IN-18 Typical Failure Mechanism

[Jeff Walton]

I recently had a IN-18 failure in a clock after about 2.5 year operation.  Over a (4) day period, the digits in one of the tubes stopped working in PWM dimmed mode.  At full brightness, the digits in the affected tube slowly lost percentages of digit coverage to where only portions of each digit would illuminate.  At the end of the observed period, the digits in the failing tube went completely dark and even higher voltage on the bench would not illuminate any of the cathodes.  The tube date code was 8505.

I am guessing that absent of any other signs of damage, a seal around one of the pins has failed and allowed the tube to leak.  It was surprising how quickly it failed and none of the other tubes in the same clock have exhibited any signs of trouble and look as good as the first day of operation.  Has anyone experienced similar failures with the IN-18?  Is this a typical failure mechanism?  Have there been any particular date codes that others have found to be problematic?

The tube has been since been replaced and the clock is happy again but curiosity leads me to ask about this experience to the rest of the members of the group.  FYI - these tubes have not been rotated since the clock was initially put into operation (no handling).  There was never any evidence of cathode poisoning during the operation of the clock.

[Dan Hollis]

I had same failure, but probably due to my rotating tubes. I plan to make
an adapter plug to relieve the strain on the pins when I rotate the tubes.

-Dan

> --
> 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/4f4d6cf9-6fba-49a6-951a-c83c96f728b1%40googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>

[gregebert]

Do the pins of your failed IN-18 have surface corrosion ? I've noticed all but one of my IN-18's have surface corrosion on all pins, and I've often wondered if the corrosion would continue at the glass interface and cause microscopic cracking up the pin until the seal breaks.

[Jeff Walton]

I would consider the pins to be “tarnished” as opposed to corroded.  That may just be semantics.  I have seen tube pins that were considerably worse looking but I included a photo of the specific tube pins which failed.  The tube has been dissected..

 

 

Jeff

--

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/e4e06860-90fd-4de7-8a64-5bc2edb3496c%40googlegroups.com.

[Dalibor]

Hi Jeff,

I guess it is a leak along one of the pins, dumet seals can corrode (it is a copper plated nickel-iron wire). I would be interested to do a leak test on a helium leak detector. Of course if you can miss the tube.

Just a very small (tiny tiny!) amount of air is enough to ruin the balance of the gas inside a nixie tube. Less than one would actually think.

Regards,

Dalibor

Dne pátek 8. července 2016 20:48:04 UTC+2 Jeff Walton napsal(a):

[gregebert]

Well.....this question was inevitable.....Could the tube be regassed ?

[Dalibor Farný]

Hello,

if the tubes is faulty, there must be a cause. If any of the pins is
leaky, it is practically not repairable. To refill the tube, you need
to open it and weld a glass tube onto remaining of previous evacuation
pipe. This is tricky because old tubes are usually made of soft glass
which is prone to cracking. Generally, it would be so much work with
uncertain result.. It doesnt make sense to refill a tube unless it is
some very rare piece.

Dalibor

2016-07-12 18:34 GMT+02:00 gregebert <greg...@hotmail.com>:
> Well.....this question was inevitable.....Could the tube be regassed ?
>

> --
> You received this message because you are subscribed to a topic in the
> Google Groups "neonixie-l" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/neonixie-l/A3_V-VYndT4/unsubscribe.
> To unsubscribe from this group and all its topics, 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/17b65b1a-2c04-4b44-882f-6f7157b3678e%40googlegroups.com.

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

--
Dalibor Farny
----------------------------------------
phone: +420 724 321 571
http://www.daliborfarny.com
FB: https://www.facebook.com/daliborfarnycom

Glow Tube Research, s.r.o.
Topolna 539, 687 11
Czech Republic
VAT ID: CZ29204551

[Charles MacDonald]

On 16-07-12 02:13 PM, Dalibor Farný wrote:
> Hello,
>
> if the tubes is faulty, there must be a cause. If any of the pins is
> leaky, it is practically not repairable. To refill the tube, you need
> to open it and weld a glass tube onto remaining of previous evacuation
> pipe.

could someone very skilled take the "stack" out of a dead tube and use
it as the "Guts" of a new tubes, or would the the cathodes end up poisoned?


--
Charles MacDonald Stittsville Ontario
cm...@zeusprune.ca Just Beyond the Fringe
No Microsoft Products were used in sending this e-mail.

[Dalibor Farný]

It is possible, the cathodes are usually stainless steel and they do
not get poissoned on air. Once the stack is mounted and sealed in new
tube, it is necessary to do ageing procedure again..

Dalibor

> --
> You received this message because you are subscribed to a topic in the
> Google Groups "neonixie-l" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/neonixie-l/A3_V-VYndT4/unsubscribe.
> To unsubscribe from this group and all its topics, 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/578541FA.4080606%40zeusprune.ca.

