On Dec 30, 2019, at 15:03, Jim Summers <kd7f@...> wrote:

?

Hi Stephen,

The division in Spokane started in 1979 with production of the 8640. ?I was in the first group of R&D engineers to transfer up from Stanford Park in May 1980, about 2 weeks after Mt St Helens blew its top. ?We were the 8642 development team. ?Stanford Park sent the RF signal generator, modulation analyzer and the audio analyzer product lines to Spokane and kept the microwave signal generator, power measurement and noise figure meter businesses. ?I was a new engineer at the time having been with HP for less than a year before transferring to Spokane.

In about 1992, Spokane division management saw the opportunity that the cellular business represented and asked group management for more resources. ?The result was a no to the resource request but to address the opportunity group management decided to relieve us of much of our then current product lines so we could concentrate on cellular test. ?As a result the signal generator charter went to Santa Rosa. ?

We were successful in cellular test until Agilent layoffs reduced the workforce below critical mass. ?The Spokane operation shut down at the end of October 2010. ?That was when I left Agilent.

The old HP building here also has other companies in it now. ?I don't know much about them though.

Jim

On Dec 30, 2019, at 2:03 PM, Stephen Hanselman <kc4sw.io@...> wrote:

Jim,

When/why did HP move the signal generators from Stanford Park to Spokane? ?I knew they had SA’s at Santa Rosa but not Sig Gens’s too.

Interesting aside about the Santa Rosa site. One of my friends has his company in one of the old buildings. The country and some “other” folks have taken what used to be “Mahogany Row”(div mgt) and turned it into an incubator with 10 or 15 small business in the old exec offices with communal support in the old support area, great re-use of the old building.?

Regards,

?

Stephen Hanselman

Datagate Systems, LLC

On Dec 30, 2019, at 12:55, Jim Summers <kd7f@...> wrote:

?

Actually, the 8642A was the "replacement" for the 8640. ?It's code name internally was Shetland since its Spurs were low :). The 42 was a multi-loop synthesizer, fairly complex, and had low AM noise in addition to low phase noise as design goals. ?

The original PSG family included the 8644, 8645 and 8665 generators. ?The 8664 was basically an 8665 without the output section covering frequencies above 3GHz. ?The 8643 was an 8644 with the transformer power supply replaced by a switch mode supply, didn't allow the 140ns frequency discriminator option and its front panel was modified to delete the synthesis mode buttons. ?The PSG family utilized a single fractional N PLL and used frequency discriminators to improve VCO or YTO phase noise. ?AM noise was a bit higher than that of the 8642.

As I recall (my long term memory may be fading a bit) the 8642 AM floor was around -160dBc/Hz, the PSG generators were around -150dBc/Hz and the generators that followed using log leveling loops were around -130dBc/Hz.

Agilent/Keysight developed another family of generators referred to as PSG after the signal generator charter moved from Spokane Division to Santa Rosa.

As an interesting piece of history, before the PSG family was named the Performance Signal Generator family, one of the ideas for the name was New Family of Generators. ?Luckily someone noticed the initials...

Jim

On Dec 30, 2019, at 11:47 AM, alwyn.seeds1 <a.seeds@...> wrote:

Dear All,

it rather depends what frequency range you need to cover and the offset from carrier at which you need the phase noise to be low.

Also, remember that signal generators have amplitude noise as well as phase noise and this is of particular concern at large offsets (> 50 kHz) from carrier.

The HP8643A and its higher frequency derivatives formed the replacement for the HP8640 cavity tuned signal generator and are particularly good on non-harmonic spurii.

Many DDS based generators have non-harmonic spurii at -70 dBc levels, which are a considerable nuisance when measuring off channel performance of receivers.

The 8643A is a big, heavy and complex beast with many special parts.

This series was replaced with the PSG series, which are still made.

Regards,

Alwyn

_____________________________________________________

Alwyn Seeds, Director
114 Beaufort Street (Management) Company Ltd.,
114 Beaufort Street,
London,
SW3 6BU

114 Beaufort Street (Management) Company Ltd., Registered in England and Wales: No. 02797775 Registered Office Address: 114 Beaufort Street, London, SW3 6BU

______________________________________________________

...also if you don't mind the weight (~25-26Kg) and you want a real service manual with detailed troubleshooting and all the schematics
then I would suggest a R&S SMHU up to 4.3GHz. Its PN is rated at <-144dBc/Hz @ 20KHz offset at 100MHz (tyoical -150dBc/Hz) and <-118dBc/Hz at 4GHz (typical -122dBc/Hz) certainly better than 8665B and it has a better service manual and all the usual modulations and stuff...I had one long ago (also had a 8664A) but I sold both
of them and now I only have a SMIQ06B and a E4432B which rarely use the latter.?

Jim,

When/why did HP move the signal generators from Stanford Park to Spokane? ?I knew they had SA’s at Santa Rosa but not Sig Gens’s too.

