Looking to ID a diode in a 1S1 sampling plug-in


Sean Turner
 

Hi all,

I recently acquired a *very* interesting Type 1S1. It is serial number B 11, and seems like it is very likely to be a prototype. It's in very nice shape physically, but has serious electrical issues. I replaced two extremely crusty filter caps on the + and -19V rails already, but that's the tip of the iceberg. The trace is permanently deflected off the bottom (but occasionally jumps when the mV/cm control is twiddled), and there is no staircase waveform at the horizontal output.

After careful testing, I believe all the GaAs semiconductors are operational. One of the five tunnel diodes might be going bad. All the socketed transistors are fine too. Taking a closer look at the blocking oscillator (which isn't running), I found that D80 and D90 are both shorted. Now, in the manual, these are called out as "ED 2007" germanium diodes. I am striking out on finding that part number anywhere. Does anyone know (1) what this diode is and (2) if there is a good substitute?

Thanks!

Sean


 

Tek P# 152-0071-00 listed in the Common Design Parts Catalog (1982) at the bottom of page 12-4 with vendor number "ED2007 Fmly" the specs are given as:

MAX WRK V: 15
MIN BKN V: 15
MAX AVG mA: 40
MAX I(R) @ VR(R) uA V: 10 @ 6
MAX V(F) @ I(F) V mA: .4 - 5
MAX C(0) pF: 2.5
T(rr) I(F)-I(R) ns * mA: 150 S 20 - .1
CASE STYLE: D0-7

I don't understand all the notation in this table, especially for T(rr}, but that's what it says. This is in the table titled "SIGNAL AND SWITCHING"

-- Jeff Dutky


Tom Lee
 

Hi Jeff,

You need the special Tek decoder ring. That T(rr) line translates to:

Reverse recovery time is 150ns, when the initial (forward) current is 20mA and the diode recovers to 0.1mA.

Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/25/2021 20:10, Jeff Dutky wrote:
Tek P# 152-0071-00 listed in the Common Design Parts Catalog (1982) at the bottom of page 12-4 with vendor number "ED2007 Fmly" the specs are given as:

MAX WRK V: 15
MIN BKN V: 15
MAX AVG mA: 40
MAX I(R) @ VR(R) uA V: 10 @ 6
MAX V(F) @ I(F) V mA: .4 - 5
MAX C(0) pF: 2.5
T(rr) I(F)-I(R) ns * mA: 150 S 20 - .1
CASE STYLE: D0-7

I don't understand all the notation in this table, especially for T(rr}, but that's what it says. This is in the table titled "SIGNAL AND SWITCHING"

-- Jeff Dutky




 

Thanks Tom. I'm still collecting my box tops.

Why anyone would subtract forward current from reverse current had me stumped.


 

(of course I meant to subtract reverse current from forward current)


Tom Lee
 

Here's what "reverse recovery" means, and how it is measured.

Diodes don't transition from "on" to "off" instantaneously. How long it takes can matter a lot. That time depends on how "on" it was to start with, and how you choose to define "off enough". The former is the initial forward current, and the latter is the reverse current that defines "off enough" (recovered). Trr is the time it takes to make that transition.

The two currents aren't being subtracted in the shorthand notation of the table. The "minus sign" is actually a hyphen, here used as a shorthand for an arrow ("the current goes from this value to that value").

Tom


--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/25/2021 20:35, Jeff Dutky wrote:
Thanks Tom. I'm still collecting my box tops.

Why anyone would subtract forward current from reverse current had me stumped.




Sean Turner
 

Ah! this reference is something I did not know existed...saving! Thanks, and thanks to Tom for providing the secret decoder ring...at least it didn't say "drink more Ovaltine". :o) This clears up what it is, but the more pressing question is there a suitable replacement? I'm not going to hold my breath about finding any originals...

Sean

On Thu, Feb 25, 2021 at 08:10 PM, Jeff Dutky wrote:


Tek P# 152-0071-00 listed in the Common Design Parts Catalog (1982) at the
bottom of page 12-4 with vendor number "ED2007 Fmly" the specs are given as:

MAX WRK V: 15
MIN BKN V: 15
MAX AVG mA: 40
MAX I(R) @ VR(R) uA V: 10 @ 6
MAX V(F) @ I(F) V mA: .4 - 5
MAX C(0) pF: 2.5
T(rr) I(F)-I(R) ns * mA: 150 S 20 - .1
CASE STYLE: D0-7

I don't understand all the notation in this table, especially for T(rr}, but
that's what it says. This is in the table titled "SIGNAL AND SWITCHING"

-- Jeff Dutky


Tom Lee
 

Love the Ovaltine reference! That's for decrypting only the most mysterious datasheets. ;)

I haven't yet looked at the schematic to see where those diodes are used but, aside from the lower forward drop owing to germanium-ness, the specs are a decent match to those of a 1N4148/1N914. I'll look at the schematic when I have a few more free nanoseconds than I have at the moment.

Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 2/25/2021 22:36, Sean Turner wrote:
Ah! this reference is something I did not know existed...saving! Thanks, and thanks to Tom for providing the secret decoder ring...at least it didn't say "drink more Ovaltine". :o) This clears up what it is, but the more pressing question is there a suitable replacement? I'm not going to hold my breath about finding any originals...

Sean

On Thu, Feb 25, 2021 at 08:10 PM, Jeff Dutky wrote:

Tek P# 152-0071-00 listed in the Common Design Parts Catalog (1982) at the
bottom of page 12-4 with vendor number "ED2007 Fmly" the specs are given as:

MAX WRK V: 15
MIN BKN V: 15
MAX AVG mA: 40
MAX I(R) @ VR(R) uA V: 10 @ 6
MAX V(F) @ I(F) V mA: .4 - 5
MAX C(0) pF: 2.5
T(rr) I(F)-I(R) ns * mA: 150 S 20 - .1
CASE STYLE: D0-7

I don't understand all the notation in this table, especially for T(rr}, but
that's what it says. This is in the table titled "SIGNAL AND SWITCHING"

-- Jeff Dutky


cmjones01
 

On Fri, Feb 26, 2021 at 4:14 AM Sean Turner <sdturne@q.com> wrote:
I recently acquired a *very* interesting Type 1S1. It is serial number B 11, and seems like it is very likely to be a prototype. It's in very nice shape physically, but has serious electrical issues. I replaced two extremely crusty filter caps on the + and -19V rails already, but that's the tip of the iceberg. The trace is permanently deflected off the bottom (but occasionally jumps when the mV/cm control is twiddled), and there is no staircase waveform at the horizontal output.
I don't know about the diode, but I'm following this thread with
interest. I also have a 1S1 in good physical shape, but it's
Heerenveen-manufactured serial number 700183. It also has the trace
permanently deflected off the bottom. I don't know if the horizontal
output is working because, when I tested it, it turned out that the
horizontal input on my mainframe was faulty. I'll be having another
look at it soon. I have a complete P6032 probe set, with all the
attenuator tips, to go with it.

Chris


 

sean,

I have 4 of the tek 152-0071-00 diodes, I can send them to you in a padded envelope if they will help
restore your 1S1 to life.

just email me off-list. they look like ordinary glass diodes, but are colored blue.

all the best,
walter (walter2 -at- sphere.bc.ca)
sphere reseaarch corp.

--
Walter Shawlee 2
Sphere Research Corp. 3394 Sunnyside Rd.
West Kelowna, BC, V1Z 2V4 CANADA
Phone: +1 (250-769-1834 -:- http://www.sphere.bc.ca
+We're all in one boat, no matter how it looks to you. (WS2)
+All you need is love. (John Lennon)
+But, that doesn't mean other things don't come in handy. (WS2)
+Nature is trying very hard to make us succeed, but nature does not depend on us.
We are not the only experiment. (R. Buckminster Fuller)


Dave Wise
 

?150ns is slow compared to a Silicon signal diode.


I looked at the schematic. AFAICT Tek did not pull any special tricks capitalizing on this diode having a specific Trr.


On the other hand, Tek had plenty of good Si diodes, like 1N3605/1N4152 which appears in many other places in the 1S1. They chose Germanium for D80 on purpose.?


D80 ensures that Q80 (selected 2N964) stays on a while once it goes on. (The duration is mostly controlled by T80.) This Blocking Oscillator, which drives Snapoff Diode D87, is a high-strung, high-performance circuit. Q80 is selected for avalanche characteristics in a special jig. I suspect that the extra Vf of Si might reduce its output enough to fail to meet spec.


Ge D90 and Si D92 ensure that Memory Gate Driver Q94's base sits slightly positive (biased off) between pulses. That's important for a Germanium transistor. D90's Vf at 6mA must be less than D92's Vf at 0.5mA so D90 must be Germanium or Schottky. Or teepee an extra diode in series with D92.


