Wavy-trace Luch

[JohnC]

Hi all,

I have a Luch 2209 which has had the grave misfortune of being my first project commenced (although not completed) after completing Mark's course. It's been on the back burner while awaiting a scrap movement. Here is the problem.

After cleaning, lubrication, and demagging, the watch has a very wavy trace. Decent amplitude (280-300 at peak) and a clean trace, but constantly changing rate. It almost never settles down - amplitude rises to 300 and falls to about 250, with the rate changing by perhaps 30s/day. There is no consistent pattern - sometimes down then up, sometimes down then flat then down again.. About the same in dial up and dial down. No consistent time interval between shifts. Hairspring is perhaps not perfect, but not badly out of poise. Beat error is around 0.9, so close enough with a fixed stud.

The watch is definitely not magnetized. I have cleaned it twice and inspected all the parts closely. The only problem I could find was a worn barrel arbor bearing, so I swapped in a new barrel bridge from a donor. No change. I thought the mainspring might be tired, so I cleaned the spring in the donor and swapped the barrels. Nothing.

I'm pretty well out of ideas at this point, so I guess I'm just going to keep swapping parts until I find the answer! If anyone has ideas, I would be very glad. Many thanks.

[JohnR725]

always nice to give us a picture. Then when you do give us pictures dial up and dial down and crown down might be nice also.

[JohnC]

Disassembled to inspect the train again, but I'll reassemble and post pictures tomorrow. Thanks.

[JohnC]

First 3 are dial down. Oscillation between about 220 and 270 degrees amplitude. As you can see, it’s not a rapid oscillation, more a slow, never-ending adjustment.

The next 3 are dial up, pendant down and pendant up. As you can see, the average rate varies considerably between dial down and any other position, but not that much variation between any of the other two. The magnitude of the amplitude variation also seems to be slightly less in positions other than dial down.

Last night I replaced the wheels in the upper train (see http://www.ranfft.de/cgi-bin/bidfun-db.cgi?10&ranfft&&2uswk&Luch_2209). The problem appeared to persist, although the replacement parts were less clean and the trace was messier. The last really suspect component is the lower train, which has some components (in particular the center pinion) which strike me as high-friction.

Overall the rundown of the train train with the pallets out is less smooth than I would like. 

Edited September 5, 2020 by JohnC
Add photos

[rodabod]

There is quite a difference in rate between dial-up and dial-down, yet not a significant difference in amplitude. Are you sure that the hairspring is breathing healthily - clean, no sticking, fouling, and the stud height is good?

[JohnR725]

I found the link with some history on your watch. I find it a little bit troubling history especially when I try to find a picture of your movement online because it appears to be there were variations in this watch. plus it's a Russian watch and their manufacturing capability for watches has been interesting.

 

2 hours ago, JohnC said:

Overall the rundown of the train train with the pallets out is less smooth than I would like. 

 

2 hours ago, JohnC said:

The problem appeared to persist, although the replacement parts were less clean and the trace was messier.

one of the problems with the modern timing machine is that you only get a tiny window of display. One time with paper tape machines if you had gear train problems and you run out enough paper tape sometimes you can see the problems. In other words like if one of the pivots was bent this would cause a power fluctuation that would repeat and you can see it on the paper tape. There is some really newer expensive Swiss machines they can run time plots which displays amplitude and rate differently and that we usually show you the variations in the powertrain.

The problem with variations in the powertrain or as you noted less smooth than you would like that will cause running issues exactly what you see on the timing machine. Then swapping with parts that need to be cleaned as you can see causes a messier trace because power through the gear train needs to be as smooth as possible plus you shouldn't have any power losses through the gear train even though the gear train itself will always have power losses.

One of the problems with randomly swapping wheels around is if you don't know what the problem you're fixing then bringing in new components that may have different problems makes for a really interesting problem solving issues. It looks like from Russian manufacturing that if there's variations which they appear to have swapping wheels may make things even worse because you may end up with incompatible components even though they may supposedly be the same.

