MTX-90 in counting circuit?

[William Lee]

I may be demonstrating my glaring lack of electronics knowledge with this question, but I figure I will learn something from it either way...

Is it possible to use MTX-90 thyratrons as trigger tubes in a ring counter circuit to build a time keeping device with them?   My inspiration in asking this question is from other member's designs that have been done with XC18 triggers or neon bulbs that use mains for timing pulses across multiple ring counting circuits.  

[threeneurons]

Grahame Marsh made a whole nixie clock using them. Here's his page on it:

http://www.sgitheach.org.uk/nixie3.html

With some "tweeking" (adjusting some part values) it should be doable with any trigger tube. Grahame's page also references the J.B. Dance book, where all kinds of counting techniques are covered.

[William Lee]

Thanks for the reply.  I'm familiar with Grahame's excellent work and some of the Dance circuits.  I just wasn't sure if I am missing anything when considering the MXT-90s for this type of project.  They have a cathode/anode and a control electrode so I thought they would work in a counting circuit, but I wasn't sure if I was missing something more subtle that would make them problematic as a choice.

[Grahame Marsh]

Another book worth finding is "Cold Cathode Tube Circuit Design" by DM Neale which contains design procedures and worked examples. It contains about 100 pages on trigger tubes as well as chapters on stabilisers, reference tubes, stepping tubes (ie dekatrons) and display tubes (ie nixies):

It's here (for a while) : http://www.sgitheach.org.uk/dmneale.pdf

Or Abe Books for <gasp> £91 </gasp>

Grahame

--
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/msg/neonixie-l/-/ij8klWG11OMJ.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Grahame Marsh]

I should have added that the main problem with the XC18 is that they don't work reliably in the dark.  So overnight the clock will sometimes crash.  A fluorescent light or UV lamp keeps it running, but that's not what I really want to do just for a clock.  My trigger tube clock is now just a demonstration clock that I fire up when required. 

This makes the Z70U (or Z700U) more attractive as it has a fourth primer electrode.

http://frank.pocnet.net/sheets/030/z/Z70U.pdf

and an application note

http://frank.pocnet.net/sheets/152/suppinfo/Z70UZ70W.pdf

I bought enough Z700U for a clock from (no connection) but not progressed it yet (having too much fun with CRTs)

michael rotenberg sa...@singlepower.com

But you might build an experimental MXT-90 ring and see how it runs in the dark?

Grahame

On 30/01/2013 19:15, William Lee wrote:

--

[Tidak Ada]

Graham, did yo ever try UV-LED's close to the tubes ? And  if so, what wavelength is the  most efficiënt ?

 

eric

---

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Grahame Marsh
Sent: woensdag 30 januari 2013 21:31
To: neoni...@googlegroups.com
Subject: Re: [neonixie-l] Re: MTX-90 in counting circuit?

[Smiffy]

On Thursday, January 31, 2013 6:33:48 AM UTC+10:30, Grahame Marsh wrote:

...

Or Abe Books for <gasp> £91 </gasp>

Ouch! And that's AbeBooks, too. So thanks for sharing :-)

But this thread is rather timely - just yesterday I went back to trying to figure out how to cascade A101 Dekatrons to form a clock/calendar. Whilst I'd figured a simple way to do it, using a tiny microcontroller between each stage (which also makes it easy to set,) barring the initial timing source, I wanted to stay away from silicon if possible. Would something like an XC18 (now that I've located a source,) be suited to this task? (Sufficiently illuminated, that is.)

[Grahame Marsh]

Eric

I've not tried UV LEDs, but I recall another design using two wire neon bulbs(?) used UV LEDs to keep the rings going.

A lot more people on the list are more knowledgeable of the physics involved, but my understanding is that using the photoelectric effect to prime the tube - knock electrons off the surface to start the ionisation process that then cascades to the tube fired, requires light photons that exceed the work function of the metal.  Therefore the higher the frequency of the light should be better.  But the other question (which I can't answer either!) is the amount of illumination required for the photoelectric effect to be sufficient to reliably prime the tubes.

Now I'll step aside and allow someone else to correct the physics...

Grahame

[threeneurons]

Unfortunately, the only trigger that has the proper signal levels to commutate a dekatron properly is the Ericsson GTE175M:

http://frank.pocnet.net/sheets/022/g/GTE175M.pdf

A dekatron's guides need to see an incoming signal of a very minimum of 60V transition, negative. Its outputs a signal that only goes ~20V, and in the wrong direction. Most triggers need a larger incoming signal. The GTE175M can do the job. All the other ones I've seen can't. In the old days, the interconnecting stage, between dekatrons was usually a vacuum tube triode. The favorite was a 12AT7. Look at figure 4, in this document:

http://frank.pocnet.net/sheets/084/7/7155.pdf

[Smiffy]

On Thursday, January 31, 2013 1:30:50 PM UTC+10:30, threeneurons wrote:

Unfortunately, the only trigger that has the proper signal levels to commutate a dekatron properly is the Ericsson GTE175M...

