In18 displaying cathode poisoning after 1000hrs or less.

[Trumpeter]

Hello all,

I have what was a nos set of tubes in a spectrum 18, running at 170v. At least two tubes display incomplete digits. I have the PIR sensor installed it may be even less than 1000 hours. Could I be doing something wrong or just a bad batch of tubes? They are all the same date code.

[Gene Segal]

What is the date code?  It may not necessarily be cathode poisoning.  Partial digit illumination can also be caused by slow outgassing, which may happen when a tube is installed in tight-fitting socket pins which do not yield to thermal expansion, thus causing separation at the pin-glass interface of the tube (((

Trumpeter

February 18, 2017 at 7:11 PM

Hello all,

I have what was a nos set of tubes in a spectrum 18, running at 170v. At least two tubes display incomplete digits. I have the PIR sensor installed it may be even less than 1000 hours. Could I be doing something wrong or just a bad batch of tubes? They are all the same date code.

--

Sent from Postbox

[Trumpeter]

I don't think it's the sockets they are fairly loose fitting. I can rock the tubes side to side with little effort.

[Trumpeter]

09 81 is the date code.

[Nicholas Stock]

Probably just bad luck....the pins on the spectrum are pretty forgiving, so even though Gene makes a good point, that's probably not what's going wrong. Funnily enough, I just had my first IN18 failure in an old PV electronics LTC clock...been running for over 3 yrs now and one tube looked like it was poisoned on a few digits. I attempted to regenerate the digits and thought I was successful, but two days later the same problems arose and some digits quickly disappeared altogether.....oh well, adios bad tube! Stuff happens and all that....

Cheers,

Nick

Sent from my iPhone

--
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/58A90E7F.8030208%40earthlink.net.
For more options, visit https://groups.google.com/d/optout.

[Trumpeter]

Oh well I guess I'll run them until the digits are unreadable. Just bummer given the cost of a new set, nothing cool is cheap these days.

[Trumpeter]

Huh, I notice that the number 8 on two tubes is failing to illuminate in the exact same place on the digit.

[Nicholas Stock]

Are we allowed to ask where you purchased the tubes from?

Sent from my iPhone

> On Feb 18, 2017, at 20:14, Trumpeter <dmatr...@gmail.com> wrote:
>
> Huh, I notice that the number 8 on two tubes is failing to illuminate in the exact same place on the digit.
>

> --
> 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/43a3c121-290c-4a90-b7d2-96e595a2fca8%40googlegroups.com.

[Jeff Walton]

Trumpter;

You can rejuvenate them if it isn't an outgas problem.   

I haven't had any tubes fail on the Spectrum 18, which has fairly loose socket pins,  but I have lost 2 to seal failures on the Blue Dream clock,  which has the really tight pin "cups".  I've also had a couple of tubes that have showed cathode poisoning which have been rejuvenated and are still OK after about 8 months of operation. 

To rejuvenate,  run the failing digits, one at a time,  on a power supply at 170-180vdc that lets you apply 8-10mA to the cathodes at that are partial and that may clear the deposits.  Some say it is OK to use 2-4x the rated current but I try to go easy.  If it is truly cathode poisoning, it could take care of the problem for a long time.   If a pin seal is failing, you may not be able to clear the cathodes and the tube will get progressively worse until all digits fail.  The process could take minutes or hours for each digit.   Check from time to time by lowering the current back to 3-4mA to see if they are cured at the regular current. 

Spares are also a good solution... 

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 an email to neoni...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/d2087e71-e96c-4ddf-9af1-94f6e247e3f6%40googlegroups.com.

[Trumpeter]

Jeff, could you reccomend a power supply capable of this procedure? Thank you for your input.

[Nicholas Stock]

I have one of these...

https://www.ebay.com/itm/262610935844

Works well for this kind of thing...

Cheers,

Nick

Sent from my iPhone

> On Feb 18, 2017, at 21:20, Trumpeter <dmatr...@gmail.com> wrote:
>
> Jeff, could you reccomend a power supply capable of this procedure? Thank you for your input.
>

> --
> 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/16b65d4e-98f8-40c0-a45a-068da673f645%40googlegroups.com.

[Jeff Walton]

I have the same tool that Nick suggested. 

http://m.ebay.com/itm/262610935844?_mwBanner=1

Jeff 

-------- Original message --------
From: Trumpeter <dmatr...@gmail.com>
Date: 2/18/17 11:20 PM (GMT-06:00)
To: neonixie-l <neoni...@googlegroups.com>
Subject: Re: [neonixie-l] In18 displaying cathode poisoning after 1000hrs or less.