So you're going to have to go through and check every single wheel and see what the end shake is words that should've a little bit a play how free is that wheel checked the pivots for defects you have to figure out where your powertrain fluctuation is an fix the problem because otherwise your timing will never be consistent.

It may be possible if it's only one problem if you watch the rates go up and down and see if it would follow up had her like every so many seconds it goes up and down. Figure out where the or the rate of repeating of the problem we can trace it back to a specific wheel.

https://www.2209watchmovement.com/history

[JohnR725]

I'm attaching something out of a timing machine manual for paper tape. One of the nice things with liquid crystal screens as they can imitate paper tape machines they just can't imitate the long running time. So here the suggestion is just run lots of tape which you can't do but you will see what the effect of power train fluctuation is

[JohnC]

Thanks to you both. Rodabod, a picture of the balance is attached. Not sure you can make it out. I don’t think the spring is perfect, but nothing is stuck together. Full disclosure, I bought a donor that was actually running in order to get the balance complete. On the original I messed up the hairspring trying to put it in poise. However the wavy trace issue was preexisting, and I was hoping to tackle that before the positional variation.

Second, John, you are definitely right about the Russian calibers. I think part of the issue is that their caliber numbers are not caliber numbers, but actually describe the movement’s capabilities - thus for example 2602 just means manual winding, no shock protection, subseconds. Seems like a great system but useless for identifying generations that may have succeeded each other over decades in production.

That being said, I am quite sure that both movements I have here are first-gens, and likely to be compatible. And lastly, I think (hope) I have identified the issue, and sadly I think it is fatal to the original movement.

The issue is, there is way too much sideshake in the barrel. Going back to look again, I can actually see score marks in the original barrel bridge because under torque the barrel is leaning over too far. Even after using the donor bridge, the slop is still too much, because the bottom bush is also out of round. I cannot find any other problem with the train, although I would add that I did try swapping out the lower train and the free running was nicer.

I don’t have the knowledge or capability to fix these bushings, so sadly I think the original movement is toast. Fortunately, I have a donor that was running before I stripped it for parts. So I think the plan is, the donor becomes the donee. 

 

[JohnR725]

40 minutes ago, JohnC said:

The issue is, there is way too much sideshake in the barrel. Going back to look again, I can actually see score marks in the original barrel bridge because under torque the barrel is leaning over too far. Even after using the donor bridge, the slop is still too much, because the bottom bush is also out of round. I cannot find any other problem with the train,

I've had this issue before on pocket watches.. In one case it had the wrong barrel  and using  a different timing machine with software I could see what the problem was.  But on the regular timing machine I couldn't tell what the problem wasn't all that was all over the place like yours.. Even though  the manuals the paper tape machines say if you run along enough plot you can figure it out well the center wheel meshing with the barrel occurs at least every five minutes  see what the paper tape I wasn't seeing it.. But only with the software this other machine when around the time plot good I see the problem..

If the barrel in the center wheel no longer mesh properly  as they  teeth come together they will bind up at least they did in my case and on a time plot you can see this slow increase of a problem and then as the tooth past it released and that was an incredible screwup for timekeeping. What you're going to have to do is close the holes that the arbor goes into and you do need to figure out whether it's the plates themselves and or the barrel. Because the arbor might be loosen the barrel equally as bad.  But until you get the barrel stabilized so it's not wobbling around you're never going to get this watch to keep time or even run very well.

[JohnC]

Yes, I think I’m going to just have to use the donor plates (and the donor everything) because it’s beyond my capability to close the bushes. You are dead right about the short display on the timing machine being a problem. A real innovation would be a machine that would save the readout over a preset period, e.g. 5 mins, and export it to a computer. 

[JohnC]

Another fun fact: despite being first-gen Luch 2209s, the two movements are distinctly different shades of brass. Says something about the consistency of raw materials in the USSR!

[JohnC]

@JohnR725 @rodabod the donor movement is in and running nicely. Traces are still a bit wavy, but good enough for a 50 year old Russian watch. Thanks for your help!