Dave Wise


In my own 1S1, sampling pulse quality is strongly affected by Q80's lead length. I only meet spec if it's crammed into the socket as tight as possible. I'm tempted to grind down the socket.


________________________________
From: TekScopes@groups.io <TekScopes@groups.io> on behalf of Tom Lee via groups.io <tomlee=ee.stanford.edu@groups.io>
Sent: Thursday, February 25, 2021 8:44 PM
To: TekScopes@groups.io
Subject: Re: [TekScopes] Looking to ID a diode in a 1S1 sampling plug-in

Here's what "reverse recovery" means, and how it is measured.

Diodes don't transition from "on" to "off" instantaneously. How long it
takes can matter a lot. That time depends on how "on" it was to start
with, and how you choose to define "off enough". The former is the
initial forward current, and the latter is the reverse current that
defines "off enough" (recovered). Trr is the time it takes to make that
transition.

The two currents aren't being subtracted in the shorthand notation of
the table. The "minus sign" is actually a hyphen, here used as a
shorthand for an arrow ("the current goes from this value to that value").

Tom


--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu<http://www-smirc.stanford.edu>

On 2/25/2021 20:35, Jeff Dutky wrote:
Thanks Tom. I'm still collecting my box tops.

Why anyone would subtract forward current from reverse current had me stumped.





Sean Turner
 

I uploaded a small album of this plug-in: https://groups.io/g/TekScopes/album?id=261202

On the third picture, another curiousity is the tunnel diodes (circled in red). These do not look like any Tek tunnel diodes I've seen, and they show signs of soldering that is decidedly not up to the standards of factory Tek soldering. Whether this is a symptom of being a prototype (serial number 11 after all) or later work, no idea. D430 (the uppermost one circled) also has a weird curve with only a very tiny negative resistance region (though it is 10 mA), and the manual calls for a TD253 which is a metal package IIRC.

Sean


Ed Breya
 

Sean, I don't know about the other diodes discussed earlier, but I have seen those tiny glob ones with the dot (in the picture) before, in Tek and other gear. They could be just a different version (stripline) package for regular TDs, or they could be "back-diodes," which are a close relative. As I recall, back-diodes are even more heavily doped than regular TDs, so they are already in forward bias at zero volts, and the interesting part of their behavior exists around the zero-crossing, and includes that small region that seems kind of like a crappy TD. Your mention of the weird curve reminded me. So, don't assume they're bad TDs just yet, until you study it some more. Somewhere in the circuit description or schematics it may become more clear. I don't recall what the usual TD part designations are, but somewhere in tekwiki, there's a nice listing and summary of many TD numbers from the old days. Look for that - it may explain what TD253 is supposed to be. Also, you can look up back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase development stages were used back then, as in more modern times, then "B11" should be very close to what went into production. The poor solder joints you mentioned, around those tiny diodes, may result from a couple of considerations. First, those things are very delicate, so anyone soldering them in or out would know they have to work fast - very fast - to get that iron tip out of there ASAP, even if the joint looks crappy, as long as it worked. Second, being a prototype, these diodes may have been swapped out or had type changes many times during development, and the same soldering issues apply. This could also mean that those diodes are not necessarily the same part number that was ultimately used in production. But, you'd think they would be close at least. Good luck in figuring it out.

Ed


Sean Turner
 

Ed,

Make no mistake, I am not wanting to solder on tunnel diodes unless I absolutely must.

Thanks for the insights!

Sean

On Sat, Feb 27, 2021 at 11:37 AM, Ed Breya wrote:


Sean, I don't know about the other diodes discussed earlier, but I have seen
those tiny glob ones with the dot (in the picture) before, in Tek and other
gear. They could be just a different version (stripline) package for regular
TDs, or they could be "back-diodes," which are a close relative. As I recall,
back-diodes are even more heavily doped than regular TDs, so they are already
in forward bias at zero volts, and the interesting part of their behavior
exists around the zero-crossing, and includes that small region that seems
kind of like a crappy TD. Your mention of the weird curve reminded me. So,
don't assume they're bad TDs just yet, until you study it some more. Somewhere
in the circuit description or schematics it may become more clear. I don't
recall what the usual TD part designations are, but somewhere in tekwiki,
there's a nice listing and summary of many TD numbers from the old days. Look
for that - it may explain what TD253 is supposed to be. Also, you can look up
back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase development
stages were used back then, as in more modern times, then "B11" should be very
close to what went into production. The poor solder joints you mentioned,
around those tiny diodes, may result from a couple of considerations. First,
those things are very delicate, so anyone soldering them in or out would know
they have to work fast - very fast - to get that iron tip out of there ASAP,
even if the joint looks crappy, as long as it worked. Second, being a
prototype, these diodes may have been swapped out or had type changes many
times during development, and the same soldering issues apply. This could also
mean that those diodes are not necessarily the same part number that was
ultimately used in production. But, you'd think they would be close at least.
Good luck in figuring it out.

