A solid state replacement for the lowly 2D21 Thyratron tube...(sorry for the On-Topic post!)...

[John Robertson]

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John\'s Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
\"Old pinballers never die, they just flip out.\"

 

[bitrex]

On 4/21/2021 1:28 PM, John Robertson wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

What happens to them as they age? is that with respect to the unused NOS
types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in service
than solid state devices, at least. If you blast a relaxation oscillator
made with a 2D21 with cold spray it barely drifts.

 

[bitrex]

On 4/21/2021 1:28 PM, John Robertson wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

What happens to them as they age? is that with respect to the unused NOS
types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in service
than solid state devices, at least. If you blast a relaxation oscillator
made with a 2D21 with cold spray it barely drifts.

 

[Peter W.]

I may have a few of either floating about in my junk boxes. If so, I would be glad to send all that I have to you as I have no discernable use for them. I seem to remember those numbers,

Gratis, of course. I will let you know.

Peter Wieck
Melrose Park, PA

 

[Peter W.]

I may have a few of either floating about in my junk boxes. If so, I would be glad to send all that I have to you as I have no discernable use for them. I seem to remember those numbers,

Gratis, of course. I will let you know.

Peter Wieck
Melrose Park, PA

 

[John Robertson]

On 2021/04/21 10:48 a.m., bitrex wrote:

On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused NOS
types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in service
than solid state devices, at least. If you blast a relaxation oscillator
made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

--
(Please post followups or tech inquiries to the USENET newsgroup)
John\'s Jukes Ltd.
MOVED to #7 - 3979 Marine Way, Burnaby, BC, Canada V5J 5E3
(604)872-5757 (Pinballs, Jukes, Video Games)
www.flippers.com
\"Old pinballers never die, they just flip out.\"

 

[bitrex]

On 4/21/2021 1:53 PM, John Robertson wrote:

On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

Maybe the hydrogen is out-gassing thru the glass over time?

 

[bitrex]

On 4/21/2021 1:53 PM, John Robertson wrote:

On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

The tube tester can test for conductance from thermionic emission but
IDK if it can test for the conductance when the tube is operating in the
ionization-cascade mode.

 

[bitrex]

On 4/21/2021 1:53 PM, John Robertson wrote:

On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

The tube tester can test for conductance from thermionic emission but
IDK if it can test for the conductance when the tube is operating in the
ionization-cascade mode.

 

[John Larkin]

On Wed, 21 Apr 2021 10:28:02 -0700, John Robertson <spam@flippers.com>
wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

Got a schematic? It might be easy.

 

[John Larkin]

On Wed, 21 Apr 2021 10:28:02 -0700, John Robertson <spam@flippers.com>
wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

Got a schematic? It might be easy.

 

[John Larkin]

On Wed, 21 Apr 2021 14:02:41 -0400, bitrex <user@example.net> wrote:

On 4/21/2021 1:53 PM, John Robertson wrote:
On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#


The tube tester can test for conductance from thermionic emission but
IDK if it can test for the conductance when the tube is operating in the
ionization-cascade mode.

A tube tester may not apply enough voltage or current capability to
fire the gas, so would miss a problem with ionization.

What\'s the Gm of a thyratron anyhow?

One could build a simple thyratron tester.

 

[John Larkin]

On Wed, 21 Apr 2021 14:02:41 -0400, bitrex <user@example.net> wrote:

On 4/21/2021 1:53 PM, John Robertson wrote:
On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#


The tube tester can test for conductance from thermionic emission but
IDK if it can test for the conductance when the tube is operating in the
ionization-cascade mode.

A tube tester may not apply enough voltage or current capability to
fire the gas, so would miss a problem with ionization.

What\'s the Gm of a thyratron anyhow?

One could build a simple thyratron tester.

 

[bitrex]

On 4/21/2021 2:14 PM, John Larkin wrote:

On Wed, 21 Apr 2021 10:28:02 -0700, John Robertson <spam@flippers.com
wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

Got a schematic? It might be easy.

The 1954 gear was probably designed to run on old-timey line voltage of
~110 AC. Dialing back the heater voltage on the thyratrons to compensate
for higher modern AC supply voltage might extend their life.

 

[bitrex]

On 4/21/2021 2:14 PM, John Larkin wrote:

On Wed, 21 Apr 2021 10:28:02 -0700, John Robertson <spam@flippers.com
wrote:

In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to work
on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s used)
on their very first control center for their Tormat (200 x magnetic
cores) memory and dropped it like a hot potato. There were numerous
service bulletins culminating in the factory providing a low cost
replacement for the control centers that used 2D21s (three of these) and
the replacement used a single 2050 tube and was a very reliable design.
Not everyone took advantage of the program and all these replacement
were used up long, long ago.

John :-#)#

Got a schematic? It might be easy.