[Trumpeter]

Thanks everyone, I'll grab one and give it a go.

[Manuel Azevedo]

Hi guys,

I also have a strange situation.
I have two Spectrums - a little older model (3 years ago) and one bought last year.

On the older unit I had 6 tubes with 9003 markings (I assume March 1990, but the order is reversed from the other tubes). Never had any problem with this tubes.
On the newer unit I had 4 tubes with 0686 markings (I assume August 1986) and 2 tubes 0982 (I assume August 1982 - coincidently these tubes are 3 or 4 mm taller than all the rest).

After running for a few months on the newer unit, I noticed on one of the 1982 tubes, that the digit 8 was starting to fade. Looked like cathode poisoning, but I already had a 1984 tube with cathode poisoning on a Kosbo 1 digit clock, and it didn't quite look 100% the same.

I've replaced the tube with one of the old unit and after a few days, the digit 8 showed again complete.

Put the tube back into the new unit and it was OK for a few months, until on one tube the 8 was back again losing visibility and a 0 on another tube was starting to fade in the bottom.

Yesterday I had enough - I replaced all the tubes between units and now all are working OK without any "poison"  symptoms.

I've noticed that the older unit's tubes seem to shine a little bit brighter than the newer unit.

I guess that both units might not be delivering the same current, hence these issues.

Regards,

Manuel

[threeneurons]

I took a look at the schematic, and anode resistors (R2-R7, direct drive) are 10K. The voltage drop across each of these resistors should be at least 40V, for 4mA MINIMUM anode current. Put a meter across the anode resistor of an offending tube, and report back the voltage.

[newxito]

I assembled one of these kits yesterday, the anode resistors are now 8.2K

[Manuel Azevedo]

On the old unit, I can see that they all are 10K 1% - it’s in an acrylic case, so easy to see.

In the afternoon, I’ll check the new unit (metal case, hard to disassemble) and make the measurements.

 

[Gene Mark Segal]

I hear of far more failures with early-mid 1980's tubes than 1989-91 ones. Do yours have a mercury giver tube in the base? It's a tiny glass rod with a heater wire wrapped around it. 

Sent from my iPhone

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/EDDF6A60-12DF-4386-931D-1A9184FDB135%40gmail.com.

[Manuel Azevedo]

The old clock was/is set at 173.1V. The new one was at 168.5V.
 
So before testing, I did the following:

·         Set both clocks to 173.1V without any tube attached
·         Set both clocks to the exact same settings (tube dimming starts at 3:00am, so no tube dimming during tests)
·         Set both clocks to setup mode so we have the same digits lit “01 01 99”
 
At each clock, I measured the following:
 
               Older      Newest
HV          173.1     173.1
R2           43.92     38.8
R3           39.1        37.1
R4           36.6        36.6
R5           35.2        37.2
R6           36.4        33.6
R7           35.6        36.9
 
My older clock seems to be a Rev5 clock, because on the documentation it says 10k for R2-R7. However, Rev5 mentions R15 and both my clocks don’t have R15. I don’t remember contacting Pete about R15, but it’s not there. Don’t see it also on the schematics of Rev5 documentation.
The Rev6 documentation mentions R2-R7 with 8k2 and does not mention R15 at all.
 
Conclusions so far:
 

• The newest clock was set for 168.5V. This by itself should not be a problem, I suppose, but I increased it to 173.1V
• While testing both clocks, I used different power supplies than their regular ones. One of them clearly had a problem: on the new clock, under load I got on HV 145V… Clearly a bad power supply. When replacing with another, both clocks had 173.1V when on load.
• I’m going to replace the new clock with a better power supply
• The “poisoned” tubes I moved to the old clock seem to be healing
• I’ve also checked all my tubes have the giver at the base

 
I’m inclined to be believe that the switched power supply might not behaving correctly. I’m replacing it on the new clock… Maybe that was the problem...
 
I’ll assemble everything again and keep an eye on the clocks.

[Gene Mark Segal]

The "8" digit has the greatest surface area and requires the most current to light, so for a partially outgassed and/or internally contaminated tube driven with a fixed current, the 8 is often the first to show symptoms of partial illumination.  

--
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/1dec2b55-291f-436b-b996-204c9da3732c%40googlegroups.com.

