Re: [neonixie-l] Digest for neonixie-l@googlegroups.com - 4 updates in 3 topics

[Mike Mitchell]

The tubeclock database (http://www.tubeclockdb.com/) is a good place to start.

Akafugu has an end-view clock with colons:  http://www.akafugu.jp/posts/products/nixie/

On Tue, Jun 24, 2014 at 3:43 AM, <neoni...@googlegroups.com> wrote:

  Today's topic summary

Group: http://groups.google.com/group/neonixie-l/topics

• Saga of the Sperry part 2 [2 Updates]
• Digest for neoni...@googlegroups.com - 1 update in 1 topic [1 Update]
• Recommendations For A Socket Nixie Clock Kit [1 Update]

 Saga of the Sperry part 2

Morris Odell <vilg...@bigpond.net.au> Jun 23 04:14AM -0700  

Hi all,
 
Part 1 described the innards of the Sperry radar indicator that I wanted to
use for a clock and it's difficult internal structure. Now I'll get started
on the design.
 
The first step was to work out what frequencies and voltages I would need.
Experience has shown that CRT clocks need to run at a scan frequency that
is related to the mains frequency in order to prevent stray electric and
magnetic fields in the environment from wobbling the display. This was not
likely to be a big issue here as the neck of the CRT was shielded and by
the time the beam got out of the shield it would be accelerating fast
enough to not be seriously affected. Nevertheless I decided on 900 Hz as it
is a multiple of both 60 and 50 and was in the design range of the power
transformers. Looking at the ratings of those transformers I worked out
that I needed about 50 watts, possibly less if I could derive the +/- 12
volt and the shift coil supplies from the mains. I also thought it would be
pretty nifty if I could drive the scan coils from a rotary resolver driven
by a stepper which in turn would be controlled by an AVR and GPS receiver.
Cool huh?
 
The only snag was that I didn't have, and would almost certainly never
have, the proper resolver! I did have a couple of surplus 50 and 400 Hz
synchros though, so why not try?
 
The 100 volt 900 Hz power supply had to be a sine wave as that was
expected by the area balance circuit. This meant that the power supply
would be a linear amplifier driven by a sine wave source. Inefficient I
know, but making a switch mode sine wave inverter was just too hard given
the difficulty in finding suitable cores to work at 900 Hz. I designed up a
900 Hz exciter using a 4046 PLL phase locked to the mains followed by an
active low pass filter to knock off enough harmonics to make it look like a
sine wave. I included a voltage controlled attenuator to allow feedback
control of the output and built it up on a little PCB - it worked very well
and supplied several other outputs that I thought I would need.
 
For the power amplifier I chose a LM3886 which was cheap and capable of
generating lots more power than I needed providing it was heat sunk
properly and supplied with enough volts. It would need a transformer to
shift the output voltage to 100 and I raided both my and several friends
junk boxes to find what I though might work. I had a suitable 80 VA mains
power transformer already - it would develop +/- 25 volts and had taps to
also get +/- 17 volts for a pair of 12 volt 3 terminal regulators, and 6
volts for the CRT heater thus allowing me to remove one of the 900 Hz
transformers. The junk box yielded a suitable very large heatsink so away I
went.
 
I got to work removing all the unnecessary stuff from the indicator,
Dismantling is definitely a lot of fun :-) Out went the heater
transformer, IF strip, the power relays, fuses and the multi-pin
connectors. All the associated wiring came out of the looms too. A few
charred power resistors and all the electrolytics were replaced as were
some messy replacement rectifiers that sat on blackened areas on the rear
panel(!). The variable video time constant parts and switching went, as did
the heading indicator and signal strength meter. After doing that there was
enough room to mount the power transformer and the big heatsink for the LM
3886. I connected it all up, attached my 400 Hz synchro and reached for the
power switch - a scary proposition for something that hadn't seen moving
electrons for along time.
 
