Electolytics in signal path??

[Cursitor Doom]

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[John Larkin]

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
<curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency
apps.

2.2u implies a really low frequency corner. What instrument might that
be?




--

John Larkin Highland Technology, Inc trk

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Cursitor Doom]

On Sun, 11 Aug 2019 10:30:01 -0700, John Larkin wrote:

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency apps.

2.2u implies a really low frequency corner. What instrument might that
be?

It's an HP 8566B spectrum analyser 100Hz-22Ghz (but that is split into
two ranges at 2.5Ghz). It's got the "YTO unlocked" error - a known issue
with these otherwise excellent vintage instruments - which I'm in the
process of chasing down.



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

At higher frequencies that ESL might show itself.

 

[bitrex]

On 8/11/19 1:30 PM, John Larkin wrote:

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency
apps.

2.2u implies a really low frequency corner. What instrument might that
be?

also depends on the age of the equipment - if it's old (okay all
HP-branded test equipment is "old" to some standard in 2019 I guess, HP
was been making equipment since the 30s or 40s?) an electrolytic might
be all they have to work with as compared to wax paper caps, IDK if they
even made them that large.

If it's "newer" I don't think good-quality monolithic SMT ceramics in
that size were available in the 1980s and early 90s and a 2.2u
through-hole or axial film cap is gonna run you what, $3 in singles?
Even HP has to hit a budget

 

[bitrex]

On 8/12/19 3:30 PM, Phil Hobbs wrote:

On 8/11/19 3:20 PM, Cursitor Doom wrote:
On Sun, 11 Aug 2019 10:30:01 -0700, John Larkin wrote:

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency apps.

2.2u implies a really low frequency corner. What instrument might that
be?

It's an HP 8566B spectrum analyser 100Hz-22Ghz (but that is split into
two ranges at 2.5Ghz). It's got the "YTO unlocked" error - a known issue
with these otherwise excellent vintage instruments - which I'm in the
process of chasing down.




An electrolytic has a lot in common with a really dead battery, so I'd
be concerned about temperature gradients causing voltage offsets.

Cheers

Phil Hobbs

it's a nice unit that I estimate the chances of CD successfully fixing
up at 'bout 0%

 

[Phil Hobbs]

On 8/11/19 3:20 PM, Cursitor Doom wrote:

On Sun, 11 Aug 2019 10:30:01 -0700, John Larkin wrote:

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency apps.

2.2u implies a really low frequency corner. What instrument might that
be?

It's an HP 8566B spectrum analyser 100Hz-22Ghz (but that is split into
two ranges at 2.5Ghz). It's got the "YTO unlocked" error - a known issue
with these otherwise excellent vintage instruments - which I'm in the
process of chasing down.

An electrolytic has a lot in common with a really dead battery, so I'd
be concerned about temperature gradients causing voltage offsets.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
https://hobbs-eo.com

 

[Phil Hobbs]

On 8/12/19 3:40 PM, bitrex wrote:

On 8/12/19 3:30 PM, Phil Hobbs wrote:
On 8/11/19 3:20 PM, Cursitor Doom wrote:
On Sun, 11 Aug 2019 10:30:01 -0700, John Larkin wrote:

On Sun, 11 Aug 2019 14:22:02 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

Greetings, gentlemen,

Would I be correct in thinking that no commercial enterprise, least of
all one like the one-time gloriously sublime HP, would put a 2.2uF
electrolytic cap in the signal path of a high-end RF board, surely?

Why not? Size or ESL might be considered for really high frequency
apps.

2.2u implies a really low frequency corner. What instrument might that
be?

It's an HP 8566B spectrum analyser 100Hz-22Ghz (but that is split into
two ranges at 2.5Ghz). It's got the "YTO unlocked" error - a known issue
with these otherwise excellent vintage instruments - which I'm in the
process of chasing down.




An electrolytic has a lot in common with a really dead battery, so I'd
be concerned about temperature gradients causing voltage offsets.


it's a nice unit that I estimate the chances of CD successfully fixing
up at 'bout 0%

Bet they taught you all about HP stuff in art school.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Cursitor Doom]

On Mon, 12 Aug 2019 21:22:19 -0400, Phil Hobbs wrote:

On 8/12/19 3:40 PM, bitrex wrote:

it's a nice unit that I estimate the chances of CD successfully fixing
up at 'bout 0%

Bet they taught you all about HP stuff in art school.

