawl last night and found that this was a false scent, I now think that the problem may stem from wear on the teeth of the ratchet-wheel, exactly as you have suggested; perhaps the points of the teeth are intended to catch against the left-hand end of the pawl before it disengages wholly from the ratchet, so that when the ratchet teeth are worn the whole geometry of the system is fatally upset. If so I suppose the fault is incurable, unless I can have a new ratchet-wheel made – or unless my improvisation works: see below.

As for the seeming disarrangement of the brass wheels, this is not in fact incorrect. The wheel farthest to the right belongs to the feed-screw shaft, which is not supported in the inner wall of the motor casing but is pivoted in an adjustable bearing in the removable outer panel; when that panel is detached, the nearer end of the feed-screw flops about freely and thus the brass gear is able to disengage from its neighbour. When the panel is in place the feed-screw is not free to rotate independently, so that clearly the gears are meshing correctly.

I did try (at half-past one this morning!) wedging part of an old pocket-watch mainspring under the hooked end of the pawl. This allowed me to wind the spring for a few clicks before it slipped out of place. I still believe that the principle was sound and that if I can fix the spring more firmly I may be able to revive the machine.

Oliver Mundy.

If I figured out correctly how this anchor is intended to work, the rounded edge on the right side is pushed down at each passage of each tooth of the gear, so that at the other side the spike is pushed inside the space between the teeth. When the gear revolves backwards, being the spike inside the space between two teeths, the spike will engage in a tooth and lock the gear.

If it is so, I think you should be able to circumvent those worn out teeths (which are definitely the issue) by adding some thickness on top of the right rounded edge of the anchor, as drawn in red in picture. By doing so, the spike should be pushed a bit more between the teeth. Personally I would add such thickness by adding a layer of bi-component glue ("liquid steel" as they call it). It's a five minutes work, and can be thinned if necessary, or removed if proven to be useless, so definitely worth trying.

Alternatively, you can also craft from scrap metal a new anchor, with a more pronounced left spike.

Just thinking, but maybe a thin washer under the "rocker" would lift it enough to engage the less worn area in the middle of the coarse winding gear, since the wear appears to be most pronounced near the edge...

Curt A wrote:Just thinking, but maybe a thin washer under the "rocker" would lift it enough to engage the less worn area in the middle of the coarse winding gear, since the wear appears to be most pronounced near the edge...

The problem with this would be that the clearance between the pawl and the panel behind it is very critical. The retaining-screw has an unthreaded portion of shank whose length fractionally exceeds the thickness of the pawl, so that the latter can still swivel freely when the screw is fully tightened. Anything placed behind the pawl would cancel out this clearance and the pawl would be jammed fast against the panel. Also, the ratchet and the pawl are both very solid items and of similar thickness – 3 or 4mm, or approximately ⅛ to ⅙ inch – so that the pawl already spans the whole thickness of the ratchet teeth and there is no unused portion of the latter to be brought into play.

Simple though it appears, there is a surprising amount of thought in the design of this little machine!

My task for tonight will be to look at the operation of the corresponding parts on my Columbia Q (this having an almost identical arrangement, but the other way up and visible edge-on through the open space between the side-plates) so as to verify exactly how the two components in question interact during winding. I am still not quite clear which end of the pawl performs the locking function; I thought it was the hooked end, but Marco has prompted me to reconsider.

Oliver Mundy.