Another Tube Hobby Failure

[MichaelB]

This makes 4 times now this clock has failed over the past 10 years or so. This is a stock clock with IN-18's. I have another set of these electronics that has never failed, but then again, I had changed the tube board to accommodate the Z566M tube. Different current demands? It has worked like a charm for years now. The failure with the IN-8 clock this time was a leaky C6. In the past the inductor has failed, but usually its one of the electrolytics in the Pwr supply stage. It's become kind of a ritual where every 2-3 years I have to pull the clock apart and play detective and figure out what's wrong after its starts blowing fuses. Kind of fun now, actually!

[Nicholas Stock]

Something else is awry there Michael. I have a friend with a TubeHobby IN18 that i built for him that’s been running for well over 6 years now with no issues. The only modification i ever made was putting an extra heatsink on the regulator as what was there (board plane) didn’t seem sufficient to me (just a hunch, I’m sure Jonas knew what he was doing..)....

Hope you get it sorted!

Sent from my iPhone

On Apr 6, 2019, at 17:15, MichaelB <badn...@badnixie.com> wrote:

This makes 4 times now this clock has failed over the past 10 years or so. This is a stock clock with IN-18's. I have another set of these electronics that has never failed, but then again, I had changed the tube board to accommodate the Z566M tube. Different current demands? It has worked like a charm for years now. The failure with the IN-8 clock this time was a leaky C6. In the past the inductor has failed, but usually its one of the electrolytics in the Pwr supply stage. It's become kind of a ritual where every 2-3 years I have to pull the clock apart and play detective and figure out what's wrong after its starts blowing fuses. Kind of fun now, actually!

--
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/07394dde-1d45-4e3f-a38a-75ca15d16c48%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

<Tube Hobby Pwr Supply Schem..jpeg>

[MichaelB]

HI Nick, 

I know you have mentioned that before, but I really don’t think anything else is wrong. As I said I have another one of these with 566M tubes and it has worked flawlessly for 8+ years. Those electronics are in a similar enclosure with less ventilation. And this is the second set of electronics i have tried in this enclosure thinking there might be something internally wrong with the PCB. (I had that problem with one of my Thomas clocks way back when) But, I believe others have had issues with these electronics. Another variable vs. your clocks might be that I have wired in a pair of NE-2 Colons in parallel vs. the stock config that uses just one per each side. I doubt that would be an issue, but it may account for a slightly greater current demand.  I don't think these clocks like anything, but an open air set up. Also, the close proximity of the 5V VREG and the IRF640 is somewhat suspect from a heat dissipation perspective. Anyway…I used a higher voltage capacity electrolytic this time, maybe I’ll this'll give me another year or so :-). and I will try your idea of adding a heat sink to the VREG, since heat seems to be the culprit here. Thx Nick

On Saturday, April 6, 2019 at 6:06:36 PM UTC-7, Pramanicin wrote:

Something else is awry there Michael. I have a friend with a TubeHobby IN18 that i built for him that’s been running for well over 6 years now with no issues. The only modification i ever made was putting an extra heatsink on the regulator as what was there (board plane) didn’t seem sufficient to me (just a hunch, I’m sure Jonas knew what he was doing..)....

Hope you get it sorted!

Sent from my iPhone

On Apr 6, 2019, at 17:15, MichaelB <badn...@badnixie.com> wrote:

This makes 4 times now this clock has failed over the past 10 years or so. This is a stock clock with IN-18's. I have another set of these electronics that has never failed, but then again, I had changed the tube board to accommodate the Z566M tube. Different current demands? It has worked like a charm for years now. The failure with the IN-8 clock this time was a leaky C6. In the past the inductor has failed, but usually its one of the electrolytics in the Pwr supply stage. It's become kind of a ritual where every 2-3 years I have to pull the clock apart and play detective and figure out what's wrong after its starts blowing fuses. Kind of fun now, actually!

--
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 neoni...@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/07394dde-1d45-4e3f-a38a-75ca15d16c48%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Mac Doktor]

On Apr 6, 2019, at 9:36 PM, MichaelB <badn...@badnixie.com> wrote:

I used a higher voltage capacity electrolytic this time, maybe I’ll this'll give me another year or so :-).

Are you using 85°C or 105°C?

and I will try your idea of adding a heat sink to the VREG, since heat seems to be the culprit here. Thx Nick

Never turn your back on a 78xx. You could try using a 317 instead (requires a resistor).