LOL! You got that right, Phil.

Actually, I'm a pretty darn good troubleshooter and have fixed a lot of
high-end gear in the past, although admittedly this one's probably the
most complex unit I've undertaken to date. Given an oscilloscope and a
schematic I can most times succeed in finding and fixing these faults.
I'm way better at fault-finding than I am at designing, anyway.
You have one of these IIRC, don't you, Phil? Have you ever had the "YTO
unlocked" error? It refers to a YIG-tuned oscillator which I've never
heard of before. I'm hopeful I can rectify the fault without having to
spend time studying the minutiae of YIG-tuned oscillators first!!



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Phil Hobbs]

On 8/13/19 1:29 PM, Cursitor Doom wrote:

On Mon, 12 Aug 2019 21:22:19 -0400, Phil Hobbs wrote:

On 8/12/19 3:40 PM, bitrex wrote:

it's a nice unit that I estimate the chances of CD successfully
fixing up at 'bout 0%

Bet they taught you all about HP stuff in art school.

LOL! You got that right, Phil.

Actually, I'm a pretty darn good troubleshooter and have fixed a lot
of high-end gear in the past, although admittedly this one's probably
the most complex unit I've undertaken to date. Given an oscilloscope
and a schematic I can most times succeed in finding and fixing these
faults. I'm way better at fault-finding than I am at designing,
anyway. You have one of these IIRC, don't you, Phil?

Yup, it's by my elbow right now.

Have you ever had the "YTO unlocked" error? It refers to a YIG-tuned
oscillator which I've never heard of before. I'm hopeful I can
rectify the fault without having to spend time studying the minutiae
of YIG-tuned oscillators first!!

A YTO is a current-controlled oscillator, as opposed to
voltage-controlled. There's a yttrium-iron garnet (YIG) cavity
resonator, whose resonance is controlled magnetically via a coil.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Cursitor Doom]

On Tue, 13 Aug 2019 14:54:31 -0400, Phil Hobbs wrote:

A YTO is a current-controlled oscillator, as opposed to
voltage-controlled. There's a yttrium-iron garnet (YIG) cavity
resonator, whose resonance is controlled magnetically via a coil.

Right, so that's this mysterious "main coil" the schematics are referring
to I would imagine. You obviously haven't had the issue with yours yet.
It's pretty common, so I'll be surprised if you escape indefinitely. The
good news is it's almost always something very cheap and easy to fix
that's gone awry; usually a trimmer pot or a fixed value capacitor so
we're talking pennies. The tricky bit is identifying exactly *which*
small, cheap component has failed.




--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Cursitor Doom]

On Tue, 13 Aug 2019 16:01:43 -0700, John Larkin wrote:

I never understood how a magnetic field can be controlled so well as to
result in that low phase noise.

I've never understood how anyone could think an SDR-based instrument
could equal an analogue device for spectral purity/low phase noise.



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Cursitor Doom]

On Tue, 13 Aug 2019 18:48:14 -0400, Phil Hobbs wrote:

There's also often a vernier coil with much lower inductance (and so
much less tuning range but higher speed) that's used for the actual
phase locking. YIG-tuned oscillators have unbeatable close-in phase
noise, which is why an 8566B can still run rings around any SDR-based
system for phase noise.

No surprise to me at all about that last point. You may have the
advantage over me if you have a physical hard copy of the service manual.
I have to scroll through 900 pages of pdf files of varying scan quality
to find anything I want, which is a PITA, but I suppose I should be damn
grateful for being able to even have that.

I haven't had mine apart yet, except for swapping out a bad A3A4 display
memory board, of which I happened to have one in stock. One fine day
I'm going to declare Fix-It Week and get a bunch of sick gear back
online. For instance:

HP 3585A DC-40 MHz spectrum analyzer (a fave of Joerg's that has an LO
unlock problem)

Tek 466A analogue storage scope--worked fine till I turned on the
storage function, and never since.

Tek TDS 694C 3-GHz, 10-GS/s scope that emits a piercing shriek from the
CRT high voltage section for the first half hour of operation. Still in
use, but needs an LCD transplant.

Tek P6249 4 GHz probe for the above, which oscillates at about 2
MHz--probably a dead cap someplace.

Plus a couple of RMS voltmeters (HP 3400A and Boonton 93A), a couple of
HP 3456A 6.5 digit DMMs, and various smaller stuff.