[Jeff Walton]

 

 

Dalibor,

 

As an expert on the materials used in the manufacture of nixie tubes, would you recommend any method of cleaning or neutralizing the corrosion of the pins and seals?  Is there any chemical that could neutralize the chemical reaction of the pins which would slow down the process? 

 

Sometimes parts are shipped in bags with materials that protect the finish.  In my early days with Texas instruments, IC’s were all shipped with a silver plating on pins which would tarnish badly if the shipping bags were unsealed and they often shipped them in containers with a chemical that was more chemically active in air to remove sulfides and etc to prevent tarnishing.  They also used coated shipping rails that were static resistant with corrosion inhibitors.   Might there be something that could be placed inside a clock housing which would sacrifice itself to prevent oxidation/corrosion of the tube pins?

 

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Dalibor
Sent: Tuesday, July 12, 2016 3:32 AM
To: neonixie-l
Subject: [neonixie-l] Re: IN-18 Typical Failure Mechanism

 

 

I guess it is a leak along one of the pins, dumet seals can corrode (it is a copper plated nickel-iron wire). I would be interested to do a leak test on a helium leak detector. Of course if you can miss the tube.

 

Just a very small (tiny tiny!) amount of air is enough to ruin the balance of the gas inside a nixie tube. Less than one would actually think.

 

Dalibor

 

Dne pátek 8. července 2016 20:48:04 UTC+2 Jeff Walton napsal(a):

 

I recently had a IN-18 failure in a clock after about 2.5 year operation.  Over a (4) day period, the digits in one of the tubes stopped working in PWM dimmed mode.  At full brightness, the digits in the affected tube slowly lost percentages of digit coverage to where only portions of each digit would illuminate.  At the end of the observed period, the digits in the failing tube went completely dark and even higher voltage on the bench would not illuminate any of the cathodes.  The tube date code was 8505.

 

I am guessing that absent of any other signs of damage, a seal around one of the pins has failed and allowed the tube to leak.  It was surprising how quickly it failed and none of the other tubes in the same clock have exhibited any signs of trouble and look as good as the first day of operation.  Has anyone experienced similar failures with the IN-18?  Is this a typical failure mechanism?  Have there been any particular date codes that others have found to be problematic?

 

The tube has been since been replaced and the clock is happy again but curiosity leads me to ask about this experience to the rest of the members of the group.  FYI - these tubes have not been rotated since the clock was initially put into operation (no handling).  There was never any evidence of cathode poisoning during the operation of the 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 email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/0ec52c6e-eefb-43f8-8e61-349fa55a8e54%40googlegroups.com.

[John Rehwinkel]

Even on that picture, pin 12 looks suspect, as if the glass around the pin is cracked.

- John

[Jeff Walton]

John, 

The envelope had already been cut open when the photo was taken.   You might note the powder from the glass cutting in the evacuation stem... 

-------- Original message --------
From: John Rehwinkel <jre...@mac.com>
Date: 7/13/2016 3:01 PM (GMT-06:00)
To: neoni...@googlegroups.com
Subject: Re: [neonixie-l] IN-18 Typical Failure Mechanism

--
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/D127611F-B763-4D4C-9096-A7DD6DAA77AE%40mac.com.

[Dalibor Farný]

Hi Jeff,

I dont think there is a simple solution to protect the pins from
corrosion. Even the air humidity can develop leaks in the seals with
time. I know that JJ vacuum tube manufacturer uses gold coating on
pins of some of their tubes to prevent corrosion (I dont know if it
has actually any effect), but this is difficult to do on existing
nixie tubes. About the leaks itself - I heard that Tesla once produced
large batch of stems with small leaks. They decided not to trash them,
but apply a film of varnish (really) around the pins on finished tubes
to stop leaking. But this is rather ridiculous story than a
recomendation ;-)

Cheers,

Dalibor

> You received this message because you are subscribed to a topic in the
> Google Groups "neonixie-l" group.
> To unsubscribe from this topic, visit
> https://groups.google.com/d/topic/neonixie-l/A3_V-VYndT4/unsubscribe.
> To unsubscribe from this group and all its topics, 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/5785aebb.515d240a.29439.191b%40mx.google.com.

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

[Terry S]

Electronic components are typically stored in nitrogen for long term storage and lead finish preservation.

Terry

[Jeff Walton]

Since my initial post about an IN-18 failure in early July, I have had two more failures of IN-18 tubes.  I currently have (5) different clocks using IN-18 tubes which were all sourced as NOS from different eBay vendors.  Interestingly, the failures have all come from a Blue Dream clock (~3 years old) and none from any of the PV clocks of which the oldest one is about 2.5 years old.  The second tube that failed was a date code 03-85 and it failed very quickly and completely like the first tube with complete failure in about 10 days from start to finish.  The latest failure is similar but has continued to work for the last month with a very slow progression of failure from the bottom on the "0", "8", "5" and "6".  The other digits are not yet affected. The date code on the latest failing tube is 05-81.