[Gene Mark Segal]

PS: correction - I should have probably wrote "partially in-gassed" as I believe the pressure inside the tubes is less than one atmosphere (someone correct me if I'm wrong), and so unwanted oxygen/nitrogen etc. gets into the tube. Any way you slice it, the tubes gets contaminated and requires higher voltage/current to achieve full ionization of the neon/argon. 

Sent from my iPhone

On Feb 19, 2017, at 12:50 AM, Manuel Azevedo <azevedo...@gmail.com> wrote:

[gregebert]

Read-up on Paschen's Law, which can explain why a leaky tube is harder to ionize after it's pressure starts to increase.

I wonder if we should store our nixies in a hyperbaric chamber with neon/argon.

[Paul Andrews]

I would expect that a leaky tube would be suffering from contamination. This wiki article is a good description of how these tubes work: https://en.wikipedia.org/wiki/Gas-filled_tube#Gas_purity

[Dekatron42]

In tubes, filled with Neon and other gases, with shiny copper pieces you can see that these copper pieces become darker and darker (black) the more air that has entered the tube, the tube can work just as designed for a long time while this process is ongoing but it will sooner or later give up.

/Martin

[John Rehwinkel]

> Read-up on Paschen's Law, which can explain why a leaky tube is harder to ionize after it's pressure starts to increase.
>
> I wonder if we should store our nixies in a hyperbaric chamber with neon/argon.

Both effects happen: the neon and argon leak out because even though they're at lower than atmospheric pressure, they're higher than the partial pressure of neon and argon outside the tube. And, more intuitively, oxygen and nitrogen leak in. Both of these effects are detrimental to the tube's operation.

Old-style frit (or "soft") seal helium-neon laser tubes have a similar problem. The helium (a very slippery gas) would slowly leak out through the seals, so although the tube would still ionize and glow, laser action wasn't supported. The cure is surprisingly simple: "soak" the tubes in a trash bag with helium in it at atmospheric pressure. With the reversed (and much larger) partial pressure difference of helium, a few years' loss could be restored in a few days. The normal setup was to leave the tube running and monitor the laser output. It would slowly rise, then begin to drop when the helium pressure began to get too great. Take the tube out of the bag then and you could get a few more years out of it.

In short, you don't need a hyperbaric chamber, atmospheric pressure will probably suffice, but you'd want to have the proper ratio of gases.

- John

[NeonJohn]

On 02/20/2017 11:44 AM, John Rehwinkel wrote:

> Both effects happen: the neon and argon leak out because even though
> they're at lower than atmospheric pressure, they're higher than the
> partial pressure of neon and argon outside the tube. And, more
> intuitively, oxygen and nitrogen leak in. Both of these effects are
> detrimental to the tube's operation.

Yes. The uber-killer, though is helium. It goes through soft glass
like water through a paper towel.

At one nuclear plant where I was setting up their radiation monitoring
system, I suddenly started losing photomultiplier tubes in liquid and
gas monitors. I traced the problem down to another contractor doing
helium leak tests on the monitors (why? they operate at atmospheric
pressure) and instead of the normal procedure of hooking the mass spec
up to the detector and then flooding the monitor with helium, they were
pressurizing the detection chamber and sniffing with a probe connected
to the leak detector.

I'm talking about exposure of just a few minutes.

>
> Old-style frit (or "soft") seal helium-neon laser tubes have a
> similar problem. The helium (a very slippery gas) would slowly leak
> out through the seals, so although the tube would still ionize and
> glow, laser action wasn't supported. The cure is surprisingly
> simple: "soak" the tubes in a trash bag with helium in it at
> atmospheric pressure. With the reversed (and much larger) partial
> pressure difference of helium, a few years' loss could be restored in
> a few days. The normal setup was to leave the tube running and
> monitor the laser output. It would slowly rise, then begin to drop
> when the helium pressure began to get too great. Take the tube out
> of the bag then and you could get a few more years out of it.

I've done several of those. Amazing how fast the He goes back in.

>
> In short, you don't need a hyperbaric chamber, atmospheric pressure
> will probably suffice, but you'd want to have the proper ratio of
> gases.

Just heat-seal the tubes in metallized mylar bags, the same stuff that
modern helium balloons are made of.

And keep helium out of your house! No helium balloons or whatnot.

I'll bet that if one did a spectroscopic analysis on the light coming
from a weak tube, one would see He in the mix.

John


--
John DeArmond
Tellico Plains, Occupied TN
http://www.tnduction.com <-- THE source for induction heaters
http://www.neon-john.com <-- email from here
http://www.johndearmond.com <-- Best damned Blog on the net
PGP key: wwwkeys.pgp.net: BCB68D77

[Trumpeter]

Are you saying the helium from leaking balloon can damage these tubes? My kids have a couple in the living room now.

