rewolf wrote:I bet none of the eraly digital watches has a stabilized power supply for the oscillator.
What is a repeater?
That is the part that keeps giving the electric impulses to keep the quartz cristal oscilating.
It is part of an 'quartz oscilator' but it is the bit that mákes and keeps the cristal oscilating.
The crystal merely resonates in response.
In the 1978 manual the stabelising of the power supply to the oscilator (cristal and repeater) and devider is mentioned as a mark of quality, as a difference between well made and compromised by cost.
It does not list which ones were which.
Ditto a trimmer being present and set, ditto the use of a pre-aged cristal.
For me it answers why early quartz watches which were not made penny wise can be specced as 3 or 5 secs/month.
Quite a few early quartzes have a provision for temperature compensation too. This dissappear from all but the ones at the very high luxury end as soon as quartz watches become cheaper in the second half of the seventies.
Non tc, non stabelised, not trimmed but véry cheap and 15-20 secs/month becomes the standard.
Understandibly as it still is 30 times better than a good (if not chronometer) mechanical and at a very much lower cost.
The temperature deviation graph of quartz you gave is indeed typical for a the current shape/frequency cristals used.
The block (sometimes bar) cristals used in some early oscilators are far more stable but cost considerably more (to cut precisely and 'trim') and are some 20 times more voluminous, The minute laser cut and trimmed tuning fork shaped cristals were very much cheaper to mass produce and since wrist temp. is pretty stable any, thus performance good enough, these became omni-present.
The whole key to the performance differences between early and late seventies onwards quartz is the production cost versus 'good enough'.
In the early years the possibilities of the technology were being explored and cost more a result than leading.
'Good enough' at at the sharpest price became the only norm with VERY few even more expensive exceptions.
The cheap image of electronic digitals killed off the high end modules alltogether.
The Seiko 0634 was the last of the tc ones whereas the analogues maintained two, three tc quartzes at the very high end.
It sééms that the Seiko S760 and S760 are a step towards well made solid state electronic digital quartz modules.
The stand alone performance indicates good structural quality of both the oscilator and the module. I have however nó, NO, concrete data about their built.
Just by coïncidance my EPD was above 50 degrees: in fault mode, yesterday and today for some time and was not synched last night.
I am going to keep is from synching again tonight and check the deviation tomorrow.
Several hours at 60 - 70 degrees shoúld give a measurable and extrapolatable deviation.
Freezing to just below -5 for a period of two days was already on the list as that gives a bigger and thus better measurable deviation
05 Mar 2011, 00:01
that the answer to why the Seiko EPD has such good stand alone performance wihout being tc was to be found in a 32 y.o. watchmaker´s schoolbook; it simply is a properly made quartz timer.
Just proper trimming I guess.