Terry Bowman, KA4HJH
"The Mac Doctor"

[Terry Kennedy]

On Saturday, April 6, 2019 at 8:15:57 PM UTC-4, MichaelB wrote:

This makes 4 times now this clock has failed over the past 10 years or so. This is a stock clock with IN-18's. I have another set of these electronics that has never failed, but then again, I had changed the tube board to accommodate the Z566M tube. Different current demands? It has worked like a charm for years now. The failure with the IN-8 clock this time was a leaky C6. In the past the inductor has failed, but usually its one of the electrolytics in the Pwr supply stage. It's become kind of a ritual where every 2-3 years I have to pull the clock apart and play detective and figure out what's wrong after its starts blowing fuses. Kind of fun now, actually!

I've had many of these clocks fail, repeatedly. It always seems to be the HV capacitor. Sometimes it takes out the fuse, sometimes it doesn't. I've had a few where the AC adapter has died. At this point when I get one back for repair, I socket the capacitor because it is a given that I'll see that clock again for the same problem. 

[Roger Brinkman]

I have one of these clocks too and it fails every three years or so on average. It’s all down to C6. Sometimes just F1 fails and other times the inductor L1 goes short circuit as well.

The Equivalent Series Resistance (ESR) of C6 goes through the roof, causing it to effectively be a capacitor in series with a resistance. This result is ringing in L1, and instability in the output voltage which increases L1’s heat dissipation and untimely failure. Ambient temperature is definitely a factor in C6’s lifespan, i.e clocks running in cooler climates may run longer before failing, but the real mechanism of failure is C6’s ability to withstand the internal heat generated by the pulse current as its ESR rises with age. The higher it’s ESR, the warmer it runs, then the quicker its electrolyte dries up.

High quality, low ESR 105 degree electrolytic capacitors designed for pulse applications like the Panasonic WL-R series will have a longer life than ‘ordinary’ electros but eventually even they will fail. I have thought about substituting a Multilayer Ceramic Capacitor in place of electrolytic C6, but currently waiting for the next failure (which is about due!) before I experiment with them. Perhaps another person can comment on their suitability for such an application?

Regards from Australia! 

--
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/0DFF2801-AAC9-4780-9779-E01A35F1612D%40gmail.com.

[Nick]

There's a number of anomalies here. The 7805 has no reverse bias diode and no 100n output cap to prevent oscillation (read the data sheet). Is the 7805 running hot? Some 7805 variants require these, some don't.

R6 has about 180V across it - is it rated for that? There may be RF oscillation on the HV output - have you scoped the HV and LV on the failing unit?

Nick

[MichaelB]

You've described my experience with this clock to a "T". If I hadn't made a custom enclosure around these electronics and its 'unique' tube spacing, I would abandon it in a heartbeat and substitute NixiChron electronics, but this would basically require building a whole new enclosure. It was my 1st enclosure build so I'm a bit attached to it. So I'll keep my fingers crossed and hope I can come up with a solution to this anomaly. Suggestions welcome

On Saturday, April 6, 2019 at 7:55:19 PM UTC-7, Roger Brinkman wrote:

I have one of these clocks too and it fails every three years or so on average. It’s all down to C6. Sometimes just F1 fails and other times the inductor L1 goes short circuit as well.

The Equivalent Series Resistance (ESR) of C6 goes through the roof, causing it to effectively be a capacitor in series with a resistance. This result is ringing in L1, and instability in the output voltage which increases L1’s heat dissipation and untimely failure. Ambient temperature is definitely a factor in C6’s lifespan, i.e clocks running in cooler climates may run longer before failing, but the real mechanism of failure is C6’s ability to withstand the internal heat generated by the pulse current as its ESR rises with age. The higher it’s ESR, the warmer it runs, then the quicker its electrolyte dries up.

High quality, low ESR 105 degree electrolytic capacitors designed for pulse applications like the Panasonic WL-R series will have a longer life than ‘ordinary’ electros but eventually even they will fail. I have thought about substituting a Multilayer Ceramic Capacitor in place of electrolytic C6, but currently waiting for the next failure (which is about due!) before I experiment with them. Perhaps another person can comment on their suitability for such an application?

Regards from Australia! 

On 7 Apr 2019, at 11:59 am, Mac Doktor <themac...@gmail.com> wrote:

On Apr 6, 2019, at 9:36 PM, MichaelB <badn...@badnixie.com> wrote:

I used a higher voltage capacity electrolytic this time, maybe I’ll this'll give me another year or so :-).