Directed to the community at large:

I am in search of a spare set of tubes, is it recommended to buy tubes from 89-93ish? These have a lower failure rate? Short of messaging sellers out of the blue with this request I do not see these listed in that way (by dates).

[Manuel Azevedo]

Just an update on this issue.

 

Indeed, it was a faulty power supply.

 

After I changed it in February, after this email, the clock did not show any signs of improvement.

 

I did change it yet with another power supply somewhere in March, but the issue maintained.

 

One of the digits started to show some discoloration and then parts of the digits where completely off.

I healed it, but this time I swapped the power supplies from the old and the new clock.

The clock was noticeably brighter and the healed tube had some discoloration, but no dark spots.

Today, 1 day later, the tube is fully healed – no more discoloured spots. On the old clock, the tubes seemed dimmer since then.

 

I measured the current drawn by each clock. The new one was drawing about 0.7A. The old one 0.5A. They both have a nwts and a PIR sensor.

 

I replaced the power supply on the old clock with a more powerful one (and of a better brand), and now both clocks are equally bright. The old one draws about 0.68A now.

 

It’s now clear that I was using a faulty power supply.

 

Manuel

--
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/cgk1rEuT1lw/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/a35478e0-184b-4449-a765-f7e9f7c0ad61%40googlegroups.com.

[gregebert]

Can you determine (or measure) the current consumed by the nixes ?  If it's a direct-drive clock and the voltages you provided are measured across the anode resistor (10K ?) the current is on the low-side, roughly 3.3 to 3.9mA. I believe the IN-18 is spec'd for 4-6mA.

[Manuel Azevedo]

Guess that’s too late for me now ☺

I’m sure all is OK now – it really was a broken power supply.

 

 

From: <neoni...@googlegroups.com> on behalf of gregebert <greg...@hotmail.com>

Reply-To: "neoni...@googlegroups.com" <neoni...@googlegroups.com>
Date: Monday, 8 May 2017 at 20:29
To: neonixie-l <neoni...@googlegroups.com>

[Trumpeter]

Hi folks,

I have had some time to troubleshoot my clock. I found that when running in night mode the tubes are supplied with around 1 mA, and when in day mode are supplied 5 mA. These readings are taken between the annode resistor (10k) and annode. My understanding is the tubes should receive between 4-6 mA. Could the lower night mode current be causing the problems with positioning? I did confirm the tubes are suffering poisioning because I have been able to rejuvenate them by running the offending digits at 10 mA for between 30 min to 3 hours. I'm thinking of running the clock on day mode continuously to see if the tubes begin to heal themselves. Thoughts? Anything else I should be looking at?

[Jeff Walton]

Were these currents measured by scope waveform?  

There is a big difference between a 1mA current at 100% duty cycle and, for example, 5mA pulses @ a 20% duty cycle, or 8mA pulses @ a 10% duty cycle.   Any of these examples will change the effective brightness.  The 1mA example will likely cause cathode poisoning.  A typical multi meter doesn't give an accurate picture of the actual voltage and current of pulsed waveforms. 

If you run the clock (long term) at full brightness, do you experience any poisoning on the seconds or minutes tubes? Those two tubes are the best to judge because all cathodes are exercised equally. 

Cathode poisoning prevention cycles require full brightness cycling to be effective and in many clocks are disabled during dimmed operation.  The brightness would be an annoyance in a dark room. 

Check to see if your voltage at the anode is within spec and that your power supply is providing full rated power under all conditions.  A faulty power supply can also "under-drive" the tubes and that will prevent the "cleaning" effect that occurs with normal operation. 

What kind of clock are you running?  In a good clock design, your tubes should "heal" (after awhile) if run at full brightness. 

Jeff 

-------- Original message --------
From: Trumpeter <dmatr...@gmail.com>

Date: 5/13/17 11:53 PM (GMT-06:00)
To: neonixie-l <neoni...@googlegroups.com>
Subject: Re: [neonixie-l] Re: In18 displaying cathode poisoning after 1000hrs or less.

--

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/9a75cb3d-6e91-4fcc-9827-a275dda57385%40googlegroups.com.

[Trumpeter]

I did not scope it but it is a direct drive not multiplexed. I don't have a scope yet. I did measure the power supply is putting out 173 volts, where I set it when I built the clock.

[Trumpeter]

Correction it's set at 170v.