SC
Stewart Cobb
Sun, Jan 12, 2025 5:42 AM
I have revived an Austron 2110 disciplined oscillator. This consists of an
Austron 1150 high-end double oven OCXO and a computer to lock it to an
incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
tuning adjustment, which the manual says to use in this case, but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
with a period of just under 8 seconds and an amplitude of about 0.42 ns
p-p. The ADEV measurement shows corresponding ripples in an otherwise
stable trace.
My interpretation of this data is that one of the two oven controllers has
gone wonky and increased its gain to the point that it's now unstable. This
might well be due to the failure of an electrolytic capacitor in an
integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be returned to
the factory for repair. After 40-odd years and 5 or 6 changes of corporate
control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
MicroChip), factory repair is no longer possible, so I need advice from the
group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And I
have the pinout from the 2110 manual. In places, the documentation refers
to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos of a
teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
any other relevant comments?
Cheers!
--Stu
I have revived an Austron 2110 disciplined oscillator. This consists of an
Austron 1150 high-end double oven OCXO and a computer to lock it to an
incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
tuning adjustment, which the manual says to use in this case, but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
with a period of just under 8 seconds and an amplitude of about 0.42 ns
p-p. The ADEV measurement shows corresponding ripples in an otherwise
stable trace.
My interpretation of this data is that one of the two oven controllers has
gone wonky and increased its gain to the point that it's now unstable. This
might well be due to the failure of an electrolytic capacitor in an
integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be returned to
the factory for repair. After 40-odd years and 5 or 6 changes of corporate
control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
MicroChip), factory repair is no longer possible, so I need advice from the
group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And I
have the pinout from the 2110 manual. In places, the documentation refers
to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos of a
teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
any other relevant comments?
Cheers!
--Stu
TV
Tom Van Baak
Sun, Jan 12, 2025 10:36 AM
its phase traces out a beautiful sine wave with a period of just
under 8 seconds
First, in TimeLab use the 'n' command to apply a notch filter. Screen
capture the result. Also compare the ADEV before and after. The center
frequency and Q may hint of the origin of the ~8 second modulation. This
won't take more than a minute of your time since you already have the
data. Hint: keep your original .TIM file; do not save over it with the
filtered version.
Then, a couple of ideas for you, pick one:
-
Since your OCXO is outside the locking range I wonder if you're
simply seeing a discipline loop trying and failing repeatedly. I've seen
other standards that go into a similar sweep mode while trying to lock.
Two ways to check. If you have a good DMM monitor the EFC voltage for 10
or 20 seconds and see if there's a waveform. Or, just ground the EFC pin
and see if the 8 second modulation goes away. Check the board /
schematic to see if this is safe, i.e., you may not want to short a DAC
output to ground.
-
Simply remove the external reference frequency so that the 2110 is
free-running. See if the sweeping continues or goes away.
-
Pull the OCXO out of the instrument and run it stand-alone from a
bench supply. If the cycling is gone then the fault is not the OCXO but
instead something in the 2110.
I still have my Austron 2110 on the bench from last we talked (about the
power supply), so if there's anything I can do to help let me know. But
first try one of the above simple diagnostics.
/tvb
On 1/11/2025 9:42 PM, Stewart Cobb via time-nuts wrote:
I have revived an Austron 2110 disciplined oscillator. This consists of an
Austron 1150 high-end double oven OCXO and a computer to lock it to an
incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
tuning adjustment, which the manual says to use in this case, but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
with a period of just under 8 seconds and an amplitude of about 0.42 ns
p-p. The ADEV measurement shows corresponding ripples in an otherwise
stable trace.
My interpretation of this data is that one of the two oven controllers has
gone wonky and increased its gain to the point that it's now unstable. This
might well be due to the failure of an electrolytic capacitor in an
integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be returned to
the factory for repair. After 40-odd years and 5 or 6 changes of corporate
control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
MicroChip), factory repair is no longer possible, so I need advice from the
group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And I
have the pinout from the 2110 manual. In places, the documentation refers
to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos of a
teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
any other relevant comments?
Cheers!
--Stu@lists.febo.com
Stu,
> its phase traces out a beautiful sine wave with a period of just
under 8 seconds
First, in TimeLab use the 'n' command to apply a notch filter. Screen
capture the result. Also compare the ADEV before and after. The center
frequency and Q may hint of the origin of the ~8 second modulation. This
won't take more than a minute of your time since you already have the
data. Hint: keep your original .TIM file; do not save over it with the
filtered version.
Then, a couple of ideas for you, pick one:
1) Since your OCXO is outside the locking range I wonder if you're
simply seeing a discipline loop trying and failing repeatedly. I've seen
other standards that go into a similar sweep mode while trying to lock.