I pulled the latest failing tube and will leave it out of circuit for a couple months and try it again later to see if it has deteriorated further.

In the meantime, I am wondering if there is any connection of the failures to the type of sockets that are used in the Blue Dream clock, which are the individual socket cups that are soldered flush to the PCB and are not flexible vs the raised socket pins with the acrylic holders that PV uses in their kits?  I do not rotate the tubes in any of the clocks (they all have cathode poisoning prevention).  The Blue Dream sockets are much more rigid than the sockets in the PV Spectrum clocks.  I'm wondering if there is a higher latent stress to the base of the tube that is contributing to a failure of the pin seals in the tubes that live in the Blue Dream?

All of the clocks are running with ~170v HV and run full brightness during the day and PWM dimmed at night so electrically, the tubes in the different clocks are being driven with similar voltages and current.  The rest of the original tubes in the Blue Dream work like the day they were new and the Blue Dream has performed flawlessly otherwise.  All of my clocks are maintained indoors and it is neither really dry or humid in any of the locations.

Attached are photos of the tube and pins; a photo of the "8" in operation and a photo showing a transition between "2" and "3" which shows no evidence of failure.  The "5" and "6" look just like the "8" in terms of the failure from the bottom.  The display photos are in the PWM dimmed mode.

I would be interested in other thoughts or opinions that anyone would have about the IN-18 failures.  Coincidence or a pattern?  The first two failures have been destroyed but the latest failure is still available for other photos. 

On Friday, July 8, 2016 at 1:48:04 PM UTC-5, Jeff Walton wrote:

I recently had a IN-18 failure in a clock after about 2.5 year operation.  Over a (4) day period, the digits in one of the tubes stopped working in PWM dimmed mode.  At full brightness, the digits in the affected tube slowly lost percentages of digit coverage to where only portions of each digit would illuminate.  At the end of the observed period, the digits in the failing tube went completely dark and even higher voltage on the bench would not illuminate any of the cathodes.  The tube date code was 05-85.

[Peter H]

I had this (I mean the pic with the 8) on several IN-18 Tubes, it was cathode poisoning, at least I thik it was. The tube were sold as NOS, 1979. All could be regenerated by a higher current for several hours, each digit.It works good again since 2 months.

I did it with the very professional and nice device from ebay:

Professional Nixie Tube Tester & Healer - For All Nixie Tubes 

 

Artikelnr.: 262483292894 , nixie_ninja

I'm not the seller of the device.

The result is not 100% but good.

[Jeff Walton]

Thanks for your suggestion.  This current IN-18 failure is definitely different from the first two failures, as the first two went completely dark for all digits in a short time and could not be lit at any voltage.  I considered those to be failure of the seals and leakage.

 

I thought that cathode poisoning exhibited a more irregular or random appearance but I can try pushing higher current through each cathode and see if it will can correct the  coverage.  I have previously rejuvenated B7971 tube segments, which showed as random dark gaps on illuminated segments by using high current.  I have not tried using the same technique on the IN-18.  Do you recall the current level that was necessary for your tubes?

 

Jeff

 

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Peter H
Sent: Tuesday, November 01, 2016 10:19 AM
To: neonixie-l
Subject: [neonixie-l] Re: IN-18 Typical Failure Mechanism

 

 

I had this (I mean the pic with the 8) on several IN-18 Tubes, it was cathode poisoning, at least I thik it was. The tube were sold as NOS, 1979. All could be regenerated by a higher current for several hours, each digit.It works good again since 2 months.

 

I did it with the very professional and nice device from ebay:

 

Professional Nixie Tube Tester & Healer - For All Nixie Tubes 

 

Artikelnr.: 262483292894 , nixie_ninja

 

I'm not the seller of the device.

The result is not 100% but good.

 

-----------------------------------------------------------------

Am Freitag, 8. Juli 2016 20:48:04 UTC+2 schrieb Jeff Walton:

I recently had a IN-18 failure in a clock after about 2.5 year operation.  Over a (4) day period, the digits in one of the tubes stopped working in PWM dimmed mode.  At full brightness, the digits in the affected tube slowly lost percentages of digit coverage to where only portions of each digit would illuminate.  At the end of the observed period, the digits in the failing tube went completely dark and even higher voltage on the bench would not illuminate any of the cathodes.  The tube date code was 8505.

