After all the exasperation this thing has caused, I just had to tell someone that I've finally got the damn thing running! Someone who wouldn't roll their eyes at me.

The clock dates to around 1934. She's a big beast, about 30 inches high by 18 and seemingly incredibly simple. Mains powers a coil; a magnet spins around inside; that turns a worm drive; in turn that rives the centre wheel and ... that's about it.

The worm drive was worn. That needed some fettling. The friction fits lacked enough friction, so they were played with. Eventually, after about three evenings on it over the last month, we have her running. There was no rush because it took a while to make a new back frame and apply a dozen coats of shellac. The original frame is a simple metal one that seems to have been painted in every colour from black to red over the decades but needs something big to support it and raise it away from the wall it will eventually hang on. More that a few times, I've told the clock that if it doesn't buck it's ideas up, I'll put a quartz in it.

Anyway, she's now running but seems to be gaining about half a minute an hour. There's no mechanism for adjusting the timekeeping on this. Does anyone have a clue as to how you regulate one of these? It's a Smiths De Luxe.

 

I love that era of technology. So many beautiful techvolutionary dead ends and offshoots.

I don’t suppose regulation would come from purely electric means? Adding voltage or something of the sort?

 

No, I had the same thought at first. This seems to run at the same rate on 200-250V and runs in the same direction regardless of polarity.

I've had the movement running (balanced on top of the Seiko) for more than a day and it seems to have settled down. I'm going to case it this weekend and see how it goes.

 

Electric wall clocks are often unregulated, and simply obtain timing off the line (mains) frequency. That's going to drift a little bit depending on the load being placed on the power grid:

(thanks)

 

That's damn near brilliant, in so many ways, from a design point of view, usability and so many other things.

It also explains the deviation I saw and how the mechanism seemed to rectify itself. I'm much more comfortable recasing it now. Might be later in the week but I'll post a photo of it reassembled.

 

If it's gaining 30s/hr, that's much more of a deviation than I'd expect from the mains; the graph from @Vitezi shows deviation of 1s/day and even that is high compared to the traditional regulation, where they would have some variation through the day (due to load balancing with supply), but the electric company would bring it back to alignment with an atomic clock reference overnight, so day-to-day drift would be zero. This was important when every official clock ran this way on this standard. However I wouldn't be surprised if the grid operators no longer hold themselves to this standard, now that fewer clocks rely on it. Still a variation of ~1% seems quite high.

 

True, but add to that the fact that I live in an old house and it was pouring with rain and ... hmm ... maybe I need an electrician.

Seriously, though, the movement settled down after running for a while.

 

She's alive!

I'm still to put her into the new frame (has to be spaced off the wall) ... just to make sure she continues to run as she should. Thanks for the advice.

 

She's nearly complete, folks ... just waiting for a little paint to dry before the last dozen screws go in. I bet you can't even see the blu-tac!

 

Weeks of works (well, a fair few evenings over the course of a few weeks); it's all working perfectly; dropped it in for PAT testing this morning ...

And, yes, that is inside the glass