INS-1 flickering

420 views
Skip to first unread message

Larry

unread,
Feb 11, 2011, 2:39:25 PM2/11/11
to neonixie-l
I'm in the final stages of putting together a clock using INS-1 tubes
for the colon dots. The dots alternate lighting every second. I'm
using an ArduinoNIX as my power supply, cathode controller, and anode
switches. Each of the INS-1 tubes uses an 180k ohm anode resistor. My
problem is that one of the INS-1 tube works fine while the other
starts it's glow at the top of the tube and then it moves to the base
of the tube. Is this common for INS-1 tube or have I just got a bad
one?

threeneurons

unread,
Feb 11, 2011, 3:53:15 PM2/11/11
to neonixie-l
> On Feb 11, 11:39 am, Larry <unmitigated_f...@earthlink.net> wrote:
> I'm in the final stages of putting together a clock using
> INS-1 tubes for the colon dots. ... Each of the INS-1 tubes
> uses an 180k ohm anode resistor. My problem is that one of the
> INS-1 tube works fine while the other starts it's glow at the
> top of the tube and then it moves to the base of the tube. Is
> this common for INS-1 tube or have I just got a bad one?

The 'flickering' INS-1 has been reported in the past. I don't recall
what the general conclusion was about this problem. It could be
several things:

1) Inadequate current. The datasheet specifies 0.5mA. That means
around 90V should be dropped across that 180K resistor. If its
significantly less (0.2mA or less) then that can be a problem

2) Polarity. The INS-1 has two metal structures inside the glass. One
is the anode, and one is the cathode. I believe the lower of the two
structures is the anode. Its shaped like a cup, emptying downward.
This structure is connected to the lead which has a bump (dot) on the
glass envelope. The cathode structure is above it, just under the
glass lens. Its shaped like a hollow cylinder. The cathode is usually
larger, since that's the electrode that gets eaten away over time. The
current should be high enough to cover this structure.

I've run some long term tests (> one month) on a few of my own INS-1s,
and none seem to flicker. Others, though, have reported the
flickering. It usually occurs after running for a few days.

Finally, I don't read Russian, so I don't know if that 0.5mA is
typical or maximum operating current.

Larry

unread,
Feb 12, 2011, 10:30:45 AM2/12/11
to neonixie-l
I looked closely at my assembly and determined that the two tubes are
wired in reverse polarity. This adds weight to your theory about
polarity being the cause of the flicker. I'll build up a couple sets,
one each with both tubes one polarity or the other and see if on pair
flickers and the other doesn't.

Larry

unread,
Feb 12, 2011, 4:03:40 PM2/12/11
to neonixie-l
Changing the polarity does the trick.

threeneurons

unread,
Feb 12, 2011, 8:20:34 PM2/12/11
to neonixie-l
On Feb 12, 1:03 pm, Larry <unmitigated_f...@earthlink.net> wrote:
> Changing the polarity does the trick.

Glad I could be of help. Mind you, all of my info about the INS-1 is
mostly empirical, meaning experimental. That means 'hit-n-miss'. Think
monkeys in a cage flinging they're poo. Oh ! I might be mistaking that
with politics, sorry.

Gene Segal

unread,
Feb 12, 2011, 8:27:01 PM2/12/11
to neoni...@googlegroups.com
I have observed the following behavior of ИНС-1:

With proper polarity connection, and rated current/voltage applied per spec sheet, SOME of the lamps exhibit unstable behavior, where they start pulsating sporadically. Others do not show this behavior. I would not call it a flicker (if flicker is defined as sporadic on/off condition), but a periodic pulsation, a kind of internal thermodynamic feedback loop, once set into pulsate mode, will tend to stay in that mode. In that sense, it is stable, but in the sense that it happens sporadically (probably a function of ambient temperature or light), it is unstable.

Gene

Larry

unread,
Feb 12, 2011, 9:29:54 PM2/12/11
to neonixie-l
Thanks to everyone's help I now have a Nixie clock that has full
Westminster chimes. It's driven by an Arduino board and uses an
Adafruit Waveshield for sound and an ArduinoNIX shield for Nixie
control, along with a 74HC595 chip to add a few more output pins. I'm
using a DS1307 RTC to keep it in time. The clock will self adjust for
DST for the next 10 years. By that time it's going to need a new
battery anyway. The clock is mounted in an old Seth Thomas case and
with the wave files playing has a wonderful resonance.

