All transistor clock

Thread Starter

upand_at_them

Joined May 15, 2010
940
Saw this on the 'Tube. It's a teacher's project for his EE class in Thailand. I thought it was pretty cool. No IC's! I like how he did the mains counter/divider.


Video is 2 years old...Sorry if this has been posted/discussed before.
 
Somewhere I saw a clock that didn't use transistors, either. There's a trick where you make a ring counter using nothing but neon lamps and resistors and someone built a clock this way.

Back in the day all kinds of counter-based instruments used this technique. There was a special tube that had this all in one envelope called a Dekatron. Someone even made a computer called the W.I.T.C.H. out of Dekatrons.
 

dcbingaman

Joined Jun 30, 2021
1,065
You could make a no transistor clock with nothing more than a battery and a DC motor and an old mechanical clock like a Grandfather clock: Remove the 'weights' for a standard Grandfather clock and connect the drum that normally gets the torque from the weights with the motor. The motor simply needs to provide a little torque to keep the pendulum moving, it would not be moving at any significant speed but simply applying the required torque.
 

dcbingaman

Joined Jun 30, 2021
1,065
Saw this on the 'Tube. It's a teacher's project for his EE class in Thailand. I thought it was pretty cool. No IC's! I like how he did the mains counter/divider.


Video is 2 years old...Sorry if this has been posted/discussed before.
You could also do an 'hour glass' clock. Using a solenoid to turn the hour glass 180 degrees each time the sand runs out. You could detect the sand running out using a simple light beam that is broken by the sand. Then use a set of mechanical relays as a binary counter, to count how many times the sand has flipped.
 

WBahn

Joined Mar 31, 2012
29,978
You could also do an 'hour glass' clock. Using a solenoid to turn the hour glass 180 degrees each time the sand runs out. You could detect the sand running out using a simple light beam that is broken by the sand. Then use a set of mechanical relays as a binary counter, to count how many times the sand has flipped.
My guess is that there would be enough uncertainty in when the sand resulted in the beam changing that the accuracy of the clock would be pretty poor.
 

dcbingaman

Joined Jun 30, 2021
1,065
My guess is that there would be enough uncertainty in when the sand resulted in the beam changing that the accuracy of the clock would be pretty poor.
That reminds me of something I always wanted to make but never got around to it. That is an 'electronic' sun dial clock. By having the 'clock' placed properly (Due South and level) have a sensor that detects the position and length of the shadow. Than send that data back to a computer in the house wirelessly and convert the data into the time of day and maybe even the time of year.
Another way might just have say a tube that can move on the azimuth and elevation with a light sensor in the bottom, the tube would track the sun by keeping the light centered at the bottom of the tube and reporting the azimuth and elevation information to a PC in the house for conversion into time of day and time of year.
 

MrSalts

Joined Apr 2, 2020
2,767
Another way might just have say a tube that can move on the azimuth and elevation with a light sensor in the bottom, the tube would track the sun by keeping the light centered at the bottom of the tube and reporting the azimuth and elevation information to a PC in the house for conversion into time of day and time of year.
And it could start your coffee every morning (and 30-minutes after every cloud passes the sun).
 

WBahn

Joined Mar 31, 2012
29,978
That reminds me of something I always wanted to make but never got around to it. That is an 'electronic' sun dial clock. By having the 'clock' placed properly (Due South and level) have a sensor that detects the position and length of the shadow. Than send that data back to a computer in the house wirelessly and convert the data into the time of day and maybe even the time of year.
Another way might just have say a tube that can move on the azimuth and elevation with a light sensor in the bottom, the tube would track the sun by keeping the light centered at the bottom of the tube and reporting the azimuth and elevation information to a PC in the house for conversion into time of day and time of year.
The movement of the tip of the shadow is actually pretty complex, so it is a lot harder than just taking a single reading and converting that to any kind of time.

But there are definitely some very interesting things you could do by collecting a year's worth of time-stamped data. One thing you might find of interest would be to plot the analemma for your location.

Using a single tube with a single light sensor in the bottom is problematic because the system has to constantly search out the max signal on the sensor and this is constantly changing. Even clouds passing overhead that have no human-perceptible effect on the light can drastically change the amount of light hitting the sensor. A number of years ago there was a full solar eclipse and where I was we reached 90% totality. But even standing outside and watching it happen using a number of devices, none of use could tell that the light we perceived had dimmed at all. We could tell a slight sense that the heat on our skin had dropped.

The normal way of doing this is to use two tubes on each axis that are set at slight angles or two sensors set on opposite sides of a divider and then keeping the signal on both sensors the same.

Today you don't need to get all fancy with physical movement, however. Just mount a cheap camera and have it take pictures at regular intervals and then write some software (there's probably stuff out there, too) to analyze the images to identify and track the shadow tip.
 
Last edited:

MrChips

Joined Oct 2, 2009
30,708
I started a 1000-year solar powered clock project but never completed it.
Initially I experimented with a slit over an LDR in an attempt to detect the midday sun. I am inclined to drop that idea and instead will attempt to detect sunrise and sunset.
 