Two ways to check. If you have a good DMM monitor the EFC voltage for 10
or 20 seconds and see if there's a waveform. Or, just ground the EFC pin
and see if the 8 second modulation goes away. Check the board /
schematic to see if this is safe, i.e., you may not want to short a DAC
output to ground.
2) Simply remove the external reference frequency so that the 2110 is
free-running. See if the sweeping continues or goes away.
3) Pull the OCXO out of the instrument and run it stand-alone from a
bench supply. If the cycling is gone then the fault is not the OCXO but
instead something in the 2110.
I still have my Austron 2110 on the bench from last we talked (about the
power supply), so if there's anything I can do to help let me know. But
first try one of the above simple diagnostics.
/tvb
On 1/11/2025 9:42 PM, Stewart Cobb via time-nuts wrote:
> I have revived an Austron 2110 disciplined oscillator. This consists of an
> Austron 1150 high-end double oven OCXO and a computer to lock it to an
> incoming reference frequency and do some statistics and drift prediction.
> It mostly works, but --
>
> After a week of warmup, the OCXO is still about 25 ppb low, which is
> outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
> tuning adjustment, which the manual says to use in this case, but --
>
> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
> TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
> with a period of just under 8 seconds and an amplitude of about 0.42 ns
> p-p. The ADEV measurement shows corresponding ripples in an otherwise
> stable trace.
>
> My interpretation of this data is that one of the two oven controllers has
> gone wonky and increased its gain to the point that it's now unstable. This
> might well be due to the failure of an electrolytic capacitor in an
> integrator loop.
>
> The manual refers to the 1150 as a "sealed unit" which must be returned to
> the factory for repair. After 40-odd years and 5 or 6 changes of corporate
> control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
> MicroChip), factory repair is no longer possible, so I need advice from the
> group ...
>
> Is oven controller failure a reasonable diagnosis?
>
> Does anyone have documentation on the 1150 oscillator? (There is a
> datasheet on KO4BB, but it describes a slightly different version. And I
> have the pinout from the 2110 manual. In places, the documentation refers
> to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
> might also be useful. But is there anything else?)
>
> Has anyone repaired a 1150, or even taken one apart? Are there photos of a
> teardown anywhere?
>
> Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
> any other relevant comments?
>
> Cheers!
> --Stu@lists.febo.com
>
BC
Bob Camp
Sun, Jan 12, 2025 1:29 PM
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test
methods improved.
The result is that there are likely many schematics of the internals of a 1150
OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed
the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic
cap. It would use something with much better leakage performance. That’s not to
say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting
the parts in the vicinity of what you just replaced. Even with the schematic, you
likely would not have that info.
One note: this all assumes the 1150 wanders around when removed from the
device and powered with bench gear. If that step hasn’t been done yet, it’s on the
list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping
for that :) ) or some sort of alternate unit cobbled into the device.
Bob
On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts time-nuts@lists.febo.com wrote:
I have revived an Austron 2110 disciplined oscillator. This consists of an
Austron 1150 high-end double oven OCXO and a computer to lock it to an
incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
tuning adjustment, which the manual says to use in this case, but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
with a period of just under 8 seconds and an amplitude of about 0.42 ns
p-p. The ADEV measurement shows corresponding ripples in an otherwise
stable trace.
My interpretation of this data is that one of the two oven controllers has
gone wonky and increased its gain to the point that it's now unstable. This
might well be due to the failure of an electrolytic capacitor in an
integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be returned to
the factory for repair. After 40-odd years and 5 or 6 changes of corporate
control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
MicroChip), factory repair is no longer possible, so I need advice from the
group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And I
have the pinout from the 2110 manual. In places, the documentation refers
to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos of a
teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
any other relevant comments?
Cheers!
--Stu
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test
methods improved.
The result is that there are likely *many* schematics of the internals of a 1150
OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed
the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic
cap. It would use something with much better leakage performance. That’s not to
say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting
the parts in the vicinity of what you just replaced. Even with the schematic, you
likely would not have that info.
One note: this all *assumes* the 1150 wanders around when removed from the
device and powered with bench gear. If that step hasn’t been done yet, it’s on the
list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping
for that :) ) or some sort of alternate unit cobbled into the device.
Bob
> On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts <time-nuts@lists.febo.com> wrote:
>
> I have revived an Austron 2110 disciplined oscillator. This consists of an
> Austron 1150 high-end double oven OCXO and a computer to lock it to an
> incoming reference frequency and do some statistics and drift prediction.