Jon

unread,
Feb 13, 2011, 4:14:37 AM2/13/11
to neonixie-l


On Feb 12, 3:30 pm, Larry <unmitigated_f...@earthlink.net> wrote:
> I looked closely at my assembly and determined that the two tubes are
> wired in reverse polarity.  This adds weight to your theory about
> polarity being the cause of the flicker.

I played quite a bit with IN-3 a few years ago and observed a similar
effect. They were much better when you wired them with the correct
polarity.

Jon.

threeneurons

unread,
Feb 13, 2011, 5:13:52 PM2/13/11
to neonixie-l
> On Feb 12, 6:29 pm, Larry <unmitigated_f...@earthlink.net> wrote:
> Thanks to everyone's help I now have a Nixie clock that has full
> Westminster chimes.  It's driven by an Arduino board and uses an
> Adafruit Waveshield for sound and an ArduinoNIX shield for Nixie
> control, along with a 74HC595 chip to add a few more output pins.  I'm
> using a DS1307 RTC to keep it in time.  

http://www.ladyada.net/make/waveshield/

Well that's a big plus for Arduino ! Having a bunch of 3rd party
attachements. I'm always a big fan of sound. No, I don't have 'golden
ears' requiring the best sound system. They can be considered 'tin
ears', if anything. When I use to go to the CES Las Vegas show, and
visit the high end speaker exhibits, I'd admire the fine wood working
and ignore the sound. Back, to nixies, I know, I've been one of the
few guys who likes a real sounding chime. There are other ways to do
this, and I've used plenty of them, but this is a big plus for a (I
suspect) large subset of this group that uses the Arduino platform.

BTW: bought a low end Arduino, a month ago, to play with. Actually,
for my more software oriented friends to play with. Its cheap, and you
don't need to be a solder monkey. I'm clearly a solder monkey.

coggs

unread,
Feb 13, 2011, 5:30:07 PM2/13/11
to neonixie-l
@Larry: Sounds like a great clock. Hope you find the time to post
pictures.
..c

TheJBW

unread,
Mar 28, 2016, 4:14:22 AM3/28/16
to neonixie-l, wave...@earthlink.net
This is an ancient topic, I know, but I wanted to chime in. Back in late December, I built a clock which uses an INS-1 as an AM/PM indicator (so on 12hr / day). With the correct polarity, the lamp was 100% fine for the first two weeks or so, and then began to exhibit the flickering phenomenon described here. It was a mild annoyance, but I didn't have time to address it, as it was soldered in place and enclosed behind acrylic. Having forgotten about it entirely for a few months, I was looking at it last week, and noticed that it doesn't flicker anymore. I've been keeping an eye on it since, and have seen no hint whatsoever of flicker. So it seems, at least for a tiny sample size, that this self oscillation / flicker may simply go away after a 'burn-in' period.

-Bill

Gene Mark Segal

unread,
Mar 28, 2016, 10:37:11 AM3/28/16
to neoni...@googlegroups.com
Bill, excellent observation! I almost didn't recognize my own post from 2011, thanks for digging that up!))

I'm curious if that INS-1 you reported about, which went stable, will stay stable long-term. Please do report. 

Best regards, Gene 
--
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 email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/1083afa8-af9b-4a44-a4b2-22080f56b52e%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

TheJBW

unread,
Mar 29, 2016, 12:20:47 AM3/29/16
to neonixie-l
Will do! I've set myself a reminder for a few months from now, so I'll update either way.

TheJBW

unread,
Oct 3, 2016, 9:01:15 PM10/3/16
to neonixie-l
Gene (and anyone else tracking this),
It's been another six months and the INS-1 in question is still soldiering on flicker-free and fully bright. If it ever goes bad again, I'll make a point of picking this thread back up.

-Bill

Gene Mark Segal

unread,
Oct 3, 2016, 9:10:01 PM10/3/16
to neoni...@googlegroups.com
Bill, thanks for reporting! That's very useful information. Looks like it went stable ultimately. It's strange that it was stable for the first two weeks; if there was a "burn in" I would think the lamp would act up right away. Maybe voltage fluctuated in the device?

Gene

TheJBW

unread,
Oct 4, 2016, 2:15:44 PM10/4/16
to neonixie-l
Gene,
It is possible, although I somewhat doubt it. The supply source is a 12V cheapo amazon 2A unit driving a Tayloredge 1364 module on a Tayloredge 1355 backplane (along with 6 IN12As).

