I wonder sometimes about how one brand or model of pocket watch can be "more accurate" than another. the watches all have the same basic parts put into the same basic cases so how is this possible? of course we know about elinvar and elinvar plus, but if 2 different brands have matching parts like that wouldn't they run the same if adjusted the same way?

Hey Joe,I think it is as much about the condition of the movement,and as they say "what day of the week it was made",as the quality of the watch. Also I think watches are kinda like cars,in that you can have two that are exactly the same,and one will be a good one,and the other will be nothing but trouble. I have had dollar watches that kept railroad time,and have also had Railroad watches that did not keep good time at all. I now have several New York Standard watches that are dead on accurate,and I also have a nice old 18 size Waltham 17 jewel model 1883 that is driving me crazy trying to get it to keep good time. It has been cleaned and adjusted,and will run perfectly for a few days,and then will start to gain time,then after a few more days it will start to lose time. Bottom line is,as you said they are all supposed to be good timekeepers,but some just won't do it.

Joe,
If you had two brand new comparable RR watches from different manufacturers and they were in the pendant up position and wound at the same time daily they would probably be very accurate. If you take those two watches and let two different people use them one may end up fast and the other slow based on the particular persons habits. Even if you had two watches made by the same manufacturer one consecutively after the other they may have to be regulated to the individual users habits. As far as adjusting two watches the same that would be very difficult, even when new there were tolerances and specifications but it's unlikely all watches timed the same throughout different positions. I would think a Hamilton 950B should keep time better than a 992B if both were brand new because I'd bet Hamilton had higher standards for their premier watch. A RR watch in GRO should be able to meet +/- 30 seconds per week if regulated for the person wearing it. Check out this link.
Hamilton Technical Data; Subject: Regulation

Here's the 992B Specification sheet.

I wonder what the 950B specifications are?

RR

thanks guys. I do remember reading somewhere that a watch will vary from person to person depending on each person's habits. any way you look at it they are a nice piece of work that we all find quite interesting.

What do collectors generally consider to be acceptable accuracy for vintage and antique pocket watches?

Part of it is that for each part there was a specification and sometimes depending on skill and luck a watch may have most of the parts closer to specs and would be more accurate. This is the same principle for many car engines of the 60's and 70's some engine even though everything was the same, one would run much better, by pure luck one engine was more in spec than the other.

If you judge from eBay "Collectors" you are tempted sometimes to suggest they go back to the factory and buy a new one.

Personally, it depends what it is, why I bought it, and how much I pay for it.

Aside from "stupid" buys, I look for non-"frankenwatches" that are running to be within about 5 - 10 minutes a day when in "unclean" running state, <2 minutes a day when "cleaned", and , <10 sec a day if "railroad" grade when properly timed and regulated.

That all said, a "blueprinted" 7 Jewel 12s Elgin I have here runs RR accurate all day long (for the past 2 weeks) in pendant up position, and all too many over-abused RR Grades will NEVER be that accurate again unless completely overhauled with new Jewels, fitments, compensated balance wheel/Hairspring assembly, and motor parts.

I worked exensively with Detroit Engine plants, Most specifically, Ford and GM, in the late 60's and early 70's and they admitted that by 1973 their quality finally rose from 10% to a level of 80+% "Blueprint engines", mainly due to the very stringent emission fines ($3000.00/vehicle in CA)they paid for each vehicle that failed a field test. I can tell you stories about that time . . . .

Then of course there is Harley Davidson . . .

Dave on the topic of Harley when they were purchased in the mid-late 70's by AMF, the "Bowling" company that was in many ways the best thing that happened to Harleys, there was a big push in relation to quality control and statistcal process control and the quality of the bike improved greatly. Used to be in the 60's with the vertical split cases that poor quality would mean that most harleys would leak and drip oil. AMF made great strides in quality and it showed although the diehards could never come to terms with the AMF next to Harley Davidson. My 99 Tour glid does not leak a drop, unless it is the operator changing oil. For the most part with all the messing with that most watches probably have seen that it is amazing that many keep good time. As with anything restoring them to the intended specs (blue-printed) would make a world of difference.