> It mostly works, but --
>
> After a week of warmup, the OCXO is still about 25 ppb low, which is
> outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
> tuning adjustment, which the manual says to use in this case, but --
>
> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
> TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
> with a period of just under 8 seconds and an amplitude of about 0.42 ns
> p-p. The ADEV measurement shows corresponding ripples in an otherwise
> stable trace.
>
> My interpretation of this data is that one of the two oven controllers has
> gone wonky and increased its gain to the point that it's now unstable. This
> might well be due to the failure of an electrolytic capacitor in an
> integrator loop.
>
> The manual refers to the 1150 as a "sealed unit" which must be returned to
> the factory for repair. After 40-odd years and 5 or 6 changes of corporate
> control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
> MicroChip), factory repair is no longer possible, so I need advice from the
> group ...
>
> Is oven controller failure a reasonable diagnosis?
>
> Does anyone have documentation on the 1150 oscillator? (There is a
> datasheet on KO4BB, but it describes a slightly different version. And I
> have the pinout from the 2110 manual. In places, the documentation refers
> to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
> might also be useful. But is there anything else?)
>
> Has anyone repaired a 1150, or even taken one apart? Are there photos of a
> teardown anywhere?
>
> Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
> any other relevant comments?
>
> Cheers!
> --Stu
> _______________________________________________
> time-nuts mailing list -- time-nuts@lists.febo.com
> To unsubscribe send an email to time-nuts-leave@lists.febo.com
MG
Michael Garvey
Sun, Jan 12, 2025 7:56 PM
It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
Mike Garvey
-----Original Message-----
From: Bob Camp via time-nuts time-nuts@lists.febo.com
Sent: Sunday, January 12, 2025 8:29 AM
To: Discussion of precise time and frequency measurement time-nuts@lists.febo.com
Cc: Bob Camp kb8tq@n1k.org
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test methods improved.
The result is that there are likely many schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
One note: this all assumes the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
Bob
On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts time-nuts@lists.febo.com wrote:
I have revived an Austron 2110 disciplined oscillator. This consists
of an Austron 1150 high-end double oven OCXO and a computer to lock it
to an incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical
coarse tuning adjustment, which the manual says to use in this case,
but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
and TimeLab (thanks, John Miles!), its phase traces out a beautiful
sine wave with a period of just under 8 seconds and an amplitude of
about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
an otherwise stable trace.
My interpretation of this data is that one of the two oven controllers
has gone wonky and increased its gain to the point that it's now
unstable. This might well be due to the failure of an electrolytic
capacitor in an integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be
returned to the factory for repair. After 40-odd years and 5 or 6
changes of corporate control (Sulzer? -> Austron -> Datum ->
Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
possible, so I need advice from the group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And
I have the pinout from the 2110 manual. In places, the documentation
refers to the 1150 as a Sulzer oscillator, so TVB's references a few
days ago might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos
of a teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO
nondestructively? Or any other relevant comments?
Cheers!
--Stu
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
an email to time-nuts-leave@lists.febo.com
It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
Mike Garvey
-----Original Message-----
From: Bob Camp via time-nuts <time-nuts@lists.febo.com>
Sent: Sunday, January 12, 2025 8:29 AM
To: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com>
Cc: Bob Camp <kb8tq@n1k.org>
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test methods improved.
The result is that there are likely *many* schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
One note: this all *assumes* the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
Bob
> On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts <time-nuts@lists.febo.com> wrote:
>
> I have revived an Austron 2110 disciplined oscillator. This consists
> of an Austron 1150 high-end double oven OCXO and a computer to lock it
> to an incoming reference frequency and do some statistics and drift prediction.
> It mostly works, but --
>
> After a week of warmup, the OCXO is still about 25 ppb low, which is
> outside the locking range of about 15 ppb. The OCXO has a mechanical
> coarse tuning adjustment, which the manual says to use in this case,
> but --
>
> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
> and TimeLab (thanks, John Miles!), its phase traces out a beautiful
> sine wave with a period of just under 8 seconds and an amplitude of
> about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
> an otherwise stable trace.
>
> My interpretation of this data is that one of the two oven controllers
> has gone wonky and increased its gain to the point that it's now
> unstable. This might well be due to the failure of an electrolytic
> capacitor in an integrator loop.
>
> The manual refers to the 1150 as a "sealed unit" which must be
> returned to the factory for repair. After 40-odd years and 5 or 6
> changes of corporate control (Sulzer? -> Austron -> Datum ->
> Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
> possible, so I need advice from the group ...
>
> Is oven controller failure a reasonable diagnosis?