The 12V supply is driving several other modules (an IN13 module that I built -- which is powered by an independent 1364 taking 12V input), a raspberry pi, an amp, and some low current incandescent indicators. While these obviously will load the power supply randomly over time, peak draw is on the order of 750mA and all of this hardware was installed weeks before the flickering began.

In any case, definitely strange and maybe a unique case, but perhaps worth using a test rig on any INS-1s that are unstable but might be saved.

-Bill

RobG

unread,
Oct 23, 2016, 10:56:51 AM10/23/16
to neonixie-l
One of the most requested features on my clock was the separator, so I have decided to add two INS-1.
Once I got my tubes, I tested large batch of them and discovered that some of the tubes start to flicker after about 15 minutes.
My first thought was that those tubes are going bad, after all, they are old and the quality is... well, just look at the lens with magnifying glass.
Since this "effect" is similar to flicker flame, where low current makes neon unstable, I figured I have to increase the current.
My default was 0.6mA (220K @ 180V,) so I bumped it to 0.75mA (180K.) That did not help, it made it little worst I think, so I figured that's not it.
I tried to go the other way and decreased the current to 0.3mA (470K) and then to 0.15mA (1M.) That helped, flicker became slower, which might be unnoticeable when tube is used in a separator. Another advantage was that the glow was not as strong as @0.6mA and now it matched glow of other Nixies (digits.)

Just for fun, I reversed polarity. Flicker was gone and the glow was brighter. However, I will not suggest using them in reverse.

Then I decided to do an experiment. I connected flickering tube in reverse without current limiting resistor (however, my power supply limits current to 15mA.) The tube started flashing white, blue, red for about 2-3 seconds and then turned bright orange (normal operation but very bright.) I kept it on for another 2 -3 seconds and turned it off. Now, when connected the right way, tube no longer flickers. I did that with several tubes and they all became flicker free (kept them on for couple of hours.) Only one tube started flickering again, but the flicker goes away after several seconds (flicker starts slow, then faster, and then stops.) I will keep them on for few days and report back.

Gene Segal

unread,
Oct 23, 2016, 6:02:56 PM10/23/16
to neoni...@googlegroups.com
Very, very informative find and excellent write-up.  Thanks!!!!

October 23, 2016 at 7:56 AM
--
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 email to neoni...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
October 4, 2016 at 11:15 AM
Gene,
It is possible, although I somewhat doubt it. The supply source is a 12V cheapo amazon 2A unit driving a Tayloredge 1364 module on a Tayloredge 1355 backplane (along with 6 IN12As).

The 12V supply is driving several other modules (an IN13 module that I built -- which is powered by an independent 1364 taking 12V input), a raspberry pi, an amp, and some low current incandescent indicators. While these obviously will load the power supply randomly over time, peak draw is on the order of 750mA and all of this hardware was installed weeks before the flickering began.

In any case, definitely strange and maybe a unique case, but perhaps worth using a test rig on any INS-1s that are unstable but might be saved.

-Bill

On Monday, October 3, 2016 at 6:10:01 PM UTC-7, Gene Segal wrote:
--
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 email to neoni...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
October 3, 2016 at 6:09 PM
Bill, thanks for reporting! That's very useful information. Looks like it went stable ultimately. It's strange that it was stable for the first two weeks; if there was a "burn in" I would think the lamp would act up right away. Maybe voltage fluctuated in the device?

Gene

On Oct 3, 2016, at 6:01 PM, TheJBW <the...@gmail.com> wrote:

--
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 email to neoni...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.
October 3, 2016 at 6:01 PM
Gene (and anyone else tracking this),
It's been another six months and the INS-1 in question is still soldiering on flicker-free and fully bright. If it ever goes bad again, I'll make a point of picking this thread back up.

-Bill

On Monday, March 28, 2016 at 7:37:11 AM UTC-7, Gene Segal wrote:
--
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 email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/e756320c-3f0e-471d-89db-93676527b2a8%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
March 28, 2016 at 7:37 AM
Bill, excellent observation! I almost didn't recognize my own post from 2011, thanks for digging that up!))

I'm curious if that INS-1 you reported about, which went stable, will stay stable long-term. Please do report. 

Best regards, Gene 

On Mar 28, 2016, at 1:14 AM, TheJBW <the...@gmail.com> wrote:

--
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 email to neoni...@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.

--
Sent from Postbox
Reply all
Reply to author
Forward
0 new messages