Are you using 85°C or 105°C?

and I will try your idea of adding a heat sink to the VREG, since heat seems to be the culprit here. Thx Nick

Never turn your back on a 78xx. You could try using a 317 instead (requires a resistor).

Terry Bowman, KA4HJH
"The Mac Doctor"

--
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 neoni...@googlegroups.com.

[johnk]

Michael, I have to comment on the logic here.

You say you have one that is quite OK and seem to use that fact to rule out the effects mentioned by Nick. Actually, I consider that the parasitic oscillations idea is reinforced by your fact, not denied by it.   When a circuit is on the verge of misbehaving there will be some that are OK and some that aren’t.

Obviously I can’t tell from here what the actual fault is.

 

Capacitors.  I am interested in the topic. Low ESR electrolytics are a curse. [Remember that fuss about the “stolen” formula not being complete and all the computer manufacturers being affected?] Some do seem to last better than expected. Some fail quickly. The LED lighting industry is possibly going to cause improvements. Those tiny electros that they hide in the base of the globe really are stressed. And some sellers still quote the expected LED life as that for the power supply too !     Oops.  [And Australian consumer law provides for lifetime warranty btw, regardless of what the manufacturer states. And we beat up Steam to the point that they allow returns!]

I am very interested in hearing about how well different brands and types of capacitor perform. Anyone tried non-electros?

I have noticed that some switching supplies driving discharge tubes for TV and laptop backlighting [yeah, old ones, not LED ones] use monolithic solid caps. One I fixed had only 2uF though iirc.  A pity that opamp capacitor multipliers don’t store energy J .

 

John K

Australia

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/22340da1-4f58-4673-bb1d-a9bdebb8fd14%40googlegroups.com.

[David Pye]

Couldn't you wire in a small HV PSU module instead, and bypass the circuitry for that one?

Not sure that it would necessarily have a better lifespan (depending on who made it...)

David

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/c55b7b6b-a7d7-4f50-8eb8-10bb79c6a208%40googlegroups.com.

[Paul Andrews]

The short lifetime ratings of electrolytics have always worried me. 8,000 hours? That’s about 1 year, though that is at rated operating conditions. This article discusses dressing of lifespan: https://www.allaboutcircuits.com/news/calculating-the-lifespan-of-electrolytic-capacitors-with-de-rating/

It seems temperature is definitely your enemy here.

[MichaelB]

I've actually thought about wiring in a Taylor supply, but space is a bit of an issue

[MichaelB]

You guys have made me curious now about the other 'good' clock I have of his. I'm going to open it up and see if there are any other differences other than the obvious being tube board I used vs. the 'stock' one.

[MichaelB]

I used a Lelon (the new Xicon) 1uf 350V 85° C vs. the 1uf 250V that's recommended.

[Roger Brinkman]

This is the Multilayer Ceramic Capacitor I had in mind in my suggestion earlier in the thread, datasheet impedance looks suitable, and I suspect its life will be much longer than an electrolytic. Even though it’s an SMD part, it should be able to easily be soldered in place of the electro. I’m going to order some and install, will post results.

Roger Brinkman 

Australia

https://au.element14.com/murata/grm55dr72e105kw01l/cap-1-f-250v-10-x7r-2220/dp/1828845?st=MLCC%20105%20250v

https://au.element14.com/murata/grm55dr72e105kw01l/cap-1-f-250v-10-x7r-2220/dp/1828845?st=MLCC%20105%20250v

--
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/9e8b78a6-afd5-4b0e-8f46-59340b36f948%40googlegroups.com.

[gregebert]

#1 Make sure you are purchasing new units from a reputable supplier; there are fake parts out there that are poor quality. Never re-use salvaged electrolytics.

#2 Heat will shorten the life of any electrolytic; it only takes a few well-placed vent holes to keep things cool.

#3 Not all electrolytics are created equal; ESR and lifetime specs specs vary by manufacturer

#4 Choose a manufacturer that provides datasheet info for extended operation; EPCOS has devices spec'd for 200,000 hours at 40C. Note that any capacitor will last longer at lower temperatures.

[Bill Notfaded]

You mean the smaller Taylor supply won't fit??? I'd find a way to squeezed John's in there. ;^}

[GastonP]

A measurement of ripple levels and frequency would help a lot in diagnosing what's happening.

Also, electrolytics have a maximum rated current, and the higher the one that passes through them, the higher the internal temperature, and in consequence a shorter life. I don't know if multilayer ceramic capacitors have this value specified, but it will affect them for sure.