Oh me oh my that's going to be a busy "week" for you. :-D.
I was able to rescue a 466B (I think it's the B IIRC) from a charity
shop. It wasn't working but I got it going quite easily and am really
pleased with it. There's something compelling about analogue storage
scopes for people of my age (don't ask). I have a modern Tek DSO as well,
but if there is something the 466 can handle it's my first choice now, as
it's so easy to intuitively set up in under 2 minutes whereas if it's the
Tek I have to RTFM every time, so 50 mins of set up minimum *if* I'm very
lucky. I suppose it would be different if I were using it every day but
I'm just a hobbyist and have so much test gear I can't possibly remember
the quirks of every single instrument. I recall you have an impressive
selection yourself so you know where I'm coming from on this no doubt.




--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Phil Hobbs]

On 8/13/19 7:01 PM, John Larkin wrote:

On Tue, 13 Aug 2019 18:48:14 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 8/13/19 5:30 PM, Cursitor Doom wrote:
On Tue, 13 Aug 2019 14:54:31 -0400, Phil Hobbs wrote:

A YTO is a current-controlled oscillator, as opposed to
voltage-controlled. There's a yttrium-iron garnet (YIG) cavity
resonator, whose resonance is controlled magnetically via a coil.

Right, so that's this mysterious "main coil" the schematics are
referring to I would imagine.

There's also often a vernier coil with much lower inductance (and so
much less tuning range but higher speed) that's used for the actual
phase locking. YIG-tuned oscillators have unbeatable close-in phase
noise, which is why an 8566B can still run rings around any SDR-based
system for phase noise.

I never understood how a magnetic field can be controlled so well as
to result in that low phase noise.

AIUI the inductance of the main coil is some henrys, so that isn't as
hard as all that.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Larkin]

On Tue, 13 Aug 2019 18:48:14 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 8/13/19 5:30 PM, Cursitor Doom wrote:
On Tue, 13 Aug 2019 14:54:31 -0400, Phil Hobbs wrote:

A YTO is a current-controlled oscillator, as opposed to
voltage-controlled. There's a yttrium-iron garnet (YIG) cavity
resonator, whose resonance is controlled magnetically via a coil.

Right, so that's this mysterious "main coil" the schematics are
referring to I would imagine.

There's also often a vernier coil with much lower inductance (and so
much less tuning range but higher speed) that's used for the actual
phase locking. YIG-tuned oscillators have unbeatable close-in phase
noise, which is why an 8566B can still run rings around any SDR-based
system for phase noise.

I never understood how a magnetic field can be controlled so well as
to result in that low phase noise.

 

[Phil Hobbs]

On 8/13/19 5:30 PM, Cursitor Doom wrote:

On Tue, 13 Aug 2019 14:54:31 -0400, Phil Hobbs wrote:

A YTO is a current-controlled oscillator, as opposed to
voltage-controlled. There's a yttrium-iron garnet (YIG) cavity
resonator, whose resonance is controlled magnetically via a coil.

Right, so that's this mysterious "main coil" the schematics are
referring to I would imagine.

There's also often a vernier coil with much lower inductance (and so
much less tuning range but higher speed) that's used for the actual
phase locking. YIG-tuned oscillators have unbeatable close-in phase
noise, which is why an 8566B can still run rings around any SDR-based
system for phase noise.

You obviously haven't had the issue with yours yet. It's pretty
common, so I'll be surprised if you escape indefinitely. The good
news is it's almost always something very cheap and easy to fix
that's gone awry; usually a trimmer pot or a fixed value capacitor
so we're talking pennies. The tricky bit is identifying exactly
*which* small, cheap component has failed.

I haven't had mine apart yet, except for swapping out a bad A3A4 display
memory board, of which I happened to have one in stock. One fine day
I'm going to declare Fix-It Week and get a bunch of sick gear back
online. For instance:

HP 3585A DC-40 MHz spectrum analyzer (a fave of Joerg's that has an LO
unlock problem)

Tek 466A analogue storage scope--worked fine till I turned on the
storage function, and never since.

Tek TDS 694C 3-GHz, 10-GS/s scope that emits a piercing shriek from the
CRT high voltage section for the first half hour of operation. Still in
use, but needs an LCD transplant.

Tek P6249 4 GHz probe for the above, which oscillates at about 2
MHz--probably a dead cap someplace.