Ah, that's a shame - on the basis that the XD18 is to be had plentifully, and at a sensible price.

Wonder if it's possible to replace a thermionic triode with a transistor of some description - rather than going the microcontroller route, which I still think seems to be a bit of an overkill, and not really in keeping with the aesthetic I'm looking for.

20V in, 60V down - just an inverter with a gain of 3, isn't it?

 

[threeneurons]

Yes, using transistors will do just fine. And you can use the nixie group's favorite, the very common MPSA42, which can withstand 300V. Here are some of my dekatron circuits:

http://threeneurons.wordpress.com/dekatron-stuff/

http://threeneurons.wordpress.com/dekatron-stuff/variable-dekatron-spinner-kit/

http://threeneurons.wordpress.com/dekatron-stuff/auxiliary-dekatron-stuff/

http://threeneurons.wordpress.com/nixie-power-supply/nixie-clock-kit-6-digit-with-dekatron-pendulum/

[Dekatron42]

You can use a lot of trigger tubes for dekatron coupling, the problem being to bias the trigger electrode correctly - a thing that is much more easy with the GTE175M.

 

There are old East German magazines showing how to use the Z70U and  the Z5823 and similar types. The problem is that they don't fit as easily as a coupling stage like the GTE175M does. They don't fit directly for all types of dekatrons as the output voltage change available across the cathode resistor on the dekatron might be too small to fit the trigger voltage necessary for the trigger tube to get a stable triggering point meaning that you will have to use a potentiometer to trim the bias point for each coupling stage. You will usually have to put a negative voltage on the cathode resistor on the dekatron instead of grounding it to get a high enough voltage to reach the necessary trigger voltage change.

 

The Z70U was used as a coupling stage between ZM1070 dekatrons and the Z5823 was used together with Z562S. Since the ZM1070 and Z562S use almost the same voltages as any other dekatron it should be possible to adjust the circuits for them to work properly with any other dekatron as long as the output voltage across the cathode resistor is enough to reach a stable triggering point with the biasing technique. I've heard that Mullard had a coupling stage with the Z70U for their Z504S (identical to the ZM1070) but I have not seen any circuit diagram for it.

 

There is also the possibility of using for instance the EF80 valve as a coupling stage like ETL (Ericsson) did for their GSA10G - have a look at circuit LK201 on page CT-12 here: http://tubehobby.com/datasheets/DEKA.pdf. The EF80 is cheaper than the 12AT7 and it also handles the cut-off for long times better than the 12AT7, the double triode 5963 would be a better choice for long cut-off periods as it was developed for just that purpose in computers.

 

/Martin

[Tidak Ada]

And GDT125T or GTD125M ? As far as I know they are also developed to manage dekatrons. Unfortunately the GDTx' are not very speedy...

 

eric

---

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of threeneurons
Sent: donderdag 31 januari 2013 4:01
To: neoni...@googlegroups.com
Cc: graham...@googlemail.com

Subject: Re: [neonixie-l] Re: MTX-90 in counting circuit?

--

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/msg/neonixie-l/-/744HLDyK7w4J.

[Tidak Ada]

Problem with thermionic tubes are the vulnerability to a kind of cathode poisoning due to lack of current in 'zero time' and the high power consumption.

 

Best should be 'computer grade tubes' like E90CC and E92CC They have specially coated cathodes

 

eric

---

From: neoni...@googlegroups.com [mailto:neoni...@googlegroups.com] On Behalf Of Smiffy
Sent: donderdag 31 januari 2013 6:38
To: neoni...@googlegroups.com

Subject: Re: [neonixie-l] Re: MTX-90 in counting circuit?

--

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/msg/neonixie-l/-/6tPXBUsloF8J.

[JohnK]

I have a brochure from a British valve maker saying that once they solved the interface resistance problem for computer valves they used the same cathodes in all the valves.

I think that all the old era triodes and double-triodes are ok. I wonder if the current Chinese valves are.

 

John K

[John Rehwinkel]

> I have a brochure from a British valve maker saying that once they solved the interface resistance problem for computer valves they used the same cathodes in all the valves.

That seems unlikely for a large valve maker, but maybe this was a smaller outfit. The truth is that large valve makers added silicon to the cathode sleeve mix to make the cathodes "activate" faster. Faster activation means less time on the activation station, which translates to lower production cost and higher profit. The drawback is that adding silicon was what caused the "sleeping sickness" in tubes held in cutoff for long periods, so the computer grade tubes cost more to make, as they were made without the silicon. However, money could be saved by using old punches, winders, and so forth, as linearity was not an issue (this is why computer grade tubes are generally not a good bet for audio or other analog/linear use).

A smaller outfit, however, may not buy enough cathode sleeve alloy to have two different formulations, and isn't as sensitive to high-volume production, so it could afford to just let all their tubes activate more slowly, and thereby use the the same allow for all their tubes. And naturally, advertise this as an advantage!