The 1954 gear was probably designed to run on old-timey line voltage of
~110 AC. Dialing back the heater voltage on the thyratrons to compensate
for higher modern AC supply voltage might extend their life.

 

[bitrex]

On 4/21/2021 1:53 PM, John Robertson wrote:

On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

Hydrogen diffusion through thin glass within tube-like operating
temperatures, where the gas is in thermal equilibrium with the glass and
the mean free path is >> glass thickness is dominated by e^(-1/T), so if
hydrogen outgassing is causing the erratic behavior whether the tube
envelope temperature is run before the \"knee\" or after may have a lot to
do with the expected service life.

 

[bitrex]

On 4/21/2021 1:53 PM, John Robertson wrote:

On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#

Hydrogen diffusion through thin glass within tube-like operating
temperatures, where the gas is in thermal equilibrium with the glass and
the mean free path is >> glass thickness is dominated by e^(-1/T), so if
hydrogen outgassing is causing the erratic behavior whether the tube
envelope temperature is run before the \"knee\" or after may have a lot to
do with the expected service life.

 

[bitrex]

On 4/21/2021 2:50 PM, bitrex wrote:

On 4/21/2021 1:53 PM, John Robertson wrote:
On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several
sources, but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits
of electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a
low cost replacement for the control centers that used 2D21s (three
of these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and
sometimes they just ignore the control signal. Yet these tubes test OK
on my mutual conductance tube checker (of course).

John :-#)#


Hydrogen diffusion through thin glass within tube-like operating
temperatures, where the gas is in thermal equilibrium with the glass and
the mean free path is >> glass thickness

I may have the pac-mans backwards there don\'t quote me on that.

 

[bitrex]

On 4/21/2021 2:50 PM, bitrex wrote:

On 4/21/2021 1:53 PM, John Robertson wrote:
On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several
sources, but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits
of electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a
low cost replacement for the control centers that used 2D21s (three
of these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and
sometimes they just ignore the control signal. Yet these tubes test OK
on my mutual conductance tube checker (of course).

John :-#)#


Hydrogen diffusion through thin glass within tube-like operating
temperatures, where the gas is in thermal equilibrium with the glass and
the mean free path is >> glass thickness

I may have the pac-mans backwards there don\'t quote me on that.

 

[bitrex]

On 4/21/2021 2:18 PM, John Larkin wrote:

On Wed, 21 Apr 2021 14:02:41 -0400, bitrex <user@example.net> wrote:

On 4/21/2021 1:53 PM, John Robertson wrote:
On 2021/04/21 10:48 a.m., bitrex wrote:
On 4/21/2021 1:28 PM, John Robertson wrote:
In my field of arcade repairs we run into older jukeboxes - one
particular model which uses the 2D21 tube and was from about 1954.

I have the spec sheet here:

https://www.flippers.com/pdfs/2D21_Thyratron_tube.pdf

It would be fun to make a solid state replacement for it as nothing
seems to exist. These tubes are still available from several sources,
but they get wonky with age.

Being able to simply wire in a TRIAC or SCR (with appropriate bits of
electronic glue) to remove this tube from the suspect list when
servicing this classic Seeburg V200 jukebox would simplify future
servicing. I always like to make things easier for the next tech to
work on machines that passed through my shop...

I should point out that Seeburg only used these 2D21s (three 2D21s
used) on their very first control center for their Tormat (200 x
magnetic cores) memory and dropped it like a hot potato. There were
numerous service bulletins culminating in the factory providing a low
cost replacement for the control centers that used 2D21s (three of
these) and the replacement used a single 2050 tube and was a very
reliable design. Not everyone took advantage of the program and all
these replacement were used up long, long ago.

John :-#)#


What happens to them as they age? is that with respect to the unused
NOS types?

I was under the impression that small-signal gas thyratrons were very
reliable in service. They\'re much less temperature sensitive in
service than solid state devices, at least. If you blast a relaxation
oscillator made with a 2D21 with cold spray it barely drifts.

The 2D21 tubes become erratic, sometimes they respond, sometimes they
are slow (the solenoid grumbles as it sluggishly engages), and sometimes
they just ignore the control signal. Yet these tubes test OK on my
mutual conductance tube checker (of course).

John :-#)#


The tube tester can test for conductance from thermionic emission but
IDK if it can test for the conductance when the tube is operating in the
ionization-cascade mode.

A tube tester may not apply enough voltage or current capability to
fire the gas, so would miss a problem with ionization.

What\'s the Gm of a thyratron anyhow?

Thyratron anode conductance is grid current-controlled when in the
ionization mode, you can read off the parameters from fig 2 of the 2D21
datasheet. For grid 1 voltages between about -1 and -5 volts and Ib
between 25 and 200 mA it\'s vaguely linear at about 50mA per 1 mA.

One could build a simple thyratron tester.