View Single Post
Old 16th October 2012, 04:27 PM   #34
Beanbag's Avatar
Join Date: Jun 2003
Posts: 3,468
As a certified horologist -- and probably the ONLY person in this forum who's actually REALLY worked on a Bulova Accutron movement -- please allow me to make a few comments.

The Bulova Accutron is a remarkable development in horology. It was a stupefying leap in wristwatch timekeeping. For once, you could wear a timepeice on your wrist that when properly maintained and adjusted could keep more accurate time than a bench or wall clock. Indeed, one of the biggest issues Bulova dealers had was when people came in claiming their Accutron wasn't keeping correct time. The typical issue was that whatever standard the owner was comparing their watch to wasn't keeping correct time. Just about every store selling Accutrons would keep a short-wave radio set to WWV so they had a reliable time standard for the client to compare their watch to.

That said, I can't for the life of me imagine why anyone attempting measurements in inertial field detection would go with an Accutron movement as part of their detection scheme. From a '60's Steampunk aspect, I admit there's a certain appeal to using such a unit, but from a standpoint of reliability, precision, and accuracy, there are MUCH better standards or references to use.

First, the Accutron is an electronically-driven MECHANICAL device. The electronics in it are extremely simple, on the order of what a ten-year-old could reliably assemble: a single transistor, a couple capacitors, several resistors, and some small wire coils. NO integrated circuits, nothing fancy, just discrete components soldered together.

The vibrating tuning fork drives the display mechanism through a ratchet-and-pawl assembly directly attached to one of the vibrating tines. This is a mechanical connection, one subject to wear, damage, lubrication issues, and requires precise adjustment to make it work correctly. It's analogous to having a small reciprocating engine driving a plunger back-and-forth 360 times a second, knocking a ratchet wheel forward one notch for each cycle.

An Accutron tuning fork DOES NOT come out of the box at the perfect frequency. It has a set of regulators – moveable eccentric counterweights – that allow you to tweak the fork into the correct frequency. The system is also magnetically-driven, originally designed to use a 1.35-volt mercury battery to drive the tuning fork coils. What you have is a system designed to operate at low magnetic field intensities; therefore, it's also sensitive to external magnetic influences as well.

For example, a direct quote from the Bulova Accutron Service Manual Series 214, page 32:
“The ACCUTRON hour and minute hands are steel. On some models, depending on the length and design of these hands, the tuning fork rate changes slightly when the hands are set near the tuning fork magnets. For this reason, it is good practice to set both hands in the lower portion of the dial before measuring the rate of an ACCUTRON timepiece with the rate recorder.”

Mind you, this is with just the hands – slivers of steel with less material than a sewing needle. The manufacturer tells you they are likely to jack with your reading. While this would be disastrous for taking “instantaneous” measurements, for regular timekeeping purposes this is NOT an issue, because over the long run the errors will average each other out. If you're measuring intervals from one “tick” to the next, however, you might as well be using a Slinky ™ as a ruler.

I like Accutrons. I think they're neat. I also think they're cranky, irritable bitches that require regular service by a trained technician who knows what they're doing, has access to the service manuals and the specialized test equipment and fixtures needed to maintain them, AS WELL as access to replacement parts. Take this to heart: A) you don't find that many trained Accutron watchmakers in the world these days; B) the manuals can be found if you dig long enough (I photocopied a set from watchmaking school); and C) while you might find some of the test gear, replacement parts are GONE. Period. Finito. Maybe some old geezer watchmaker has a stash of index wheels (the part most commonly worn out) in their bench, plus a few other expendables, but there is NO place where you can send off an order and expect to get something back other than a “you gotta be kidding” note.

The Accutron magnets are also 60's technology, namely alnico (if I remember correctly). They are KNOWN to lose their magnetism over time. Bulova mentions this also in the service manuals. Their solution? Replace 'em. But you can't get 'em any more.

Based on these shortcomings, let me ask this question: who in their right mind would base an experimental apparatus to detect such a minute phenomenon around a detector that hasn't been produced in well over thirty years, with no available parts for refurbishing what would have to be a used unit, few (if any) people available to work on it, and has such well-known sources of error?

A few other issues come to mind. I don't know if the transistor used is germanium or silicon. Because of the low operating voltage, 1.35 volts, I would guess germanium because of the lower band-gap voltage. Regardless, either germanium or silicon, there is NO temperature compensation in the driving circuitry. None. Change the temperature of any transistor (germanium or silicon), and you change the operating characteristics. And with germanium, the temperature effects are greater. It wouldn't be a problem in a wrist watch because, surprisingly, the temperature inside the metal watch case remains fairly constant, since it's against a warm body nominally around 98 degrees F. Built in to some test apparatus, however, it's hard to say what the temperature would be. Any sign if the Accutron movement was inside a temp-controlled housing?

Because the tuning fork is mechanically-linked to the rest of the watch drive train, problems with the movement “downstream” will actually be reflected in the action of the tuning fork. If, for say, one of the gears (actually called wheels in watchmaking, even if they have teeth) further down in the train has a worn or defective tooth, that will cause a regular disruption in the driving system every time that tooth rotates around and engages another wheel, causing the preceding wheels to bog down, causing the index wheel not to move as freely, thereby “fighting” the next impact from the pawl on the tuning fork arm, causing a change in the normal free vibration. The tuning fork is not a “free” mechanism, but interacts directly with the rest of the movement by physical contact.

No, an Accutron isn't even a decent choice for a detector. I think they chose it because it hearkens back to the Good 'Ol Days and lends a character of common sense and when Right was Right. It's nostalgic. However, there are MUCH better detectors that use the same principle, like a quartz crystal. Drive it electrically, but you can measure its frequency WITHOUT having to connect to it mechanically, say with an acoustic sensor, so you're not influencing it while taking your measurement.

You get more readings as well. With an Accutron, you get 360 per second. With even a modest crystal used in a cheap quartz watch, you get 32768 “readings” a second, two orders of magnitude better.

In short: In my opinion, this guy's study was built on a foundation of Silly Putty.


Nothing divides an indivisible nation quite as well as religion.

Know god, no peace.
No god, know peace.

If Jesus is the answer, it must be a real dumb question.
Beanbag is offline   Quote this post in a PM   Nominate this post for this month's language award Copy a direct link to this post Reply With Quote Back to Top