Interesting aside about the Santa Rosa site. One of my friends has his company in one of the old buildings. The country and some “other” folks have taken what used to be “Mahogany Row”(div mgt) and turned it into an incubator with 10 or 15 small business in the old exec offices with communal support in the old support area, great re-use of the old building.?

Regards,

?

Stephen Hanselman

Datagate Systems, LLC

On Dec 30, 2019, at 12:55, Jim Summers <kd7f@...> wrote:

?

Actually, the 8642A was the "replacement" for the 8640. ?It's code name internally was Shetland since its Spurs were low :). The 42 was a multi-loop synthesizer, fairly complex, and had low AM noise in addition to low phase noise as design goals. ?

The original PSG family included the 8644, 8645 and 8665 generators. ?The 8664 was basically an 8665 without the output section covering frequencies above 3GHz. ?The 8643 was an 8644 with the transformer power supply replaced by a switch mode supply, didn't allow the 140ns frequency discriminator option and its front panel was modified to delete the synthesis mode buttons. ?The PSG family utilized a single fractional N PLL and used frequency discriminators to improve VCO or YTO phase noise. ?AM noise was a bit higher than that of the 8642.

As I recall (my long term memory may be fading a bit) the 8642 AM floor was around -160dBc/Hz, the PSG generators were around -150dBc/Hz and the generators that followed using log leveling loops were around -130dBc/Hz.

Agilent/Keysight developed another family of generators referred to as PSG after the signal generator charter moved from Spokane Division to Santa Rosa.

As an interesting piece of history, before the PSG family was named the Performance Signal Generator family, one of the ideas for the name was New Family of Generators. ?Luckily someone noticed the initials...

Jim

On Dec 30, 2019, at 11:47 AM, alwyn.seeds1 <a.seeds@...> wrote:

Dear All,

it rather depends what frequency range you need to cover and the offset from carrier at which you need the phase noise to be low.

Also, remember that signal generators have amplitude noise as well as phase noise and this is of particular concern at large offsets (> 50 kHz) from carrier.

The HP8643A and its higher frequency derivatives formed the replacement for the HP8640 cavity tuned signal generator and are particularly good on non-harmonic spurii.

Many DDS based generators have non-harmonic spurii at -70 dBc levels, which are a considerable nuisance when measuring off channel performance of receivers.

The 8643A is a big, heavy and complex beast with many special parts.

This series was replaced with the PSG series, which are still made.

Regards,

Alwyn

_____________________________________________________

Alwyn Seeds, Director
114 Beaufort Street (Management) Company Ltd.,
114 Beaufort Street,
London,
SW3 6BU

114 Beaufort Street (Management) Company Ltd., Registered in England and Wales: No. 02797775 Registered Office Address: 114 Beaufort Street, London, SW3 6BU

______________________________________________________

On Dec 30, 2019, at 12:55, Jim Summers <kd7f@...> wrote:

?

Actually, the 8642A was the "replacement" for the 8640. ?It's code name internally was Shetland since its Spurs were low :). The 42 was a multi-loop synthesizer, fairly complex, and had low AM noise in addition to low phase noise as design goals. ?

The original PSG family included the 8644, 8645 and 8665 generators. ?The 8664 was basically an 8665 without the output section covering frequencies above 3GHz. ?The 8643 was an 8644 with the transformer power supply replaced by a switch mode supply, didn't allow the 140ns frequency discriminator option and its front panel was modified to delete the synthesis mode buttons. ?The PSG family utilized a single fractional N PLL and used frequency discriminators to improve VCO or YTO phase noise. ?AM noise was a bit higher than that of the 8642.

As I recall (my long term memory may be fading a bit) the 8642 AM floor was around -160dBc/Hz, the PSG generators were around -150dBc/Hz and the generators that followed using log leveling loops were around -130dBc/Hz.

Agilent/Keysight developed another family of generators referred to as PSG after the signal generator charter moved from Spokane Division to Santa Rosa.

As an interesting piece of history, before the PSG family was named the Performance Signal Generator family, one of the ideas for the name was New Family of Generators. ?Luckily someone noticed the initials...

Jim

On Dec 30, 2019, at 11:47 AM, alwyn.seeds1 <a.seeds@...> wrote:

Dear All,

it rather depends what frequency range you need to cover and the offset from carrier at which you need the phase noise to be low.

Also, remember that signal generators have amplitude noise as well as phase noise and this is of particular concern at large offsets (> 50 kHz) from carrier.

The HP8643A and its higher frequency derivatives formed the replacement for the HP8640 cavity tuned signal generator and are particularly good on non-harmonic spurii.

Many DDS based generators have non-harmonic spurii at -70 dBc levels, which are a considerable nuisance when measuring off channel performance of receivers.

The 8643A is a big, heavy and complex beast with many special parts.

This series was replaced with the PSG series, which are still made.