 

I am guessing that absent of any other signs of damage, a seal around one of the pins has failed and allowed the tube to leak.  It was surprising how quickly it failed and none of the other tubes in the same clock have exhibited any signs of trouble and look as good as the first day of operation.  Has anyone experienced similar failures with the IN-18?  Is this a typical failure mechanism?  Have there been any particular date codes that others have found to be problematic?

 

The tube has been since been replaced and the clock is happy again but curiosity leads me to ask about this experience to the rest of the members of the group.  FYI - these tubes have not been rotated since the clock was initially put into operation (no handling).  There was never any evidence of cathode poisoning during the operation of the clock.

--

[Nicholas Stock]

I've used this device to recover a couple of tubes before...the max current the device will give is 10 mA I believe. This proved fine for the IN18 tubes....which get quite warm after a while!!!

Nick

--
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+unsubscribe@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/5818cf0c.41e1240a.28851.a56c%40mx.google.com.

[Mich...@aol.com]

Pete,

 

I suggest Better linking to ebay items like this:

http://www.ebay.com/itm/262483292894

 

I thought something was strange when it was trying to make me login with a German login.

 

It's a pretty unit, although I have never experienced a poisoned tube yet to need it. (Fortunately)

 

Michail

 

 

In a message dated 11/1/2016 8:18:51 A.M. Pacific Daylight Time, peter.hu...@gmail.com writes:

I had this (I mean the pic with the 8) on several IN-18 Tubes, it was cathode poisoning, at least I thik it was. The tube were sold as NOS, 1979. All could be regenerated by a higher current for several hours, each digit.It works good again since 2 months.

I did it with the very professional and nice device from ebay:

Professional Nixie Tube Tester & Healer - For All Nixie Tubes 

 

Artikelnr.: 262483292894 , nixie_ninja

I'm not the seller of the device.

The result is not 100% but good.

Am Freitag, 8. Juli 2016 20:48:04 UTC+2 schrieb Jeff Walton:

I recently had a IN-18 failure in a clock after about 2.5 year operation.  Over a (4) day period, the digits in one of the tubes stopped working in PWM dimmed mode.  At full brightness, the digits in the affected tube slowly lost percentages of digit coverage to where only portions of each digit would illuminate.  At the end of the observed period, the digits in the failing tube went completely dark and even higher voltage on the bench would not illuminate any of the cathodes.  The tube date code was 8505.

I am guessing that absent of any other signs of damage, a seal around one of the pins has failed and allowed the tube to leak.  It was surprising how quickly it failed and none of the other tubes in the same clock have exhibited any signs of trouble and look as good as the first day of operation.  Has anyone experienced similar failures with the IN-18?  Is this a typical failure mechanism?  Have there been any particular date codes that others have found to be problematic?

The tube has been since been replaced and the clock is happy again but curiosity leads me to ask about this experience to the rest of the members of the group.  FYI - these tubes have not been rotated since the clock was initially put into operation (no handling).  There was never any evidence of cathode poisoning during the operation of the 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 email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/ad2e8bfc-7e02-4e9c-a82d-0dc01a6c6b20%40googlegroups.com.

[gregebert]

Results will vary by manufacturer/tube construction.

I have a few 5031 nixies that appear to have cathode-poisoning (partial cathode illumination, and in some cases only the interconnects glow rather than the digit), but they have not had much improvement with high-current 'cleansing'. Nominal current is 2.2mA, and I've been pumping up to 20mA (the limit of my HV supply). It takes quite a bit more voltage to get that current, in the ballpark of 200V, compared to typical 150-160V in normal operation. Yeah, they get really warm.....These are early production nixies, so I'm certain they lack better manufacturing techniques used on the 5092.

Fortunately, none of my 'better' nixies have required rejuvenation.

BTW, if you have a current-limited supply you can actually melt a tube (such as a 0A2) and watch the glass shrivel around the red-hot metal guts. Hilarious.

[Alic]

The sockets on the blue dream seem to be less forgiving indeed : I managed to crack an IN-18 not long ago because the pins were not straight enough on that tube I bought on eBay. :-(
That was on my latest clock.
The 2 blue dream clocks I assembled before that still work well after 3 and 4 years. The tubes for these older clocks where purchased from different sources: GS-tubes sold me ones from '75 and Dieter from '92.

[Nick]

Some kits I've built which use pin sockets (MillMax etc.) use over-large holes for the pins so they have some movement - you number the tubes 1-4, corresponding to their placement on the board - the pins are tacked onto the board with the tube in place, then soldered fully - this way the pins for a particular position align correctly with the tube in that position, thus minimizing the stress on that tube. This rationale obviously breaks down if you have to change a tube for some reason.

Personally, I don't like soldered MillMax pins much as their use normally assumes that all tubes are perfect and identical. Sockets are good as they allow for variation.

Nick