Relay-controlled Nixie Clock

318 views
Skip to first unread message

ahochan

unread,
Aug 7, 2013, 5:50:11 AM8/7/13
to neoni...@googlegroups.com
Hi,

I'm thinking of creating a Nixie clock controlled by mechanical 5V relays.

I won't do anything fancy, just static drive with the Nixie anode connected
to 180VDC through a resistor and one relay for each cathode. The relay coils
will be connected to uC pins via either individual transistors or perhaps
a darlington array. Like so:

  180V
   |
   R (anode resistor)
   |
 Nixie
| ... |
1 relay for each cathode (connected to GND)

Schematic here:

I'll probably use IN-4 tubes, but that may change.

I have two questions:

* Does my schematic look sane?

* Can anyone recommend a relay to use?
  I'm looking for a relay that can be switched from 5VDC. There are many
  available in small form factors, but most seem to be rated for 250VAC/30VDC
  or similar. (I guess a relay rated for a lower DC voltage might be ok too,
  since it will only pull 3-4mA)

I want a mechanical relay... since the main point of this exercise is to get
the clicking sound when the relays switch.

Thanks in advance.

Tidak Ada

unread,
Aug 7, 2013, 6:24:52 AM8/7/13
to neoni...@googlegroups.com
Did you consider the power drain of such a clock?
I was also looking to build a meter-clock with all flip-flops build with relays and also relay's for the D/A converters. the sum of the current was shocking
You could consider to use SMD-relays, but even in that case the power drain is considerable!
Even the neon and dekatron clocks by Sgitheach  [ http://www.sgitheach.org.uk/index.html ] are only used for demonstration because of their power consumption!
You could think about building only the minutes stage in relay logic or even the better only the first divider of that stage. that makes noise enough.
 
Anyhow an interesting idea!
 
eric

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of ahochan
Sent: woensdag 7 augustus 2013 11:50
To: neoni...@googlegroups.com
Subject: [neonixie-l] Relay-controlled Nixie Clock

--
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 post to this group, send an email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/32a67b85-9496-4f62-84dc-c76bfff0d83b%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Dekatron42

unread,
Aug 7, 2013, 11:42:27 AM8/7/13
to neoni...@googlegroups.com
Hi,
 
You can use bistable relays with one coil, they change state for every pulse, so to turn them on you pulse them once and to turn them off again you pulse them once again - there are also bistable relays with two coils where one coil is used to turn them on and the other to turn them off. This way you will not draw any power when the relays are not switching states.
 
You can also use uniselectors, relay coils that move a rotating arm across several contacts. These are harder to find but turn up now and again on eBay. There is a movie here: http://www.youtube.com/watch?v=zUB_CUab5gY , and another one here ( close to 40 seconds in the movie) : http://www.youtube.com/watch?v=YajwGdbXV1w
 
These types of clocks where built during the 60's and 70's and you can find circuit diagrams for them in old magazines. There were even clocks which where driven by motors which turned rotary switches, with a set of gears they made it possible for seconds, minutes and hours and even AM/PM indicators.
 
/Martin

Matthew Cameron

unread,
Aug 7, 2013, 2:44:29 PM8/7/13
to neoni...@googlegroups.com
If you are going for the click sound, also consider using a 10 position stepper switch. One stepper can control an entire digit and
could be used to advance the next stepper. These, and other like them were used in rotary phone systems. Some of those have a
release relay, so you can reset a digit to 0 or off from any state. Since these require more current, I would expect for them to be
louder than modern 5V relays, but they draw nothing when they are not moving. I would expect for the old steppers to sound quite
monotonous over time, but maybe less so if put in a tight wooden enclosure or something.

Mechanical Nixie tube Clock:
http://www.youtube.com/watch?v=YajwGdbXV1w
(no relation to me)

ahochan

unread,
Aug 7, 2013, 8:08:34 PM8/7/13
to neoni...@googlegroups.com, mc1...@matthewc.net
Thanks for the tips... bistable relays seem to be the way to go, at least for a prototype.

Stepper switches / uniselectors look fascinating, and I may go for that instead if I can source some.

greg...@hotmail.com

unread,
Aug 8, 2013, 11:41:46 AM8/8/13
to neoni...@googlegroups.com, mc1...@matthewc.net

A word of caution about 10-position stepper relays, etc: These devices are quite old, probably 40+ years, and the contacts will have oxidation and other contamination that makes them intermittent. Several years ago I threw out my stepper-relay clock because the contacts were too much hassle to keep cleaning, and a few of them were nearly worn-out. The units I had were not sealed; maybe there are  sealed devices in existence, so if you could find some, that might be a game-changer.

