Four Letter Word clock choreography

[Richard Scales]

I am working on a design for a FLW clock using B-7971 tubes with the original Cinch sockets. Having built a prototype I can see that these sockets are super kind to the tubes.

Given that a FLW clock displaying HHMM will have a very static display most of the time, I am wondering how often it might be suitable to display a WORD.

I already have something happening at 30 seconds past the minute, either the date, temperature, pressure, humidity or a scrolling message.

For the rest of the minute, and given that there are 1000's of FLW's - would once every 15 seconds be too little or too much?

Can anyone advise as to how existing FLW clocks handle this?

- Richard

[gregebert]

Shouldn't be a problem; these are sturdy tubes. I have an 8-tube clock and every 30 seconds it displays the day (eg, SATURDAY) and the date (eg JAN 30). No signs of segments weakening after several years. I have a PIR sensor so most of the time the display is running a walking segment pattern (1 segment of 1 tube). The ^ symbol is rarely used and shows no signs of poisoning. Tubes are direct-drive with no PWM dimming, and there are current limiters to keep the segment current and total anode current within datasheet limits.

[Richard Scales]

I too have a PIR connected so it is only ever on when needed, direct drive and PWM is available to dim if required.
At 30 seconds past it does the date/temp etc thing - but as it is only a four letter word clock, I was thinking of putting a word up every 15 seconds.
Is that too many, too few or about right ?

 - Richard

[gregebert]

I would pick an interval that is a prime number, so that the exact second where the FLW, date, etc comes up varies slightly. So instead of 15, use 17 or 19. Most of my clocks do their alternate thing at 15 and 45 seconds, and it's predictably annoying, and annoyingly predictable. I'm gonna go back and recode that.

If you have other clocks in the room, then 15 seconds is probably OK; but if it's the only clock I would stretch it out closer to 30.

[Richard Scales]

If i have other clocks in the room !!!!! - if anyone were to see my rooms - they would have me taken away!!!!

Seriously though - I love your prime number idea. I'll be giving that some thought - thank you.

- Richard

[David]

My original Ray Weisling FLW from early 2002 has been displaying words continuously from when I built it back then to now.

Ray may have run out of integrity, but he made a damn fine FLW algorithm. The slow 30ish second display of each word is soothing yet interesting. 

David Forbes, Tucson AZ

On Jan 30, 2025, at 9:40 PM, Richard Scales <ric...@scalesweb.co.uk> wrote:

I am working on a design for a FLW clock using B-7971 tubes with the original Cinch sockets. Having built a prototype I can see that these sockets are super kind to the tubes.

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/5e139118-cf01-4243-b8ab-39f05151d635n%40googlegroups.com.

[newxito]

In my clocks, I set every hour a random second for the alternate thing. I think this is an acceptable compromise.

[Richard Scales]

I like that too though, as this clock has no seconds to display - I want to make it slightly more dynamic - hence the desire to perhaps have 3 x random words + the date stuff during each minute.

I'll have a go at making it as random as possible.

- Richard

[gregebert]

There's no law against displaying minutes:seconds on a 4-digit clock, so that can help mix things up a bit.

BTW, I modified my LED clock (ewww....aren't LEDs blasphemy for nixie addicts ??) to display the date every 31 seconds (instead of every 30), so that lessens the monotony.