This project was built by a friend's older brother as a science fair project, many years ago. Just for fun, he recently decided to recreate and update it.
It probably won't fit on anyone's wrist.
It's an interesting "read", though---
Vacuum Tube Quartz Clock Part I
There were counter tubes with 10 neon bulbs within one tube made for this exact purpose. Perhaps sourcing some of those would make sense.
There's a gold mine of information on ring counters here: http://www.dos4ever.com/ring/ring.html
Tom
Last edited:
> divide by 32768 to get to a 1 second pulse. If done with digital flipflops this will need 15 dual triodes. Then you need to divide by 60 (twice) to get minutes and hours. That adds another 12 dual triodes.
The 1959 experimenter seems to get from 120KC to 60Hz, 2,000:1, in just four dual-triodes.
That 120KHz-6KHz (in one stage!) divider must be incredibly fussy. See attached.
If that performance could be done over several stages, your 32,768:1 could be done in four dual-triodes. Four more to do 60:1 twice.
Plus the master oscillator and buffer.
And we have not mentioned display drivers.
Bah. He did not say it had to be "accurate". A 1Hz R-C oscillator is trivial, and may not drift an hour a day. Still leaves the 60-twice and display problems.
___________________
> if it's unreasonable, try actually telling me
I will tell you it is unreasonable.
There could be a watch-maker turned tube-maker who could custom fabricate this. Although people make DIY tubes, the ones I have seen are bulky "gee wizz" bottles, not micro-miniature. It is pretty clear none of us are at that level.
The 1959 experimenter seems to get from 120KC to 60Hz, 2,000:1, in just four dual-triodes.
That 120KHz-6KHz (in one stage!) divider must be incredibly fussy. See attached.
If that performance could be done over several stages, your 32,768:1 could be done in four dual-triodes. Four more to do 60:1 twice.
Plus the master oscillator and buffer.
And we have not mentioned display drivers.
Bah. He did not say it had to be "accurate". A 1Hz R-C oscillator is trivial, and may not drift an hour a day. Still leaves the 60-twice and display problems.
___________________
> if it's unreasonable, try actually telling me
I will tell you it is unreasonable.
There could be a watch-maker turned tube-maker who could custom fabricate this. Although people make DIY tubes, the ones I have seen are bulky "gee wizz" bottles, not micro-miniature. It is pretty clear none of us are at that level.
Attachments
Anyone ever work on a electronic organ (the musical instrument) that used vacuum tubes or neon bulbs for the frequency dividers? Remember how fussy they were? It took me several months to get a 1940's vintage Baldwin organ working. It used 6SN7's for the dividers. The power transformer alone weighed about 20 pounds.
So far this thread has addressed only two of several issues that I see with a project of this magnitude, the frequency dividers and the reference.
In order to build a clock you need a reference. In years past this was usually the 50 or 60 Hz power line. It is not available in a battery powered device, so usually a 32,768 KHz crystal is used. Our Chinese indoor outdoor thermometer / clock uses a low frequency radio receiver to pick up WWVL, a radio broadcast frequency standard on 60KHz.
The reference frequency must be divided down to a 1 pulse per second clock which advances the counters each second. Some fancy clocks use a GPS module to provide a super accurate 1 pulse per second (1PPS) "tick." This allows skipping the reference requirement since it is built in to GPS.
Having a 1PPS tick, you will need a "minutes" counter and an "hours" counter. There were some specialized tubes made for this purpose 50 or so years ago but they are scarce today, and their magnets are often too weak for proper operation.
Assuming all of the above is built and working, how are you actually going to display the time. Years ago the only options were Nixie tubes, or custom displays with individual incandescent or neon light bulbs. LED's and LCD's did not exist in the tube era.
Another issue is the power supply. Each vacuum tube will consume 1 to 3 watts just to heat the cathode, and that's before any useful functions are done. 50 to 300 volts are needed to operate the functioning elements inside the tube.
SO:
Wristwatch, impossible.
Big box that's somewhat moveable, possible but unlikely without careful design and planning.
Something that works, but takes up most of a large bench top as a "proof of concept" (to prove that it can be done) doable, but still a complex design that requires some design and construction skill.
The digital wristwatch didn't become possible until IC and LED technology advanced enough to allow all the above functions to be stuffed into a rather large and clunky watch that ate batteries like my grandkids eat cookies! I remember the first Pulsar watch from 1972. A professor from the University of Miami was showing his off in the store where I worked in the early 70's. It cost about $2000 back then.
https://en.wikipedia.org/wiki/Pulsar_(watch)
So far this thread has addressed only two of several issues that I see with a project of this magnitude, the frequency dividers and the reference.
In order to build a clock you need a reference. In years past this was usually the 50 or 60 Hz power line. It is not available in a battery powered device, so usually a 32,768 KHz crystal is used. Our Chinese indoor outdoor thermometer / clock uses a low frequency radio receiver to pick up WWVL, a radio broadcast frequency standard on 60KHz.