Marley

Joined Apr 4, 2016
502
I started a 1000-year solar powered clock project but never completed it.
Initially I experimented with a slit over an LDR in an attempt to detect the midday sun. I am inclined to drop that idea and instead will attempt to detect sunrise and sunset.
Very hard to make any kind of machine that will last 1000 years. There are some mechanical clocks in churches that are many hundreds of years old and still working. Not without some maintenance, I'm sure!

What is the oldest electronic device still working? How would you build a very long-lasting electronic clock? It's an interesting design problem.
 

MrSalts

Joined Apr 2, 2020
2,767
What is the oldest electronic device still working?
Maybe not the oldest, but most impressive (to me) is the NASA Voyager space probes. Operating since 1977 in severe cold of deep space.

Also old (and very concerning), are the 1960's era x-Ray machines that I saw, still in use, when I was in Haiti in 2018.

Then, very cool, are some of the early guitar amps from the 1950s - some with complete overhauls and may not contain any original parts at this point other than the output transformers if anything.

also, some completely overhauled radios from the 1920s are still in operation - I have an old "Detroitola" from 1929. I am ready to decommission it because of lack of AM content and I don't want to rebuild it again.
 

dcbingaman

Joined Jun 30, 2021
1,065
The movement of the tip of the shadow is actually pretty complex, so it is a lot harder than just taking a single reading and converting that to any kind of time.

But there are definitely some very interesting things you could do by collecting a year's worth of time-stamped data. One thing you might find of interest would be to plot the analemma for your location.

Using a single tube with a single light sensor in the bottom is problematic because the system has to constantly search out the max signal on the sensor and this is constantly changing. Even clouds passing overhead that have no human-perceptible effect on the light can drastically change the amount of light hitting the sensor. A number of years ago there was a full solar eclipse and where I was we reached 90% totality. But even standing outside and watching it happen using a number of devices, none of use could tell that the light we perceived had dimmed at all. We could tell a slight sense that the heat on our skin had dropped.

The normal way of doing this is to use two tubes on each axis that are set at slight angles or two sensors set on opposite sides of a divider and then keeping the signal on both sensors the same.

Today you don't need to get all fancy with physical movement, however. Just mount a cheap camera and have it take pictures at regular intervals and then write some software (there's probably stuff out there, too) to analyze the images to identify and track the shadow tip.
I have not had time to read this article, (skimmed it though). It appears there is a lot to know about an 'analemma'. I did not even know what the term meant until you mentioned it:

https://en.wikipedia.org/wiki/Analemma
 

k1ng 1337

Joined Sep 11, 2020
940
The movement of the tip of the shadow is actually pretty complex, so it is a lot harder than just taking a single reading and converting that to any kind of time.

But there are definitely some very interesting things you could do by collecting a year's worth of time-stamped data. One thing you might find of interest would be to plot the analemma for your location.

Using a single tube with a single light sensor in the bottom is problematic because the system has to constantly search out the max signal on the sensor and this is constantly changing. Even clouds passing overhead that have no human-perceptible effect on the light can drastically change the amount of light hitting the sensor. A number of years ago there was a full solar eclipse and where I was we reached 90% totality. But even standing outside and watching it happen using a number of devices, none of use could tell that the light we perceived had dimmed at all. We could tell a slight sense that the heat on our skin had dropped.

The normal way of doing this is to use two tubes on each axis that are set at slight angles or two sensors set on opposite sides of a divider and then keeping the signal on both sensors the same.

Today you don't need to get all fancy with physical movement, however. Just mount a cheap camera and have it take pictures at regular intervals and then write some software (there's probably stuff out there, too) to analyze the images to identify and track the shadow tip.
A camera could take a photo of the crosshair shadow produced by this sundial.

Stainless_steel_bifilar_sundial_(dial) (1).jpg
 

MrSalts

Joined Apr 2, 2020
2,767
But what happens on a cloudy day?
You could look at you mobile phone, or the TV or your laptop or estimate, or phone a friend, or, think back about how you got the time when your VCR was flashing 12:00, or... - anyone else have some ideas?
Edit: why are you only concerned with cloudy days? Should we assume you already had a plan for overnight that you were not concerned with those hours? ;)
 

WBahn

Joined Mar 31, 2012
29,978
But what happens on a cloudy day?
The same thing that happens when your watch or phone battery battery dies or your house loses power -- you get your time from some other source.

I don't see anywhere that anyone is suggesting that the clocks they are talking about would be the only -- or even primary -- means of telling time.

What is your 1000-year clock going to do on days that it is overcast at sunrise or sunset?
 

MrChips

Joined Oct 2, 2009
30,708
My 1000-year solar powered clock uses a 32768Hz quartz for time keeping.
It can keep time to better than a few minutes per year. It uses solar time to correct for drifting and does the correction slowly over months. It does not require daily correction. It does not depend on direct sunlight for correction. Daybreak and nightfall is detected and averaged over days. It does not matter if there is heavy cloud cover.
 

MrSalts

Joined Apr 2, 2020
2,767
My 1000-year solar powered clock uses a 32768Hz quartz for time keeping.
It can keep time to better than a few minutes per year. It uses solar time to correct for drifting and does the correction slowly over months. It does not require daily correction. It does not depend on direct sunlight for correction. Daybreak and nightfall is detected and averaged over days. It does not matter if there is heavy cloud cover.
Cool, a 1000-year solar powered, all transistor clock with no ICs. Wow.
 
Top