Ed


cmjones01
 

I've had a look in my 1S1 and the tunnel diodes seem to be a mix of tiny
beads like the ones in the OP's photo, and the gold "top-hat" style package
I've always associated with Tek tunnel diodes.

One explanation for the trace being stuck off the bottom of the screen is
also clear. The Nuvistor and a nearby transistor are entirely missing, so
there will be no staircase being generated. Typically, the Nuvistor isn't
the same as the ones in any of my other plugins so I can't even borrow one
to try. Time to go shopping for an 8056...

Chris

On Sun, 28 Feb 2021, 02:49 Sean Turner, <sdturne@q.com> wrote:

Ed,

Make no mistake, I am not wanting to solder on tunnel diodes unless I
absolutely must.

Thanks for the insights!

Sean

On Sat, Feb 27, 2021 at 11:37 AM, Ed Breya wrote:


Sean, I don't know about the other diodes discussed earlier, but I have
seen
those tiny glob ones with the dot (in the picture) before, in Tek and
other
gear. They could be just a different version (stripline) package for
regular
TDs, or they could be "back-diodes," which are a close relative. As I
recall,
back-diodes are even more heavily doped than regular TDs, so they are
already
in forward bias at zero volts, and the interesting part of their behavior
exists around the zero-crossing, and includes that small region that
seems
kind of like a crappy TD. Your mention of the weird curve reminded me.
So,
don't assume they're bad TDs just yet, until you study it some more.
Somewhere
in the circuit description or schematics it may become more clear. I
don't
recall what the usual TD part designations are, but somewhere in tekwiki,
there's a nice listing and summary of many TD numbers from the old days.
Look
for that - it may explain what TD253 is supposed to be. Also, you can
look up
back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase
development
stages were used back then, as in more modern times, then "B11" should
be very
close to what went into production. The poor solder joints you mentioned,
around those tiny diodes, may result from a couple of considerations.
First,
those things are very delicate, so anyone soldering them in or out would
know
they have to work fast - very fast - to get that iron tip out of there
ASAP,
even if the joint looks crappy, as long as it worked. Second, being a
prototype, these diodes may have been swapped out or had type changes
many
times during development, and the same soldering issues apply. This
could also
mean that those diodes are not necessarily the same part number that was
ultimately used in production. But, you'd think they would be close at
least.
Good luck in figuring it out.

Ed





Tom Lee
 

The 8056 is the most transistor-like of all Nuvistors. If you are not insistent on preserving original factory innards, you can replace that Nuvistor with a JFET. You can even plug it directly into the Nuvistor socket, so no mechanical mods are necessary. For details on that, see the tekwiki page on the 422. The MPF102 suggested for the latter is probably not a bad choice here, either. The gm is a little on the low side, but that shouldn't matter much here at all. You will have to readjust the staircase DC level trimmer, but that's par for the course even when replacing a Nuvistor with another Nuvistor.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 3/1/2021 13:14, cmjones01 wrote:
I've had a look in my 1S1 and the tunnel diodes seem to be a mix of tiny
beads like the ones in the OP's photo, and the gold "top-hat" style package
I've always associated with Tek tunnel diodes.

One explanation for the trace being stuck off the bottom of the screen is
also clear. The Nuvistor and a nearby transistor are entirely missing, so
there will be no staircase being generated. Typically, the Nuvistor isn't
the same as the ones in any of my other plugins so I can't even borrow one
to try. Time to go shopping for an 8056...

Chris

On Sun, 28 Feb 2021, 02:49 Sean Turner, <sdturne@q.com> wrote:

Ed,

Make no mistake, I am not wanting to solder on tunnel diodes unless I
absolutely must.

Thanks for the insights!