Incidentally, the stepper relay unit I had was a pre-nixie era digital voltmeter that used individual incandescent bulbs that side-illuminated plastic numerals.  I morphed it into an interesting clock, but it was too unreliable, bulky, and quite frankly, rather ugly. Resting In Pieces.......


Mark Moulding

unread,
Aug 8, 2013, 12:51:36 PM8/8/13
to neoni...@googlegroups.com


The clicky sound won't be very loud, but still audible if you use reed relays, and many of these have the added advantage of being low current enough to drive directly off the output pin of a microcontroller (25 mils, or even less).  It's a little difficult to find ones rated for 250 or 300 volts; most have contacts rated for 100 volts, but I suspect that they would still work fine in this application.

Here's one that only draws 10 mA, and has the suppression diode built in: http://www.goldmine-elec-products.com/prodinfo.asp?number=G18683    Since you'd only have one relay per digit energized at any time, the total current drain for the relays would be 60 mA, or 300 mW.  Too much for a wristwatch, but completely reasonable for any plugged in device.  I bet the nixie DC-DC converter will draw more.

If a bit of noise doesn't bother you, and you *do* have a beefy power supply, you could make one using the digit wheels from a pinball machine.  These are mounted on large rotary stepper switches, which draw *amps* while switching, but only for 50 mSec or so.  They're pretty loud, though, and if you use these, you might as well just leave the number wheel mounted and have a pinball machine clock - maybe with a nixie remote repeater display...
~~
Mark Moulding

Mats Engstrom

unread,
Aug 24, 2013, 6:10:49 AM8/24/13
to neoni...@googlegroups.com
Why is the power drain such a huge issue in the thread? The OP doesn't plan to do anything fancy like having a huge chain of relays to decode bcd, just simply driving them from a microcontroller straight up.

A six-digit nixie clock would only have 6 relays on at the same time, and a normal small 5 volt relay consumes about 150mW. That's just 0.9 W in total  - hardly anything to be concerned about especially since the nixie digits draw 400 mW each!


--
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 post to this group, send an email to neoni...@googlegroups.com.

Tidak Ada

unread,
Aug 24, 2013, 6:19:19 AM8/24/13
to neoni...@googlegroups.com
It's just the challenge to do everything with relay's, including the dividers. No MC nonsense !
People seems to have been forgotten that dividers and clocks where build out of flip-flops, discrete or with logic IC's...
 
eric


From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Mats Engstrom
Sent: zaterdag 24 augustus 2013 12:11
To: neoni...@googlegroups.com
Subject: Re: [neonixie-l] Re: Relay-controlled Nixie Clock

Mats Engstrom

unread,
Aug 24, 2013, 6:34:21 AM8/24/13
to neoni...@googlegroups.com
@eric: Sure, but that is not the issue here.  The OP clearly stated the desire to build it with a MC - and then only 6 relays need to be energized at a time. 


Dekatron42

unread,
Aug 24, 2013, 9:27:08 AM8/24/13
to neoni...@googlegroups.com
With uniselectors and bistable relays, none have to be energized except when switching! ;)
 
/Martin

Tidak Ada

unread,
Aug 24, 2013, 12:00:48 PM8/24/13
to neoni...@googlegroups.com
Unfortuniately as you already said before, uniselectors are difficult do get nowdays. In a lesser way that also applies to bistable realays.
 
An other way, surrely in case you don't mind your 'antique' clock is equiped with a quartz clockwork, you can build your nixie clock with a microprocessor and a single (bistable) relay for the tick-tack, but why souldn't you use a loudspeaker and a 1 sec pulse instead?
 
Ok, it is my mainspring to use as much as possible old technology in my development of a meter clock with 'steam time'  moving coil meters as the read out. And if you then plan to build 64 or 128 bit D/A converters with relay's for each meter, you can make yourself an image of the curent drain. It will be much more as some of the electron tube based clocks that are published on the web....
 
eric
 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Dekatron42
Sent: zaterdag 24 augustus 2013 15:27
Reply all
Reply to author
Forward
0 new messages