Regards,

Alwyn

_____________________________________________________

Alwyn Seeds, Director
114 Beaufort Street (Management) Company Ltd.,
114 Beaufort Street,
London,
SW3 6BU

114 Beaufort Street (Management) Company Ltd., Registered in England and Wales: No. 02797775 Registered Office Address: 114 Beaufort Street, London, SW3 6BU

______________________________________________________

Dear All,

it rather depends what frequency range you need to cover and the offset from carrier at which you need the phase noise to be low.

Also, remember that signal generators have amplitude noise as well as phase noise and this is of particular concern at large offsets (> 50 kHz) from carrier.

The HP8643A and its higher frequency derivatives formed the replacement for the HP8640 cavity tuned signal generator and are particularly good on non-harmonic spurii.

Many DDS based generators have non-harmonic spurii at -70 dBc levels, which are a considerable nuisance when measuring off channel performance of receivers.

The 8643A is a big, heavy and complex beast with many special parts.

This series was replaced with the PSG series, which are still made.

Regards,

Alwyn

_____________________________________________________

Alwyn Seeds, Director
114 Beaufort Street (Management) Company Ltd.,
114 Beaufort Street,
London,
SW3 6BU

114 Beaufort Street (Management) Company Ltd., Registered in England and Wales: No. 02797775 Registered Office Address: 114 Beaufort Street, London, SW3 6BU

______________________________________________________

?

does it have to be HP only?

None of those older HP models can beat the Aeroflex SGD or SGA and they are very light weight and with tons of features
but very hard to find...I missed a couple of SGD at a good price because i didnt have the money to buy...

then there is R&S like SMIQ with OCXO option, again their PN is better than 8664/8665 and although they are very heavy
it is still not as heavy as 8664/8665 I think...also they have tons of features and modulations that 86xx never had and they are not that expensive these days...

Hi All,

I recently acquired a 4195a w/o any of the cables.? I thought about making up a set, but instead found this:

A bit on the spendy side, but look to be well built.? Not affiliated with the seller, but can tell you it's honest and good.? FWIW.??

Sadly, there is still work to be done on my 4195a as it appears to not be responding to any signals.

Ryan N7QJ

On Sunday, December 29, 2019, 02:07:22 AM PST, Tam Hanna <tamhan@...> wrote:

Hello,

sorry for being a bit slow in responding here. But please, count me in. I also have a HP 4195A which "smells" sometimes - my wife and I did not have the time to look at it in detail yet.

But I can remember that there is one PCB inside which is said to have problems with capacitors. If you take a look inside and find something, let me know!

Tam

With best regards
Tam HANNA 

Enjoy electronics? Join 15k7 other followers by visiting the Crazy Electronics Lab at  

On 2019. 12. 26. 5:42, John Pease via Groups.Io wrote:

On Wed, Dec 25, 2019 at 11:40 AM, Dave McGuire wrote:

On 12/25/19 2:35 PM, John Pease via Groups.Io wrote:

Hi Dave,
I saw your post about the 50 pin 4195A cables. I was going to make them
myself with ribbon cable and IDC connectors, but saw this option on line:



They seem to be well made and reasonably priced. I think the 2 foot
length is good.?

Are there any gotchas?

No gotchas; these are standard Amp Champ connectors; "25 pair" cables
in the telco world. Just be sure to get the correct gender. I use a
two-footer on my 4195A and it's fine.

-Dave

--
Dave McGuire, AK4HZ
New Kensington, PA

?Thanks Dave,
It will be fun to get this unit working again. I pulled it out of the bin at work. I powered it up and smelled something burning so I have a bit of work to do.
I really liked the 4195A? and always wanted one. The price was right!!

Over the last few years, I acquired 2 4396A's which I like. Have you run across a CLIP for these?

John

On Sat, Dec 28, 2019 at 11:38 PM Keith Monahan <keith@...> wrote:

What I need is to buy one of those 100bt cards and tone out the connections
off the 21143 and the FX2. Or a schematic would work for that card. Or a
PCI labeled pinout on the FX2 connector for the main board.

It's crossed my mind to drop a converter board or short cable that goes FX2
to white 32-bit PCI........... because why not?

Glen: if you know anyone selling a card reasonably.....

If you are just curious to reverse engineer the PCI bus pinout of the
100-pin Hirose FX2 connectors you could start with the AMCC PCI
MatchMaker S5933Q on the PCI board. From a quick look at a photo of
the PCI board the side with pins 1 - 40 of the S5933Q is directly
adjacent to the 100-pin connector and several traces can be visually
followed between the S5933Q and the 100-pin connector on the top layer
of the PCB. Tracing out all of the connections between the two would
take a bit of time, but not too bad.

Too bad the 1FC3-0004 Dino GSC-to-PCI Bridge External Reference
Specification (ERS) document doesn't include a pinout for that chip to
trace the connections from that end.

I don't currently have a dead 16700B to part out for the 10/100BT
daughter board. Seems like it would be a shame to part out a fully
working 16700B, but then again maybe individual parts could eventually
be sold for for a total greater more than a complete mainframe.