Well, the good news was that the CRT lit up OK (whew!) and the waveform
generators all worked but that's where it ended. It was pretty obvious that
my little synchros would never develop enough secondary current to
adequately deflect the beam. In addition the LM3886 was getting very hot
and working far too hard to force enough power through my junk box
transformers into the unit, even with a power factor correction capacitor -
now that's a phrase I never thought I'd ever use again! On top of all that,
the 900 Hz transformers howled like banshees. It was obvious a radical
rethink was going to be needed!
 
That's where I'll leave it for now, so stay tuned for the next exciting
installment folks!
 
Morris


 

yen...@internode.on.net Jun 24 04:04AM +0930  

Hmmm, love the story so far. Sitting on the edge my chair re the
900Hz. I am trying to get back to a project that needs approx.
200W power at 1600 - 1800 Hz [A Rebecca IV]. I have made a couple of
tiny 400Hz supplies for selsyns and looked at the variable frequency
supplies often on ebay [for spindle motors]. One cheap supply spec
said 0 - 2500Hz. Turns out that is period not frequency :-(    It
is 0 - 400Hz.
 
I have enough 'junque' to try to make a  FET 'vibrator' supply or a
high power audio amp. Have even bought a couple of transformer design
books.
 
Looking forward to the next instalment.....
 
John K
 
----- Original Message -----
From: neoni...@googlegroups.com
To:
Cc:
Sent:Mon, 23 Jun 2014 04:14:21 -0700 (PDT)
Subject:[neonixie-l] Saga of the Sperry part 2
 
Hi all,
 
Part 1 described the innards of the Sperry radar indicator that I
wanted to use for a clock and it's difficult internal structure. Now
I'll get started on the design.
 
The first step was to work out what frequencies and voltages I would
need. Experience has shown that CRT clocks need to run at a scan
frequency that is related to the mains frequency in order to prevent
stray electric and magnetic fields in the environment from wobbling
the display. This was not likely to be a big issue here as the neck of
the CRT was shielded and by the time the beam got out of the shield it
would be accelerating fast enough to not be seriously affected.
Nevertheless I decided on 900 Hz as it is a multiple of both 60 and 50
and was in the design range of the power transformers. Looking at the
ratings of those transformers I worked out that I needed about 50
watts, possibly less if I could derive the +/- 12 volt and the shift
coil supplies from the mains. I also thought it would be pretty nifty
if I could drive the scan coils from a rotary resolver driven by a
stepper which in turn would be controlled by an AVR and GPS receiver.
Cool huh? 
 
The only snag was that I didn't have, and would almost certainly
never have, the proper resolver! I did have a couple of surplus 50 and
400 Hz synchros though, so why not try?
 
The 100 volt 900 Hz  power supply had to be a sine wave as that was
expected by the area balance circuit. This meant that the power supply
would be a linear amplifier driven by a sine wave source. Inefficient
I know, but making a switch mode sine wave inverter was just too hard
given the difficulty in finding suitable cores to work at 900 Hz. I
designed up a 900 Hz exciter using a 4046 PLL phase locked to the
mains followed by an active low pass filter to knock off enough
harmonics to make it look like a sine wave. I included a voltage
controlled attenuator to allow feedback control of the output and
built it up on a little PCB - it worked very well and supplied several
other outputs that I thought I would need.
 
For the power amplifier I chose a LM3886 which was cheap and capable
of generating lots more power than I needed providing it was heat sunk
properly and supplied with enough volts. It would need a transformer
to shift the output voltage to 100 and I raided both my and several
friends junk boxes to find what I though might work. I had a suitable
80 VA mains power transformer already - it would develop +/- 25 volts
and had taps to also get +/- 17 volts for a pair of 12 volt 3 terminal
regulators, and 6 volts for the CRT heater thus allowing me to remove
one of the 900 Hz transformers. The junk box yielded a suitable very
large heatsink so away I went. 
 
I got to work removing all the unnecessary stuff from the indicator,
Dismantling is definitely a lot of fun :-) Out went the heater
transformer,  IF strip, the power relays, fuses and the multi-pin
connectors. All the associated wiring came out of the looms too. A few
charred power resistors and all the electrolytics were replaced as
were some messy replacement rectifiers that sat on blackened areas on
the rear panel(!). The variable video time constant parts and
switching went, as did the heading indicator and signal strength
meter. After doing that there was enough room to mount the power
transformer and the big heatsink for the LM 3886. I connected it all
up, attached my 400 Hz synchro and reached for the power switch - a
scary proposition for something that hadn't seen moving electrons for
along time.
 