>
> Does anyone have documentation on the 1150 oscillator? (There is a
> datasheet on KO4BB, but it describes a slightly different version. And
> I have the pinout from the 2110 manual. In places, the documentation
> refers to the 1150 as a Sulzer oscillator, so TVB's references a few
> days ago might also be useful. But is there anything else?)
>
> Has anyone repaired a 1150, or even taken one apart? Are there photos
> of a teardown anywhere?
>
> Does anyone have advice on how to open a "sealed" OCXO
> nondestructively? Or any other relevant comments?
>
> Cheers!
> --Stu
> _______________________________________________
> time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
> an email to time-nuts-leave@lists.febo.com
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com
BC
Bob Camp
Sun, Jan 12, 2025 8:17 PM
Hi
….. same thing when I worked at / ran / managed engineering at Motorola, Harris Crystal / Cinox / EG&G / Oscillatek / Vectron / McCoy. I think that was a bit over 40 years of fun. Weird that other than Motorola, that’s all morphed into one company.
Bob
On Jan 12, 2025, at 2:56 PM, Michael Garvey r3m1g4@verizon.net wrote:
It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
Mike Garvey
-----Original Message-----
From: Bob Camp via time-nuts time-nuts@lists.febo.com
Sent: Sunday, January 12, 2025 8:29 AM
To: Discussion of precise time and frequency measurement time-nuts@lists.febo.com
Cc: Bob Camp kb8tq@n1k.org
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test methods improved.
The result is that there are likely many schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
One note: this all assumes the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
Bob
On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts time-nuts@lists.febo.com wrote:
I have revived an Austron 2110 disciplined oscillator. This consists
of an Austron 1150 high-end double oven OCXO and a computer to lock it
to an incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical
coarse tuning adjustment, which the manual says to use in this case,
but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
and TimeLab (thanks, John Miles!), its phase traces out a beautiful
sine wave with a period of just under 8 seconds and an amplitude of
about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
an otherwise stable trace.
My interpretation of this data is that one of the two oven controllers
has gone wonky and increased its gain to the point that it's now
unstable. This might well be due to the failure of an electrolytic
capacitor in an integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be
returned to the factory for repair. After 40-odd years and 5 or 6
changes of corporate control (Sulzer? -> Austron -> Datum ->
Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
possible, so I need advice from the group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And
I have the pinout from the 2110 manual. In places, the documentation
refers to the 1150 as a Sulzer oscillator, so TVB's references a few
days ago might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos
of a teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO
nondestructively? Or any other relevant comments?
Cheers!
--Stu
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
an email to time-nuts-leave@lists.febo.com
Hi
….. same thing when I worked at / ran / managed engineering at Motorola, Harris Crystal / Cinox / EG&G / Oscillatek / Vectron / McCoy. I think that was a bit over 40 years of fun. Weird that other than Motorola, that’s all morphed into one company.
Bob
> On Jan 12, 2025, at 2:56 PM, Michael Garvey <r3m1g4@verizon.net> wrote:
>
> It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
>
> I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
> Mike Garvey
>
> -----Original Message-----
> From: Bob Camp via time-nuts <time-nuts@lists.febo.com>
> Sent: Sunday, January 12, 2025 8:29 AM
> To: Discussion of precise time and frequency measurement <time-nuts@lists.febo.com>
> Cc: Bob Camp <kb8tq@n1k.org>
> Subject: [time-nuts] Re: Austron 1150 OCXO advice?
>
> Hi
>
> Chances are that over the “life” of the 1150 it got redesigned multiple times.
> Parts supplies changed. Design approach improved. Production and test methods improved.
>
> The result is that there are likely *many* schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
> one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
>
> Simple answer (ok not so simple) is to pop it open and start tracing this and that.
>
> If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
>
> Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
>
> One note: this all *assumes* the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
>
> Lots of fun !!!!
>
> Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
>
> Bob
>
>
>
>> On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts <time-nuts@lists.febo.com> wrote:
>>
>> I have revived an Austron 2110 disciplined oscillator. This consists
>> of an Austron 1150 high-end double oven OCXO and a computer to lock it
>> to an incoming reference frequency and do some statistics and drift prediction.
>> It mostly works, but --
>>
>> After a week of warmup, the OCXO is still about 25 ppb low, which is
>> outside the locking range of about 15 ppb. The OCXO has a mechanical
>> coarse tuning adjustment, which the manual says to use in this case,
>> but --
>>
>> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
>> and TimeLab (thanks, John Miles!), its phase traces out a beautiful
>> sine wave with a period of just under 8 seconds and an amplitude of
>> about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
>> an otherwise stable trace.
>>
>> My interpretation of this data is that one of the two oven controllers
>> has gone wonky and increased its gain to the point that it's now
>> unstable. This might well be due to the failure of an electrolytic
>> capacitor in an integrator loop.
