I like to wind my watches twice a day, morning and night. Occasionally, I forget and this way they're less apt to run down and stop.

I realize this causes the watch to be powered by a tightly wound spring most of the time. Does this cause the watch to run significantly faster than it would if it was only wound once a day?

Hi Christopher,

Most watches are adjusted for isochronism that allows for change in power.
over the years improvments in design mainsprings, hairspring/balance etc have grealty reduced this error.

Christopher, here is a link to an interesting discussion regarding this subject....thought you may be interested in reading it....

When to wind your watch.

Good Luck....

Regards,
Jerry

In my experience some watches run faster when they are near the low end of the "wind-up". I postulate that is because the fully wound-higher torqued escape wheel will actually "kick" the balance wheel harder, causing a slightly longer "swing" or period of oscillation each way causing a slight slowing down, at low wind, the lower torque "kick" gives a reduced period of oscillation making the rate a little faster.

This could be excaberated by a watch that is somewhat "out of beat" as well.

David,

I see the same thing with some watches running faster near low end of wind. I wonder though if the speed of oscillation with a good kick at full wind would be offset by the slower speed of the balance at low wind with less travel for the next oscillation? Isn't this part of the isochronism that would be calculated by the designer of the movement?

In m y opinion, two designs were dramatically capable of reducing this effect;

"Stopworks" on the motor restricted mainspring power to the watch to a shorter range of "wind". It stopped the watch before it reached the lower "speedier" end of the wind, and indeed it stopped the wind-up of the watch before the extreme high end of wind was reached.

True "motor barrel" design used on most all Waltham 16s movements from the 1888 model on drove the wheel train from the inner motor arbor.* Using the inner diameter of the motor spring reduces this efect quite considerably as the torque vector of the ID of the spring motor changing "with time" was less dramatic.

* Most of the other named "motor barrels" used by Illinois and Hamilton were actually "going barrels", meaning they still drove the wheel train from teeth cut on the OD of the main barrel and only used the inner arbor to rewind the watch.

Modern(alloy)mainsprings produce a much more uniform power curve, given the fact that most watches PW & WW are designed to run for 30-40+ hour of run time the lower power curve is not encountered when wound fully at the start of the days use, and every 24hr thereafter. Therefore any loss in mainspring power would occur during the period of least activity, overnight this reducing isochronism errors and those induced during normal carrying of the watch. Further isochronism compensation can be achieved by adjustment of the overcoil which allows for making timing corrections at upper or lower power levels to a point where for all purposes eliminated. Also the timing of a watch has to be considered as a average over a extended period of use not a single short period of recording its accuracy.

I wind mine at least 2-3 times a day, in many cases I don't know if they have a tired old spring or a newer alloy spring since some of the watches I have were purchased with the connation of "recently being serviced" and I found no need to tear them down to verify that as long as it ran and kept good time. Probably heat is the worst offenders to keeping consistant time, yes the balance wheel is designed to offset this but that does not include any expansion of the plates, axis and any effect heat/cold has on the oil being used (newer synthetic oils should not be affect as much). Jewels expand/contract at a different rate than metal, metal types expand/contract at different rates. The position also also will impact what results you see, different positions give different frictional bearing loads even if you look at factory specifications for 992 or 950's you can see some of the deviations as a result of the frictional bearing load differences. So when I wind if I am timing a watch I usually wind it, lay it on it's back or hang it on a hook by the bow. In the end with all of the mechanical problems a typical watch has to overcome to keep time I never stop being amazed on how accurate they can be.

I've pondered the effect of temperature change on mechanical watches. It seems to me that when a watch is carried, it resides in a pocket which is a few layers of cloth away from a warm human body. When the time is checked, it is removed from the pocket and cradled in a warm hand for a few seconds before being returned to the pocket. In hot weather and cold, the watch should stay close to the temperature of its owner.

The exception would be if it's placed on the nightstand in an unheated room overnight while its owner slept.

The clothes would insulate the watch from body heat or you could look at it two ways, one side is heated by the warmth of ones body and the other side is cooled or heated by the outside temperature. The typical person carries a watch with the crystal toward your body and the case back out (to protect the crystal) metal is a better conductor of heat than glass and in some ways part of the watch has two temperature zones. Second topic is not an unheated room but a room at normal temperature, typically around 72-75 where the watch would spend 8-10 hours (sleep time or down time) and the rest of the time during work at another temperature range (ambient or body temp). Now that would not even take into consideration many users that used a leather holster type arrangement like many trolley drivers did where the watch was in the leather case and it could be flipped up to see the time. Again temperature change is probably the worst offender to accuracy and the temp change cycle is is not always exact in duration.

I've always been fascinated by the fact that anytime I carry a watch in a case with heavy silver content, it gains and holds body heat like nothing else I own, especially 18s size cases.

I find it interesting to be able to pull one out in the winter and actually use it as a hand warmer, they get that warm in my pocket. Doesn't seem to affect the watch's time keeping capability one way or the other.

Regard! Mark

At one time Hamilton used a mid-point in the temperature accuracy process between the cold and hot temperatures but this was eliminated with the advent of the Elinvar hairspring and balance, that is the very reason that Hamilton and later others spent time and effort to find a material for the hairspring that was not affected by temperature changes, all for the purpose of improving accuracy. Sometimes look at the specifications for the high point and low point test temperatures and what is allowed for variations and this is a consitant temperature over a 24hr period and not cycling from room temperature, to body temperature, to outside ambient temperature which most watches are exposed. Maybe it doesn't affect watches but it appears that most brands spent a lot of research effort to overcome this either via testing /engineering or via exoctic materials or composition materials used for parts.