Plus a couple of RMS voltmeters (HP 3400A and Boonton 93A), a couple of
HP 3456A 6.5 digit DMMs, and various smaller stuff.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Phil Hobbs]

On 8/13/19 7:31 PM, Cursitor Doom wrote:

On Tue, 13 Aug 2019 18:48:14 -0400, Phil Hobbs wrote:

There's also often a vernier coil with much lower inductance (and so
much less tuning range but higher speed) that's used for the actual
phase locking. YIG-tuned oscillators have unbeatable close-in phase
noise, which is why an 8566B can still run rings around any SDR-based
system for phase noise.

No surprise to me at all about that last point. You may have the
advantage over me if you have a physical hard copy of the service manual.
I have to scroll through 900 pages of pdf files of varying scan quality
to find anything I want, which is a PITA, but I suppose I should be damn
grateful for being able to even have that.

I haven't had mine apart yet, except for swapping out a bad A3A4 display
memory board, of which I happened to have one in stock. One fine day
I'm going to declare Fix-It Week and get a bunch of sick gear back
online. For instance:

HP 3585A DC-40 MHz spectrum analyzer (a fave of Joerg's that has an LO
unlock problem)

Tek 466A analogue storage scope--worked fine till I turned on the
storage function, and never since.

Tek TDS 694C 3-GHz, 10-GS/s scope that emits a piercing shriek from the
CRT high voltage section for the first half hour of operation. Still in
use, but needs an LCD transplant.

Tek P6249 4 GHz probe for the above, which oscillates at about 2
MHz--probably a dead cap someplace.

Plus a couple of RMS voltmeters (HP 3400A and Boonton 93A), a couple of
HP 3456A 6.5 digit DMMs, and various smaller stuff.

Oh me oh my that's going to be a busy "week" for you. :-D.
I was able to rescue a 466B (I think it's the B IIRC) from a charity
shop. It wasn't working but I got it going quite easily and am really
pleased with it. There's something compelling about analogue storage
scopes for people of my age (don't ask). I have a modern Tek DSO as well,
but if there is something the 466 can handle it's my first choice now, as
it's so easy to intuitively set up in under 2 minutes whereas if it's the
Tek I have to RTFM every time, so 50 mins of set up minimum *if* I'm very
lucky. I suppose it would be different if I were using it every day but
I'm just a hobbyist and have so much test gear I can't possibly remember
the quirks of every single instrument. I recall you have an impressive
selection yourself so you know where I'm coming from on this no doubt.

My situation is a bit different as it's my day job. Also, I furnished
my lab at IBM with the same or very similar equipment when it was new.
For instance, I used an HP 8566A fairly intensively back in 1982-3 when
I was doing satcom frequency control stuff.

Knowing the equipment is the main reason I've been able to assemble a
very capable lab of my own for ~ $60k. It's probably more than $2M at
the original list prices.

It's a great time to start out on your own.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[Cursitor Doom]

On Tue, 13 Aug 2019 19:43:25 -0400, Phil Hobbs wrote:

Knowing the equipment is the main reason I've been able to assemble a
very capable lab of my own for ~ $60k. It's probably more than $2M at
the original list prices.

It's a great time to start out on your own.

Not at my age! (don't ask).
Mine too comes out at around $2.5m worth at new prices. Only worth a
small fraction of that value today, but hey-ho.

Better go do some work I guess. <shrug>



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

 

[Phil Hobbs]

On 8/13/19 7:33 PM, Cursitor Doom wrote:

On Tue, 13 Aug 2019 16:01:43 -0700, John Larkin wrote:

I never understood how a magnetic field can be controlled so well as to
result in that low phase noise.

I've never understood how anyone could think an SDR-based instrument
could equal an analogue device for spectral purity/low phase noise.

It may happen. Modern frequency synthesizer chips are pretty amazing
compared to the old single-loop analogue variety.

Cheers

Phil Hobbs


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Miles, KE5FX]

On Tuesday, August 13, 2019 at 4:01:51 PM UTC-7, John Larkin wrote:

I never understood how a magnetic field can be controlled so well as
to result in that low phase noise.

As long as the number of turns of wire doesn't change, which goes without
saying, and the current through the wire doesn't change, which is easy
enough to ensure, the field remains more or less perfectly stable.
There are hysteresis issues, but they don't affect steady-state operation.

The 8566 is nothing special in the phase-noise department, but they got
the UI and the display right, and it is *stable*. I have several other
SAs around here, some much newer and much more powerful than my 8566B.
The 8566B will only be carried out of here after I am, while the others
come and go.

-- john, KE5FX