>>
>> The manual refers to the 1150 as a "sealed unit" which must be
>> returned to the factory for repair. After 40-odd years and 5 or 6
>> changes of corporate control (Sulzer? -> Austron -> Datum ->
>> Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
>> possible, so I need advice from the group ...
>>
>> Is oven controller failure a reasonable diagnosis?
>>
>> Does anyone have documentation on the 1150 oscillator? (There is a
>> datasheet on KO4BB, but it describes a slightly different version. And
>> I have the pinout from the 2110 manual. In places, the documentation
>> refers to the 1150 as a Sulzer oscillator, so TVB's references a few
>> days ago might also be useful. But is there anything else?)
>>
>> Has anyone repaired a 1150, or even taken one apart? Are there photos
>> of a teardown anywhere?
>>
>> Does anyone have advice on how to open a "sealed" OCXO
>> nondestructively? Or any other relevant comments?
>>
>> Cheers!
>> --Stu
>> _______________________________________________
>> time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
>> an email to time-nuts-leave@lists.febo.com
> _______________________________________________
> time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com
>
TK
Tom Knox
Sun, Jan 12, 2025 9:29 PM
Hi Stu,
I really like the 2110, I hope you can resolve the issues. I may be repeating previous comments, but the unit has two modes Resume/Coast and Start/Halt, I believe both allow the unit to free run, not exactly sure of the difference between the two mode and if one would be better than another but that should offer some insights.
Actually, there is a lot of information you can see from the front panel including the ET (electronic tune) voltage.
If you get it running, I have the details I can send you for a higher contrast LED backlite LCD.
Good Luck
Tom Knox
SR Test and Measurement Engineer
Phoenix Research Group / Ascent Concepts and Technology
4870 Meredith Way Apt 102
Boulder, Co 80303
Formerly of:
357 Fox Lane
Superior Co 80027
303-554-0307
actast@hotmail.com
https://www.thedenverchannel.com/news/marshall-fire/superior-man-moving-forward-after-losing-dream-research-lab-during-marshall-fire
"Peace is not the absence of violence, but the presence of Justice" Both MLK and Albert Einstein
From: Tom Van Baak via time-nuts time-nuts@lists.febo.com
Sent: Sunday, January 12, 2025 3:36 AM
To: time-nuts@lists.febo.com time-nuts@lists.febo.com
Cc: Tom Van Baak tvb@LeapSecond.com
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Stu,
its phase traces out a beautiful sine wave with a period of just
under 8 seconds
First, in TimeLab use the 'n' command to apply a notch filter. Screen
capture the result. Also compare the ADEV before and after. The center
frequency and Q may hint of the origin of the ~8 second modulation. This
won't take more than a minute of your time since you already have the
data. Hint: keep your original .TIM file; do not save over it with the
filtered version.
Then, a couple of ideas for you, pick one:
-
Since your OCXO is outside the locking range I wonder if you're
simply seeing a discipline loop trying and failing repeatedly. I've seen
other standards that go into a similar sweep mode while trying to lock.
Two ways to check. If you have a good DMM monitor the EFC voltage for 10
or 20 seconds and see if there's a waveform. Or, just ground the EFC pin
and see if the 8 second modulation goes away. Check the board /
schematic to see if this is safe, i.e., you may not want to short a DAC
output to ground.
-
Simply remove the external reference frequency so that the 2110 is
free-running. See if the sweeping continues or goes away.
-
Pull the OCXO out of the instrument and run it stand-alone from a
bench supply. If the cycling is gone then the fault is not the OCXO but
instead something in the 2110.
I still have my Austron 2110 on the bench from last we talked (about the
power supply), so if there's anything I can do to help let me know. But
first try one of the above simple diagnostics.
/tvb
On 1/11/2025 9:42 PM, Stewart Cobb via time-nuts wrote:
I have revived an Austron 2110 disciplined oscillator. This consists of an
Austron 1150 high-end double oven OCXO and a computer to lock it to an
incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
tuning adjustment, which the manual says to use in this case, but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
with a period of just under 8 seconds and an amplitude of about 0.42 ns
p-p. The ADEV measurement shows corresponding ripples in an otherwise
stable trace.
My interpretation of this data is that one of the two oven controllers has
gone wonky and increased its gain to the point that it's now unstable. This
might well be due to the failure of an electrolytic capacitor in an
integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be returned to
the factory for repair. After 40-odd years and 5 or 6 changes of corporate
control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
MicroChip), factory repair is no longer possible, so I need advice from the
group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And I
have the pinout from the 2110 manual. In places, the documentation refers
to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos of a
teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
any other relevant comments?