Sean

On Sat, Feb 27, 2021 at 11:37 AM, Ed Breya wrote:

Sean, I don't know about the other diodes discussed earlier, but I have
seen
those tiny glob ones with the dot (in the picture) before, in Tek and
other
gear. They could be just a different version (stripline) package for
regular
TDs, or they could be "back-diodes," which are a close relative. As I
recall,
back-diodes are even more heavily doped than regular TDs, so they are
already
in forward bias at zero volts, and the interesting part of their behavior
exists around the zero-crossing, and includes that small region that
seems
kind of like a crappy TD. Your mention of the weird curve reminded me.
So,
don't assume they're bad TDs just yet, until you study it some more.
Somewhere
in the circuit description or schematics it may become more clear. I
don't
recall what the usual TD part designations are, but somewhere in tekwiki,
there's a nice listing and summary of many TD numbers from the old days.
Look
for that - it may explain what TD253 is supposed to be. Also, you can
look up
back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase
development
stages were used back then, as in more modern times, then "B11" should
be very
close to what went into production. The poor solder joints you mentioned,
around those tiny diodes, may result from a couple of considerations.
First,
those things are very delicate, so anyone soldering them in or out would
know
they have to work fast - very fast - to get that iron tip out of there
ASAP,
even if the joint looks crappy, as long as it worked. Second, being a
prototype, these diodes may have been swapped out or had type changes
many
times during development, and the same soldering issues apply. This
could also
mean that those diodes are not necessarily the same part number that was
ultimately used in production. But, you'd think they would be close at
least.
Good luck in figuring it out.

Ed





cmjones01
 

Ah, thank you, I'll look in to replacing it with a FET and will see what I
have in stock. I presume I'll have to do some modifications to keep the
heater chain current right: it looks like that interacts with the +19V
supply.

While a FET may be practical, there's part of me that really likes the idea
of a plugin containing both a load of tunnel diodes and a Nuvistor, for
maximum use of obscure devices!

Chris

On Mon, 1 Mar 2021, 22:34 Tom Lee, <tomlee@ee.stanford.edu> wrote:

The 8056 is the most transistor-like of all Nuvistors. If you are not
insistent on preserving original factory innards, you can replace that
Nuvistor with a JFET. You can even plug it directly into the Nuvistor
socket, so no mechanical mods are necessary. For details on that, see
the tekwiki page on the 422. The MPF102 suggested for the latter is
probably not a bad choice here, either. The gm is a little on the low
side, but that shouldn't matter much here at all. You will have to
readjust the staircase DC level trimmer, but that's par for the course
even when replacing a Nuvistor with another Nuvistor.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 3/1/2021 13:14, cmjones01 wrote:
I've had a look in my 1S1 and the tunnel diodes seem to be a mix of tiny
beads like the ones in the OP's photo, and the gold "top-hat" style
package
I've always associated with Tek tunnel diodes.

One explanation for the trace being stuck off the bottom of the screen is
also clear. The Nuvistor and a nearby transistor are entirely missing, so
there will be no staircase being generated. Typically, the Nuvistor isn't
the same as the ones in any of my other plugins so I can't even borrow
one
to try. Time to go shopping for an 8056...

Chris

On Sun, 28 Feb 2021, 02:49 Sean Turner, <sdturne@q.com> wrote:

Ed,

Make no mistake, I am not wanting to solder on tunnel diodes unless I
absolutely must.

Thanks for the insights!

Sean

On Sat, Feb 27, 2021 at 11:37 AM, Ed Breya wrote:

Sean, I don't know about the other diodes discussed earlier, but I have
seen
those tiny glob ones with the dot (in the picture) before, in Tek and
other
gear. They could be just a different version (stripline) package for
regular
TDs, or they could be "back-diodes," which are a close relative. As I
recall,
back-diodes are even more heavily doped than regular TDs, so they are
already
in forward bias at zero volts, and the interesting part of their
behavior
exists around the zero-crossing, and includes that small region that
seems
kind of like a crappy TD. Your mention of the weird curve reminded me.
So,
don't assume they're bad TDs just yet, until you study it some more.
Somewhere
in the circuit description or schematics it may become more clear. I
don't
recall what the usual TD part designations are, but somewhere in
tekwiki,
there's a nice listing and summary of many TD numbers from the old
days.
Look
for that - it may explain what TD253 is supposed to be. Also, you can
look up
back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase
development
stages were used back then, as in more modern times, then "B11" should
be very
close to what went into production. The poor solder joints you
mentioned,
around those tiny diodes, may result from a couple of considerations.
First,
those things are very delicate, so anyone soldering them in or out
would
know
they have to work fast - very fast - to get that iron tip out of there
ASAP,
even if the joint looks crappy, as long as it worked. Second, being a
prototype, these diodes may have been swapped out or had type changes
many
times during development, and the same soldering issues apply. This
could also
mean that those diodes are not necessarily the same part number that
was
ultimately used in production. But, you'd think they would be close at
least.
Good luck in figuring it out.

