Talking of nixie tube clocks... (Digi-Time)

[TonyDuell]

... I have one that contains no ICs, no transistors, and no valves (other than the nixie tubes).

It's called a Digi-Time and is electromechanical. The attached photos show the unit in its case and the chassis taken from the top front. The silver metal wheel on the right is on the spindle of a synchronous motor rotating at 1rpm.

Every revolution it moves the 10-tooth star wheel to the left on one position. This is on the spindle of a 10-position rotary switch that operates the unit-minutes nixie tube.

The 6 tooth star wheel next to it operates the ten-minutes nixie tube. It is moved on one tooth by a pin on the unit-minutes wheel. This of course is a 6 position rotary switch

In turn that moves on the 12 tooth star wheel next to it which is on the spindle of a 12 position rotary switch. This operates both the 'hours' nixie tubes, I guess there are some diodes involved in decoding that.

The smoothing capacitor for the HT supply for the nixie tubes (half-wave rectified mains) needs replacing. I must get it working!

[Buzby123]

Was that a commercially available unit, or somebody's clever homebrew ?

[Ed_Dinning]

Looking at the gears it seems to have shades od the WW cardboard clock of about 40 years ago

Ed

[TonyDuell]

I think it was commerical, the nameplate on the front and the lacing-cord tied wiring under the chassis look too good for a homebrew. And it has a 'Disconnect mains before removing back' label on it which I would not expect on a homebrew. But probably a small company.

[thermionic]

That’s very interesting and ingenious!

How was time set?



SimonT.

[Radio Wrangler]

The GEC RC410 shortwave receiver was a very early frequency synthesised job. Its tuning knob operated a train of "Geneva Wheels" similar to this, but true geneva wheels can rotate in either direction, like a tape counter on a recorder. Each wheel had 10-position switches on it, programming the synthesiser. As you turned the tuning knob it clicked round and you felt a stiffer click as it incremented the next wheel. It got worse, at frequency transitions where several digits incremented at once you needed the handle on the tuning knob to force the change.

The MHz switch and wheel was brought out to another large knob so you didn't have to wind the whole way, and the 10kHz and 100kHz wheels came out to smaller knobs.

On top of all this, the switches had extra poles to operate the frequency display nixies.

So while this tuned the frequency synthesiser, there was the matter of offset-tuning the RF stages. This was handled by a preselector with relay-switched bands and a multi-ganged variable capacitor turned by a servo motor and gearbox. I think the law of the capacitor was arranged to make the preselector frequency reasonably predictable and it too was tuned from the switch bank of the geneva wheels.

An interesting receiver, but with huge SRBP printed circuit boards warping as you looked at them, it was a nightmare to keep one in functional condition. The drag of the switch trains made tuning too unpleasant for tuning around for short wave listening. It was only marginally better than those sets with thumbwheel switches.

David

[Peter88gate]

I built a nixie clock in 1965 based on the same mechanical principles as Tony's one. It was based on a Practical Electronics article during that year. I set my clock going in Jan 1966 and it has been operating continuously ever since, apart from a short interlude when I moved house in 1976!

See my previous post https://www.vintage-radio.net/forum/...7&postcount=20

Peter

[mark_in_manc]

What a nice thing. I'd like a nixie clock; I bought a few tubes but somehow I think I am scared of working with things with lots of legs (the control circuits, I am OK with the tubes themselves!) ; I am a massive neophyte with digital electronics. I do have a master clock in use which produces one-per-minute pulses though - this would be a neat way to do it!

[TonyDuell]

Originally Posted by thermionic

That’s very interesting and ingenious!

How was time set?



SimonT.

There's a slot in the back cover (with a proper plastic trim around it which is another reason I think it's commercial). You stick a finger in and twidde the wheels.

[wireless_john]

How interesting! Tempting me back to the shed except that I've too many other things on the go! The March 1965 issue of Practical Electronics has been added to my list now.

https://worldradiohistory.com/UK/Pra...cs-1965-03.pdf

[TonyDuell]

That's very interesting. There's a tagboard of resistors at the back of the chassis of the Digi-Time and the values are much the same as those in the PE article (lots of 47k, the odd 180k).

It's not the same circuit, there's an electrolytic capacitor can under the chassis in the Digi-time. But it looks rather similar.

[The Philpott]

Tony's clock is a great concept, there's something about electro-mechanical that appeals to me much more than pure electronics. If based on the same principles but prettier inside, a steampunk version with a see-through acrylic case would be expensive to make but might sell well (and would incidentally be a rare iteration of steampunk that i would not sneer at.)

Dave

[G6Tanuki]

I remember visiting STC's labs in Harlow some time in the very-early-80s [when TXE4A exchanges were still a thing] and they had a Nixie-clock whose counting circuitry was built entirely out of old Uniselectors.

Click.. Click.. Click... At certain times it would do a 'ratchet round back to zero' for one or more of the digits which produced much rattling.

[hamid_1]

There are lots of homebrew nixie clocks on Mike's Electric Stuff website, including a couple using mechanical uniselectors (presumably from a Strowger telephone exchange) and an electronic nixie clock using neon ring counters - no IC's.

https://electricstuff.co.uk/nixiegallery.html

Back in 2005 that site inspired me to build my own nixie clock from a kit. It's still going strong, over 150,000 hours later! The nixie tubes haven't burned out yet.

There was another electromechanical digital clock called Lumitime produced commercially by Tamura of Japan. Not acually a nixie clock, it used a neon-lit 7 segment display but no electronics. Gear wheels turned switches on or off which made the display segments light up in the correct sequence to form numbers.

Here's some videos of the Lumitime :

https://www.youtube.com/watch?v=SVG9veCKogU

https://vimeo.com/411154906

and a thread from another forum with some pictures of the mechanism:

https://www.eevblog.com/forum/repair...ock-mechanism/

[Radio_Dave]

I cobbled together a nixie clock using uniselectors. The noise was horrendous!
https://www.youtube.com/watch?v=zUB_CUab5gY

David

[Trevor]

I built a Nixie tube clock Used the mains as the clock then 4017b decade counters with MPSA42 as the tube drivers it ran for a no of years even had a alarm function Eventaly retired due to tired Tubes

[Dekatron]

I restored a TIMEON 2000 Nixie clock that had a similar motor drive, I was lucky to come across a magazine that described it and I published details here: https://groups.google.com/g/neonixie...m/Nb9pY-j3Fq8J

I also have a TIMEON prototype that uses the very rare ZM1250 dot matrix Nixie tubes, you can see the ZM1251 here: https://groups.google.com/g/neonixie...m/sA-SibKDo58J and in this thread if you scroll down a bit: https://www.circuitsonline.net/forum/view/96134/27 . This clock is non working with one broken ZM1250 and one that has leaked so it shines more blueish/purple than orange. There are several hundred silicon diodes and almost as many hand soldered wires which makes up a very complex matrix which transform the switch connections into separate digits on each ZM1250 and it is almost impossible to repair without having to de-solder a lot as there are a few circuit boards stacked tightly with components soldered from one board to the other.

I do have a few spares but it is almost like soldering surface mount components to solder these ZM1250's.