Cheers!
--Stu@lists.febo.com
Hi Stu,
I really like the 2110, I hope you can resolve the issues. I may be repeating previous comments, but the unit has two modes Resume/Coast and Start/Halt, I believe both allow the unit to free run, not exactly sure of the difference between the two mode and if one would be better than another but that should offer some insights.
Actually, there is a lot of information you can see from the front panel including the ET (electronic tune) voltage.
If you get it running, I have the details I can send you for a higher contrast LED backlite LCD.
Good Luck
Tom Knox
SR Test and Measurement Engineer
Phoenix Research Group / Ascent Concepts and Technology
4870 Meredith Way Apt 102
Boulder, Co 80303
Formerly of:
357 Fox Lane
Superior Co 80027
303-554-0307
actast@hotmail.com
https://www.thedenverchannel.com/news/marshall-fire/superior-man-moving-forward-after-losing-dream-research-lab-during-marshall-fire
"Peace is not the absence of violence, but the presence of Justice" Both MLK and Albert Einstein
________________________________
From: Tom Van Baak via time-nuts <time-nuts@lists.febo.com>
Sent: Sunday, January 12, 2025 3:36 AM
To: time-nuts@lists.febo.com <time-nuts@lists.febo.com>
Cc: Tom Van Baak <tvb@LeapSecond.com>
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Stu,
> its phase traces out a beautiful sine wave with a period of just
under 8 seconds
First, in TimeLab use the 'n' command to apply a notch filter. Screen
capture the result. Also compare the ADEV before and after. The center
frequency and Q may hint of the origin of the ~8 second modulation. This
won't take more than a minute of your time since you already have the
data. Hint: keep your original .TIM file; do not save over it with the
filtered version.
Then, a couple of ideas for you, pick one:
1) Since your OCXO is outside the locking range I wonder if you're
simply seeing a discipline loop trying and failing repeatedly. I've seen
other standards that go into a similar sweep mode while trying to lock.
Two ways to check. If you have a good DMM monitor the EFC voltage for 10
or 20 seconds and see if there's a waveform. Or, just ground the EFC pin
and see if the 8 second modulation goes away. Check the board /
schematic to see if this is safe, i.e., you may not want to short a DAC
output to ground.
2) Simply remove the external reference frequency so that the 2110 is
free-running. See if the sweeping continues or goes away.
3) Pull the OCXO out of the instrument and run it stand-alone from a
bench supply. If the cycling is gone then the fault is not the OCXO but
instead something in the 2110.
I still have my Austron 2110 on the bench from last we talked (about the
power supply), so if there's anything I can do to help let me know. But
first try one of the above simple diagnostics.
/tvb
On 1/11/2025 9:42 PM, Stewart Cobb via time-nuts wrote:
> I have revived an Austron 2110 disciplined oscillator. This consists of an
> Austron 1150 high-end double oven OCXO and a computer to lock it to an
> incoming reference frequency and do some statistics and drift prediction.
> It mostly works, but --
>
> After a week of warmup, the OCXO is still about 25 ppb low, which is
> outside the locking range of about 15 ppb. The OCXO has a mechanical coarse
> tuning adjustment, which the manual says to use in this case, but --
>
> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!) and
> TimeLab (thanks, John Miles!), its phase traces out a beautiful sine wave
> with a period of just under 8 seconds and an amplitude of about 0.42 ns
> p-p. The ADEV measurement shows corresponding ripples in an otherwise
> stable trace.
>
> My interpretation of this data is that one of the two oven controllers has
> gone wonky and increased its gain to the point that it's now unstable. This
> might well be due to the failure of an electrolytic capacitor in an
> integrator loop.
>
> The manual refers to the 1150 as a "sealed unit" which must be returned to
> the factory for repair. After 40-odd years and 5 or 6 changes of corporate
> control (Sulzer? -> Austron -> Datum -> Symmetricom -> MicroSemi ->
> MicroChip), factory repair is no longer possible, so I need advice from the
> group ...
>
> Is oven controller failure a reasonable diagnosis?
>
> Does anyone have documentation on the 1150 oscillator? (There is a
> datasheet on KO4BB, but it describes a slightly different version. And I
> have the pinout from the 2110 manual. In places, the documentation refers
> to the 1150 as a Sulzer oscillator, so TVB's references a few days ago
> might also be useful. But is there anything else?)
>
> Has anyone repaired a 1150, or even taken one apart? Are there photos of a
> teardown anywhere?
>
> Does anyone have advice on how to open a "sealed" OCXO nondestructively? Or
> any other relevant comments?
>
> Cheers!
