On Wednesday, July 6, 2011 6:27:00 PM UTC-7, Arfa Daily wrote:

However, just because a design is 50 years old, and was 'elegant' in its
concept, that does not make it necessarily 'good', and by today's standards,
a design that purports to be an extremely accurate timepiece, but which runs
at an incorrect speed when the batteries are replaced, probably legitimately
counts as being 'crap'.

Not really. The Accutron power supply was specifically a stable Hg cell,
and it isn't 'replace' that makes it run awry, it's 'substitute nonstandard
supplies' that causes the issue. There's only so much compensation you
can do with a total semiconductor complement of one transistor.

"whit3rd" wrote in message
...
On Wednesday, July 6, 2011 6:27:00 PM UTC-7, Arfa Daily wrote:

However, just because a design is 50 years old, and was 'elegant' in its
concept, that does not make it necessarily 'good', and by today's
standards,
a design that purports to be an extremely accurate timepiece, but which
runs
at an incorrect speed when the batteries are replaced, probably
legitimately
counts as being 'crap'.

Not really. The Accutron power supply was specifically a stable Hg cell,
and it isn't 'replace' that makes it run awry, it's 'substitute
nonstandard
supplies' that causes the issue. There's only so much compensation you
can do with a total semiconductor complement of one transistor.

But that's not what was being said. The actual point that I was commenting
on was where a reference was made to a note in the user guide regarding the
specific condition of it running fast when a new (correct type) battery was
fitted, until its terminal voltage fell to the main life voltage. It is this
undesirable characteristic that causes the problem, if you try to use
alternative chemistry cells that have a *permanently* higher terminal
voltage than the original.

Arfa

Arfa Daily wrote:
Not really. The Accutron power supply was specifically a stable Hg cell,
and it isn't 'replace' that makes it run awry, it's 'substitute
nonstandard
supplies' that causes the issue. There's only so much compensation you
can do with a total semiconductor complement of one transistor.

Let's not forget that the concept of a three terminal voltage regulator,
which makes so many things possible these days was yet to be invented.

In the early 1960's a voltage regulator had to be built opon either a
zener diode (which required a higher battery voltage) or a gas filled
"vacuum" tube.

Today if you want 5 volts, you use a 7+ volt power supply and a three
terminal regulator. If you want 3.3 you use 5 volt supply and three
terminal regulator.

If you want three volts from a battery, you use a lithum cell, but it
"wanders" from around 3.7 volts off the production line, to around 3.0 volts
when almost exhausted. Most of it's life it produces 3.3 to 3.5 volts
depending upon the exact chemistry, load, age, etc.

In 1960, you had no such luxury. I'm sure they existed sooner, but I never
heard of a Litium battery until Canon started to use them in their EOS cameras
in the late 1980's.

Also note that a modern digital watch uses a more simple in concept circuit,
it just has a crystal oscilator that runs at 3.57mHz (the old NTSC color
burst frequency) and counts 3,570,000 (or whatever the exact number is)
clock pulses and moves the second hand.

Easily done with modern CMOS and SMT. Not easily done with Sputnik level
technology.

I had at one time a clock that worked that way, except with a lower
frequency oscillator (32kHz?) made by MacKay Dymek (part of HP and
their logo was an upside down HP logo).
It fit a 19 inch rack, was at least a foot tall and used Stroeger switches
to count.

You could hear it change the hour two floors away.

Geoff.

--
Geoffrey S. Mendelson N3OWJ/4X1GM
Making your enemy reliant on software you support is the best revenge.

"whit3rd" wrote in message
...
On Wednesday, July 6, 2011 6:27:00 PM UTC-7, Arfa Daily wrote:

However, just because a design is 50 years old, and was 'elegant' in its
concept, that does not make it necessarily 'good', and by today's
standards,
a design that purports to be an extremely accurate timepiece, but which
runs
at an incorrect speed when the batteries are replaced, probably
legitimately
counts as being 'crap'.

Not really. The Accutron power supply was specifically a stable Hg cell,
and it isn't 'replace' that makes it run awry, it's 'substitute
nonstandard
supplies' that causes the issue. There's only so much compensation you
can do with a total semiconductor complement of one transistor.

Your correction is not correct. The mercury cells did start out at slightly
higher voltage. It a period of operation (several days I believe) until the
voltage fail to its stable point.

"Geoffrey S. Mendelson" wrote:

Arfa Daily wrote:
Not really. The Accutron power supply was specifically a stable Hg cell,
and it isn't 'replace' that makes it run awry, it's 'substitute
nonstandard
supplies' that causes the issue. There's only so much compensation you
can do with a total semiconductor complement of one transistor.

Let's not forget that the concept of a three terminal voltage regulator,
which makes so many things possible these days was yet to be invented.

In the early 1960's a voltage regulator had to be built opon either a
zener diode (which required a higher battery voltage) or a gas filled
"vacuum" tube.

Today if you want 5 volts, you use a 7+ volt power supply and a three
terminal regulator. If you want 3.3 you use 5 volt supply and three
terminal regulator.

If you want three volts from a battery, you use a lithum cell, but it
"wanders" from around 3.7 volts off the production line, to around 3.0 volts
when almost exhausted. Most of it's life it produces 3.3 to 3.5 volts
depending upon the exact chemistry, load, age, etc.

In 1960, you had no such luxury. I'm sure they existed sooner, but I never
heard of a Litium battery until Canon started to use them in their EOS cameras
in the late 1980's.

Also note that a modern digital watch uses a more simple in concept circuit,
it just has a crystal oscilator that runs at 3.57mHz (the old NTSC color
burst frequency) and counts 3,570,000 (or whatever the exact number is)
clock pulses and moves the second hand.


The NTSC burst frequency is 3,579,545 Hz.


I've never seen a watch with the color crystal. CMOS uses more power
at higher frequencies, which is why they use a 32,768 Hz crystal.
fifteen biniary dividers gives you the one second pulse needed at the
lowest power requirements.


Easily done with modern CMOS and SMT. Not easily done with Sputnik level
technology.

I had at one time a clock that worked that way, except with a lower
frequency oscillator (32kHz?) made by MacKay Dymek (part of HP and
their logo was an upside down HP logo).
It fit a 19 inch rack, was at least a foot tall and used Stroeger switches
to count.

You could hear it change the hour two floors away.

--
It's easy to think outside the box, when you have a cutting torch.