The reference frequency must be divided down to a 1 pulse per second clock which advances the counters each second. Some fancy clocks use a GPS module to provide a super accurate 1 pulse per second (1PPS) "tick." This allows skipping the reference requirement since it is built in to GPS.
Having a 1PPS tick, you will need a "minutes" counter and an "hours" counter. There were some specialized tubes made for this purpose 50 or so years ago but they are scarce today, and their magnets are often too weak for proper operation.
Assuming all of the above is built and working, how are you actually going to display the time. Years ago the only options were Nixie tubes, or custom displays with individual incandescent or neon light bulbs. LED's and LCD's did not exist in the tube era.
Another issue is the power supply. Each vacuum tube will consume 1 to 3 watts just to heat the cathode, and that's before any useful functions are done. 50 to 300 volts are needed to operate the functioning elements inside the tube.
SO:
Wristwatch, impossible.
Big box that's somewhat moveable, possible but unlikely without careful design and planning.
Something that works, but takes up most of a large bench top as a "proof of concept" (to prove that it can be done) doable, but still a complex design that requires some design and construction skill.
The digital wristwatch didn't become possible until IC and LED technology advanced enough to allow all the above functions to be stuffed into a rather large and clunky watch that ate batteries like my grandkids eat cookies! I remember the first Pulsar watch from 1972. A professor from the University of Miami was showing his off in the store where I worked in the early 70's. It cost about $2000 back then.
https://en.wikipedia.org/wiki/Pulsar_(watch)
Regarding cold-cathode counting and indicator tubes, see this link for some special Philips counting and indicator tubes exhibited at the 40th anniversary of the NVHR (Dutch association for the history of the radio):
40 jarig jubileum NVHR
or this link for a large number of dekatrons, which together form a reliable but slow computer exhibited at the British National Museum of Computing:
First generation - WITCH & EDSAC | The National Museum of Computing
40 jarig jubileum NVHR
or this link for a large number of dekatrons, which together form a reliable but slow computer exhibited at the British National Museum of Computing:
First generation - WITCH & EDSAC | The National Museum of Computing
Using nuvistors in a subharmonic dividers chain rather than multivibrators could result in a (relatively) compact and reliable reference frequency source.
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA457231
http://www.dtic.mil/cgi-bin/GetTRDoc?AD=ADA457231
It's interesting that everyone is thinking in a digital mode. Like dividers, and numeral displays and such.
Why not analog? Use a volt or amp meter as your display. Just make it display a slow voltage ramp. You can do a combined minute and second hand, or split them into two meters. Easy.
Now all you need is an accurate ramp generator. An RC circuit, perhaps.
Why not analog? Use a volt or amp meter as your display. Just make it display a slow voltage ramp. You can do a combined minute and second hand, or split them into two meters. Easy.
Now all you need is an accurate ramp generator. An RC circuit, perhaps.
oh stop being silly... everyone knows if your going to build an aircraft carrier out of wood, it has to be teak... I mean c'mon.
It's interesting that everyone is thinking in a digital mode. Like dividers, and numeral displays and such.
Why not analog? Use a volt or amp meter as your display. Just make it display a slow voltage ramp. You can do a combined minute and second hand, or split them into two meters. Easy.
Now all you need is an accurate ramp generator. An RC circuit, perhaps.
holy carp... that's soo simple it might actually work.
Well I always wanted to build a clock like that, but never have. I had thought to make it transistor based, but why not tubes? You might be able to get away with a single dual element tube, or just 2 of the miniatures types. The RC time constant is going to be loooooooong, that's the major hurtle.
(totally off the subject) but... have you ever heard of the Little Bear B5 Headphone Amp?
Yeah... Just bought one off Amazon. It's a portable tube amp for headphones. I am currently burning mine in. Even before burn-in it sounds real nice.
I also got a pair of Philips SHL3300/28 headphones, and a shielded cable to go between my phone and the amp to keep them apart to reduce the microphonics that would result from being too close to the wifi and bluetooth signals. I gotta say, it's the best Easter present I have gotten in a looong time.
these are the most neutral headphones I have ever heard... absolutely flat. I had to actually increase the bass (for my own personal taste at 250,125, and 31)... I can't help it but I am a bit of a BASS head... but with a focus on BASS heavy classic rock (like Black Sabbath, etc). These headphones drop lower than a lead balloon... and with the addition of the tubes, it sounds fantastic.
Member
Joined 2009
Paid Member
My first 'real' pair of headphones (I'm not counting the Bose noise cancelling I bought for flying) were Grado's entry level SR60. The response was so ragged that I couldn't stop my ears bleeding so I sold them. Now, if these Philips are actually good then maybe I'd dip my toes back into the waters....
As for a tube based headphone amp - you really need very little power. I tried making a SS buffer for a tuner and the volume control hardly gets off the rest-stop!
As for a tube based headphone amp - you really need very little power. I tried making a SS buffer for a tuner and the volume control hardly gets off the rest-stop!
This is the best pic I can get as I don't have a screwdriver to open it.
An externally hosted image should be here but it was not working when we last tested it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.