> --Stu@lists.febo.com
>
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
JL
Jim Lux
Mon, Jan 13, 2025 5:15 PM
House numbers for parts are a pain - at JPL, we often generate house numbers (or, as Mike writes, a drawing number), not because the parts are special, but just as a way to track them.
The other problem is that it might have a generic 2N number - there's a JFET that is known for low leakage, low noise (2N 44??) , but only if it's the one from some fab in the UK. Yes, it meets all the JEDEC 2N parameters, so it's legitimately labeled with the 2N number, but those ones are special.
And even if there's a legitimate part number, they may have done a special run. We had a SBIR contractor at JPL who made some nice vector modulators in GaAs (or maybe InGaP) for either X or Ka band (Can't remember). We published papers on them, part number and all. Then 5 years later someone wrote me asking where we got them, and it turns out while it had a standard number, they made a single wafer, packaged 25 of them, and that's it.
On Sun, 12 Jan 2025 15:17:33 -0500, Bob Camp via time-nuts time-nuts@lists.febo.com wrote:
Hi
….. same thing when I worked at / ran / managed engineering at Motorola, Harris Crystal / Cinox / EG&G / Oscillatek / Vectron / McCoy. I think that was a bit over 40 years of fun. Weird that other than Motorola, that’s all morphed into one company.
Bob
On Jan 12, 2025, at 2:56 PM, Michael Garvey wrote:
It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
Mike Garvey
-----Original Message-----
From: Bob Camp via time-nuts
Sent: Sunday, January 12, 2025 8:29 AM
To: Discussion of precise time and frequency measurement
Cc: Bob Camp
Subject: [time-nuts] Re: Austron 1150 OCXO advice?
Hi
Chances are that over the “life” of the 1150 it got redesigned multiple times.
Parts supplies changed. Design approach improved. Production and test methods improved.
The result is that there are likely many schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
Simple answer (ok not so simple) is to pop it open and start tracing this and that.
If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
One note: this all assumes the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
Lots of fun !!!!
Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
Bob
On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts wrote:
I have revived an Austron 2110 disciplined oscillator. This consists
of an Austron 1150 high-end double oven OCXO and a computer to lock it
to an incoming reference frequency and do some statistics and drift prediction.
It mostly works, but --
After a week of warmup, the OCXO is still about 25 ppb low, which is
outside the locking range of about 15 ppb. The OCXO has a mechanical
coarse tuning adjustment, which the manual says to use in this case,
but --
When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
and TimeLab (thanks, John Miles!), its phase traces out a beautiful
sine wave with a period of just under 8 seconds and an amplitude of
about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
an otherwise stable trace.
My interpretation of this data is that one of the two oven controllers
has gone wonky and increased its gain to the point that it's now
unstable. This might well be due to the failure of an electrolytic
capacitor in an integrator loop.
The manual refers to the 1150 as a "sealed unit" which must be
returned to the factory for repair. After 40-odd years and 5 or 6
changes of corporate control (Sulzer? -> Austron -> Datum ->
Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
possible, so I need advice from the group ...
Is oven controller failure a reasonable diagnosis?
Does anyone have documentation on the 1150 oscillator? (There is a
datasheet on KO4BB, but it describes a slightly different version. And
I have the pinout from the 2110 manual. In places, the documentation
refers to the 1150 as a Sulzer oscillator, so TVB's references a few
days ago might also be useful. But is there anything else?)
Has anyone repaired a 1150, or even taken one apart? Are there photos
of a teardown anywhere?
Does anyone have advice on how to open a "sealed" OCXO
nondestructively? Or any other relevant comments?
Cheers!
--Stu
time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
an email to time-nuts-leave@lists.febo.com
House numbers for parts are a pain - at JPL, we often generate house numbers (or, as Mike writes, a drawing number), not because the parts are special, but just as a way to track them.
The other problem is that it might have a generic 2N number - there's a JFET that is known for low leakage, low noise (2N 44??) , but only if it's the one from some fab in the UK. Yes, it meets all the JEDEC 2N parameters, so it's legitimately labeled with the 2N number, but those ones are special.
And even if there's a legitimate part number, they may have done a special run. We had a SBIR contractor at JPL who made some nice vector modulators in GaAs (or maybe InGaP) for either X or Ka band (Can't remember). We published papers on them, part number and all. Then 5 years later someone wrote me asking where we got them, and it turns out while it had a standard number, they made a single wafer, packaged 25 of them, and that's it.
On Sun, 12 Jan 2025 15:17:33 -0500, Bob Camp via time-nuts <time-nuts@lists.febo.com> wrote:
Hi
….. same thing when I worked at / ran / managed engineering at Motorola, Harris Crystal / Cinox / EG&G / Oscillatek / Vectron / McCoy. I think that was a bit over 40 years of fun. Weird that other than Motorola, that’s all morphed into one company.