- John

[David Forbes]

On 1/31/13 8:30 AM, John Rehwinkel wrote:
>
> The drawback
> is that adding silicon was what caused the "sleeping sickness" in tubes held in cutoff
> for long periods, so the computer grade tubes cost more to make, as they were made
> without the silicon. However, money could be saved by using old punches, winders,
> and so forth, as linearity was not an issue (this is why computer grade tubes are
>generally not a good bet for audio or other analog/linear use).
>

The funny thing is that here are some audio files who think that the
computer tubes are *better* for audio use, so they have driven up the
price of IBM 6211s.


--
David Forbes, Tucson AZ

[Dekatron42]

John, thank you for explaining this! I've been looking for a good explanation for quite some time, I've not put much effort into searching but I've been curious what was the difference.

[JohnK]

When the subject came up on TCA a few years back I pointed to the published
statement... it was generally agreed there that all companies used improved
cathodes. I don't remember whether anyone came up with a black and white
statement like the one I had.
I'll ask again.

A quick look at Kohl [the 1960 edition] reveals that the cathode is
extremely complex. Interface resistance affected all tubes but only some
circuits were particularly sensitive to it. The problems experienced in
wideband scope circuits comes to mind.
As the component production methods and controls improved I think that
better quality materials came cheaper and were able to be used wider. That
effect can be seen in most electronics - eg resistors.

The broad statements about computer tubes and interface resistance aren't
correct.
Quoting from "Electron Tube Life Factors", US Army Signal Corps Development
Lab, Ft Monmouth, 1959.....
" Cathode Interface Resistance.
Cathode interface resistance was measured by Burroughs Corporation on the
tubes of the second life test run only for dynamic and static conditions.
The tubes of the cutoff conditions were life tested during the second run
only and all of them were measured for cathode interface resistance.
Burroughs reported that continual calibration of the McNarry bridge for
measuring interface resistance was made by artificial impedance networks and
by means of stabilized calibrated tubes. Although correlation with other
interface resistance tests were not fully satisfactory, the results were
consistent and reproducible on the equipment used.

The values of interface resistance measured at the end of the 5000-hr life
test are listed in Table 12-10 for the three test conditions in the order of
increasing silicon content in the cathode sleeves. The pattern of growth of
interface resistance is not uniform for all of the tube types. Only for
Types JAN-6AG7 and JAN-6AN5 did interface resistance grow to high values for
low duty factor and smaller values for high duty factor. The cathode sleeves
of these two types had the largest and the smallest amounts of silicon.
Those tubes having intermediate percentages of silicon departed from the
pattern of the high-silicon tubes and in two cases (Types JAN-5687 and
JAN-12AT7) the tubes on cutoff operation had the smallest values of
interface resistance. This is the reverse of what has come to be expected."

The paragraphs following this discuss the spectrographic examination and the
correlation to the actual tubes tested. Some of the data "should be assumed
to have questionable validity".
But a conclusion was still stated..." The evidence of this life test and
spectrographic analysis does not indicate any conclusive pattern of growth
for interface resistance."

It is interesting that the results of the JAN-12AT7 surprised them.
"The pattern of interface resistance growth was unusual in that th etubes on
cutoff life test conditions developed the smallest resistance, 2 ohms,
during the 5000 hr. The tubes life tested under dynamic conditions, which
were equivalent to the cutoff condition in this task, developed 46 ohms of
interface resistance. Surprisingly, the tubes life tested under standard MIL
conditions developed the highest interface resistance. This was 220 ohms at
5000 hr. The reason for this behavior is not readily apparent. The cathode
sleeve material was analyzed by Burroughs Corporation and found to contain a
small percentage of silicon, 0.016 percent. This amount is small enough for
the cathode to be considered to be made of passive material and not to
contribute greatly to the growth of interface resistance."

This brings me back to the complexity issue. The mechanisms were not fully
understood or characterised. For some tube type results the conclusion was
along the lines of eg JAN-7AK7 "The silicon content of the cathode sleeves
is relatively low at 0.015 percent, and this undoubtedly contributes to the
low interface resistance values in all of the tube lots."
It is an interesting 173 page book.

Interface resistance is a problem, to varying degrees, in dynamic circuits.
Manufacturers would have striven to produce the 'perfect' cathode - I want
to see more hard data. Obviously costs have to be considered, but
manufacturers can't afford to be plagued with erratic unexplained failures
and 'modern' production techniques rely on statistical control. The complete
short term failure of the very important proximity fuze program is an
example that did occur. One manufacturer was able to make reliable filaments
[much to the chagrine of the competitors].
However, when that particular single billet of tungsten was exhausted that
manufacturer began to suffer the same failures. The not-so-large piece of
tungsten made an incredibly large number of filaments; the wire was so fine.


John Kaesehagen
Australia.





jk

----- Original Message -----
From: "John Rehwinkel" <jre...@mac.com>
To: <neoni...@googlegroups.com>
Sent: Friday, February 01, 2013 2:00 AM
Subject: Re: [neonixie-l] Re: MTX-90 in counting circuit?