Well, the good news was that the CRT lit up OK (whew!) and the
waveform generators all worked but that's where it ended. It was
pretty obvious that my little synchros would never develop enough
secondary current to adequately deflect the beam. In addition the
LM3886 was getting very hot and working far too hard to force enough
power through my junk box transformers into the unit, even with a
power factor correction capacitor - now that's a phrase I never
thought I'd ever use again! On top of all that, the 900 Hz
transformers howled like banshees. It was obvious a radical rethink
was going to be needed!
 
That's where I'll leave it for now, so stay tuned for the next
exciting installment folks!
 
Morris
 
--
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 [1].
To post to this group, send email to neoni...@googlegroups.com [2].
To view this discussion on the web, visit
https://groups.google.com/d/msgid/neonixie-l/c5f1b0c1-1242-49de-8153-f7908a986a8b%40googlegroups.com
[3].
For more options, visit https://groups.google.com/d/optout [4].
 
 
Links:
------
[1] mailto:neonixie-l+...@googlegroups.com
[2] mailto:neoni...@googlegroups.com
[3]
https://groups.google.com/d/msgid/neonixie-l/c5f1b0c1-1242-49de-8153-f7908a986a8b%40googlegroups.com?utm_medium=email&utm_source=footer
[4] https://groups.google.com/d/optout


 

 Digest for neoni...@googlegroups.com - 1 update in 1 topic

"Dave B." <bla...@gmail.com> Jun 23 10:21AM -0700  

I'm not sure what you mean by "top tubes" unless you mean the IN-17 variety that have wires instead of pins.
 
I built the "Lena" a couple months ago and really like it. http://nixiekits.eu/
 
The kit is very high quality and well thought-out, and the PIC programming is quite clever in terms of the various effects that can be configured. I must say, though, that the fact that the IN-12 tubes use an upside-down "2" filament for the "5" drives me a bit crazy, so I try to never look at my clock when there's a "5" in the hours or minutes ;^) . Also, Lena uses IN-17s for the seconds digits. You would probably want the Lara kit, which has the same mainboard but a different daughtercard for the six IN-2s.
 
Of course all of those tubes have "on the top" digits; if that's what you mean then of course none of those will do. In that case, check out "Sven reloaded" which uses side-view nixies.
 
BTW all of those kits have some SMT components, which are pre-soldered. But the Lena was still a fairly challenging kit to build, in a good way. You HAVE to follow the instructions closely or you will goof something up, I'm sure if it.
 
cheers,
 
Dave B.
 
On 6/23/2014 12:55 AM, neoni...@googlegroups.com wrote:
 
Today's topic summary
 
Group: http://groups.google.com/group/neonixie-l/topics
 
Recommendations For A Socket Nixie Clock Kit [1 Update]
 
Recommendations For A Socket Nixie Clock Kit
 
Dman777 <darin....@gmail.com> Jun 22 02:37PM -0700
 
Hi,
 
I am looking for recommendations for a Nixie clock kit that uses 6 *socket*
 
tubes(please note this is *not top tubes*).


 

 Recommendations For A Socket Nixie Clock Kit

Paul Parry <xd...@tiscali.co.uk> Jun 23 01:26AM -0700  

On Sunday, 22 June 2014 22:37:04 UTC+1, Dman777 wrote:
 
> http://www.pvelectronics.co.uk/index.php?main_page=product_info&cPath=1&products_id=122
> This uses IN-4 Tubes. I like this clock...but I am not a fan of the case
> and I would to have column lights
 
I do a different case for this kit:
http://www.bad-dog-designs.co.uk/images/Contemp/Desk_3.jpg

Although get the Halo kit as you can use a wide variey of end view tubes, I
personally use GN-4's as they are excellent tubes.

Cheers,
Paul