Bob
> On Jan 12, 2025, at 2:56 PM, Michael Garvey wrote:
>
> It is perhaps useful to realize that oscillator designs are/were considered by many suppliers as trade secrets and the availability of schematics is quite limited. Further, some components were screened (eg for noise performance) and will show as a drawing number with little information about what the original part was.
>
> I was an engineering manager at FTS, Datum, Symmetricom, etc for 30 years.
> Mike Garvey
>
> -----Original Message-----
> From: Bob Camp via time-nuts
> Sent: Sunday, January 12, 2025 8:29 AM
> To: Discussion of precise time and frequency measurement
> Cc: Bob Camp
> Subject: [time-nuts] Re: Austron 1150 OCXO advice?
>
> Hi
>
> Chances are that over the “life” of the 1150 it got redesigned multiple times.
> Parts supplies changed. Design approach improved. Production and test methods improved.
>
> The result is that there are likely *many* schematics of the internals of a 1150 OCXO. You need the one that matches up with the OCXO you have. The “real”
> one might have been a marked up copy of the one in the filing cabinet that showed the “in production” changes ….
>
> Simple answer (ok not so simple) is to pop it open and start tracing this and that.
>
> If there is a true integrator in the control loop it’s not going to have an electrolytic cap. It would use something with much better leakage performance. That’s not to say the heater isn’t the issue.
>
> Next up once you find the “problem” is to work out the test approach for selecting the parts in the vicinity of what you just replaced. Even with the schematic, you likely would not have that info.
>
> One note: this all *assumes* the 1150 wanders around when removed from the device and powered with bench gear. If that step hasn’t been done yet, it’s on the list ahead of tearing into the OCXO.,
>
> Lots of fun !!!!
>
> Alternative is to replace the OCXO, either with another 1150 (have fun shopping for that :) ) or some sort of alternate unit cobbled into the device.
>
> Bob
>
>
>
>> On Jan 12, 2025, at 12:42 AM, Stewart Cobb via time-nuts wrote:
>>
>> I have revived an Austron 2110 disciplined oscillator. This consists
>> of an Austron 1150 high-end double oven OCXO and a computer to lock it
>> to an incoming reference frequency and do some statistics and drift prediction.
>> It mostly works, but --
>>
>> After a week of warmup, the OCXO is still about 25 ppb low, which is
>> outside the locking range of about 15 ppb. The OCXO has a mechanical
>> coarse tuning adjustment, which the manual says to use in this case,
>> but --
>>
>> When I look at the main output with a TinyPFA (thanks, Erik Kaashoek!)
>> and TimeLab (thanks, John Miles!), its phase traces out a beautiful
>> sine wave with a period of just under 8 seconds and an amplitude of
>> about 0.42 ns p-p. The ADEV measurement shows corresponding ripples in
>> an otherwise stable trace.
>>
>> My interpretation of this data is that one of the two oven controllers
>> has gone wonky and increased its gain to the point that it's now
>> unstable. This might well be due to the failure of an electrolytic
>> capacitor in an integrator loop.
>>
>> The manual refers to the 1150 as a "sealed unit" which must be
>> returned to the factory for repair. After 40-odd years and 5 or 6
>> changes of corporate control (Sulzer? -> Austron -> Datum ->
>> Symmetricom -> MicroSemi -> MicroChip), factory repair is no longer
>> possible, so I need advice from the group ...
>>
>> Is oven controller failure a reasonable diagnosis?
>>
>> Does anyone have documentation on the 1150 oscillator? (There is a
>> datasheet on KO4BB, but it describes a slightly different version. And
>> I have the pinout from the 2110 manual. In places, the documentation
>> refers to the 1150 as a Sulzer oscillator, so TVB's references a few
>> days ago might also be useful. But is there anything else?)
>>
>> Has anyone repaired a 1150, or even taken one apart? Are there photos
>> of a teardown anywhere?
>>
>> Does anyone have advice on how to open a "sealed" OCXO
>> nondestructively? Or any other relevant comments?
>>
>> Cheers!
>> --Stu
>> _______________________________________________
>> time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send
>> an email to time-nuts-leave@lists.febo.com
> _______________________________________________
> time-nuts mailing list -- time-nuts@lists.febo.com To unsubscribe send an email to time-nuts-leave@lists.febo.com
>
_______________________________________________
time-nuts mailing list -- time-nuts@lists.febo.com
To unsubscribe send an email to time-nuts-leave@lists.febo.com