Correct me if I'm wrong, but isn't it true that true interchangeability of parts wasn't fully achieved until the advent of the Hamilton 992B? Given that fact, weren't earlier watches prone to wider design tolerances and likely to exhibit individual variances from the overall spec, if not from the minute they rolled off the line, then later as they became worn, were serviced, repaired and etc?

Chris I am not sure part interchangability would matter since each part has a spec tolerance +/- so even with the 992B each part and screw had a spec tolerance. I doubt that one screw for a watch was spec'd extremely differently than another screw. More the point we were talking about is if you had a watch movement that had a greater percent of the parts close the the middle of the spec then it would be closer to blueprint. Another movement may have some parts closer to the limit on the plus side of the tolerance and some on the negative side of the tolerance spec but even with the 992B the parts would suffer the same problem. So any part not only screws but the train, axis wind wheels etc would fall into this area. Consider the balance staff, if it was just a bit thicker (+ spec) in diameter it would have a greater chance of increased friction from the jewel, the jewel hole on the negative side of the spec then even more possibility of additional friction even if just a bit more. The balance staff length a bit short, more slop, a bit long etc.

Interesting. I see what you're saying. Am I missing something, though? If earlier watches weren't completely interchangeable, wouldn't that mean that the overall design tolerances were generally looser? If they were as tight as what we see in the 992B, then wouldn't they also be just as interchangeable? I guess what I'm driving at is that interchangeability also implies stricter design tolerance and therefore increased watch to watch reliability and accuracy. Or am I stuck in a semantic trap?

I would think the craftmanship in the late 1890's up to the early 1930's was probably hands and shoulders above when the 992B was produced. In the early time frame labor was a much lower part of the overall cost of the watch and the materials was probaly a larger portion of the cost so waste was more of an issue. If you look at the cost of the watches produced early on the cost was much higher for the average worker than the 992B was, the 992B was selling in a fast disappearing market. In the case of the 992B more parts (screws) were made to be used in multiple locations thus reducing the need for different part numbers. I am not sure these screws were produced at a tigher tolerance only that the same type of screw could be used in more locations, which itself is more of a cost cutting matter. I have a tendency to think if the watches in the late 1890's and early 1900's had access to the special alloys for the balance, hairspring and mainspring I tend to think the overall accuracy of those watches would have been better than the 992B or the 571 BWR's . I am sure that will evoke some heavy debate. From the mid-50's on the cost of labor was a larger portion of any good produced so quality started slipping badly.

I tend to agree with Claude about the craftmanship being better back then. today the technology is obviously more advanced, but I think that the feeling for making something with true hands on craftmanship is lacking.

I have to agree with Claude too for the sake of the amount of time spent fitting the watches to the most precise settings in the earlier issue machines. Remember, these are NOT power transmission systems like our car transmissions, but rather a pinion "leaf" the "rolls" across the teeth of the wheels in a "frictionless" fashion. That said, the allowance for pinions and wheels was pretty large, as much as 20% (+/- 10%) geometry difference would make little change in the "friction" or "accuracy" of the watch.

What WAS done in the earlier models was a more detailed alignment of the pivots jeweled or not, and that does make a difference in the watch accuracy.

Later movements got around this by their materials and geometries, not anything more "accurately machined". In every case, the Jeweled watches that worked the best were clearance "fitted" to their individual pivots, expecially for the balance wheel.

Those are my opinions about this.

As for Harleys, I was struck by the "AMF Effect" when they changed the earlier forged aluminum triple trees to cheap Japanese sand castings for the Sportster. I saw one split on a friend's bike dumping him and his wife unceremoniously on the street for about 200 yards, not a pretty sight!.

In 1972 I changed to Kawasaki (900Z1) and then a Honda (V65) until H-D came out with the "evo" in the 1980's, which still used the 1920's "dribble down oil system" and a sump pump, only about 60 years behind the rest of the world.

Dave my first bike was a 81 Yamaha 550 Maxim, 3 yrs and 36K miles later I purchased a 84 Honda V-65 Magna, now I have a 99 Tour Glide Twin Cam 88 and a 89 Yamaha FJ1200 when I get the urge for some speed. The Kaw 900 was probably the first true superbike, the Kaw 500 and 750 2 stroke triples were monsters but with wimpy frames the a power band that was peaky as heck I place the Kaw 900 as being the true super-bike. On the watches I think in the early days when labor was cheap they could be more fit-intensive.