Ed











Tom Lee
 

Yes, you might have to kludge in an appropriate resistor (around 39 ohms, 1-2W) across the heater terminals. Good catch!

JFETs themselves are rapidly becoming obscure devices, but if you want to cram in some other weird things, I suggest a UJT, which has always had my vote for the most unnecessary semiconductor device ever commercialized. Might as well throw in a Shockley diode, and a lump of galena with phosphor bronze catwhisker, just for good measure. Adjust for best tone.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 3/1/2021 13:44, cmjones01 wrote:
Ah, thank you, I'll look in to replacing it with a FET and will see what I
have in stock. I presume I'll have to do some modifications to keep the
heater chain current right: it looks like that interacts with the +19V
supply.

While a FET may be practical, there's part of me that really likes the idea
of a plugin containing both a load of tunnel diodes and a Nuvistor, for
maximum use of obscure devices!

Chris

On Mon, 1 Mar 2021, 22:34 Tom Lee, <tomlee@ee.stanford.edu> wrote:

The 8056 is the most transistor-like of all Nuvistors. If you are not
insistent on preserving original factory innards, you can replace that
Nuvistor with a JFET. You can even plug it directly into the Nuvistor
socket, so no mechanical mods are necessary. For details on that, see
the tekwiki page on the 422. The MPF102 suggested for the latter is
probably not a bad choice here, either. The gm is a little on the low
side, but that shouldn't matter much here at all. You will have to
readjust the staircase DC level trimmer, but that's par for the course
even when replacing a Nuvistor with another Nuvistor.

-- Cheers,
Tom

--
Prof. Thomas H. Lee
Allen Ctr., Rm. 205
350 Jane Stanford Way
Stanford University
Stanford, CA 94305-4070
http://www-smirc.stanford.edu

On 3/1/2021 13:14, cmjones01 wrote:
I've had a look in my 1S1 and the tunnel diodes seem to be a mix of tiny
beads like the ones in the OP's photo, and the gold "top-hat" style
package
I've always associated with Tek tunnel diodes.

One explanation for the trace being stuck off the bottom of the screen is
also clear. The Nuvistor and a nearby transistor are entirely missing, so
there will be no staircase being generated. Typically, the Nuvistor isn't
the same as the ones in any of my other plugins so I can't even borrow
one
to try. Time to go shopping for an 8056...

Chris

On Sun, 28 Feb 2021, 02:49 Sean Turner, <sdturne@q.com> wrote:

Ed,

Make no mistake, I am not wanting to solder on tunnel diodes unless I
absolutely must.

Thanks for the insights!

Sean

On Sat, Feb 27, 2021 at 11:37 AM, Ed Breya wrote:

Sean, I don't know about the other diodes discussed earlier, but I have
seen
those tiny glob ones with the dot (in the picture) before, in Tek and
other
gear. They could be just a different version (stripline) package for
regular
TDs, or they could be "back-diodes," which are a close relative. As I
recall,
back-diodes are even more heavily doped than regular TDs, so they are
already
in forward bias at zero volts, and the interesting part of their
behavior
exists around the zero-crossing, and includes that small region that
seems
kind of like a crappy TD. Your mention of the weird curve reminded me.
So,
don't assume they're bad TDs just yet, until you study it some more.
Somewhere
in the circuit description or schematics it may become more clear. I
don't
recall what the usual TD part designations are, but somewhere in
tekwiki,
there's a nice listing and summary of many TD numbers from the old
days.
Look
for that - it may explain what TD253 is supposed to be. Also, you can
look up
back-diodes, and see if the description matches the curve you saw.

You are right that this unit is a prototype. If the same phase
development
stages were used back then, as in more modern times, then "B11" should
be very
close to what went into production. The poor solder joints you
mentioned,
around those tiny diodes, may result from a couple of considerations.
First,
those things are very delicate, so anyone soldering them in or out
would
know
they have to work fast - very fast - to get that iron tip out of there
ASAP,
even if the joint looks crappy, as long as it worked. Second, being a
prototype, these diodes may have been swapped out or had type changes
many
times during development, and the same soldering issues apply. This
could also
mean that those diodes are not necessarily the same part number that
was
ultimately used in production. But, you'd think they would be close at
least.
Good luck in figuring it out.

Ed









snapdiode
 

Needs a coherer.


 

Q999 155-9999-99 Coherer *Selected by Tek

:)