HP 8568B FREQ ZERO Adj?

[bench_knob]

Hi Y'all,

Newbie owner or a (very sweet) HP-8568B spectrum analyzer, am running the "Initial Power On and Calibration" proceedure as described in "Programming and Operation Manual".  Hah!

Re; 'Recall 9', step 6, does any one know what is meant by, "Maximize response with FREQ ZERO adjustment", while another calibration document says, "Adjust for maximum signal amplitude".

Turning the FREQ ZERO adjustment changes two things on the display, 1) a slowly (10 sec sweep) sweeping line is drawn on a gratical line at the very bottom, ie, a flat horizontal line or 2) about 7 lines up, and if adjusted somewhere in the lower middle, it draws a gently sloping upward line, the only other thing that changes on the display, 3) is a dB amplitude readout number in upper right of the display, which jumps between -17dB and -10.7dB.

What is expected to be adjusted?  Line or number? The document isn't clear enough for my understanding of things. I've looked through all of the documents that I have, and I have not been able to discern what is intended.

Any help is appreciated.

TIA

bench_knob

[MarkPalmer]

The signal display and amplitude readout should correspond to one another, I would use the readout in this instance.  You want it set to be as close to your -10.00 dB calibrator input signal as you can adjust it.  The .7 dBm loss you see can be from your cable, I wouldn't worry too much about it, you just want to adjust to the maximum amplitude.  Ironically, this adjustment was deleted from the display sections of later production 856X analyzers. 

You'll enjoy using the 8568B.  These analyzers are very powerful RF measuring tools. 

-Mark-
   

[G0HZU]

I believe that [RECALL] [9] is a factory preset to configure the analyser such that you can align/peak up the response of the analyser when in 'zero span' mode.

It selects a low RBW setting (30Hz?) and by adjusting the FREQ ZERO preset on the front panel of the display unit you are effectively tuning for the very peak of the 30Hz RBW filter when in zero span mode where the analyser isn't sweeping but is acting as a fixed tuned receiver. You need the analyser to be aligned with the RBW filter to get the best from this mode and this is one reason for this test.

So you are supposed to tune for max line level on the display. i.e. tune it so the horizontal sweeping CRT trace is as high as possible on the display.
Note:

Although it appears that you are adjusting for amplitude in this test, you are really aligning the 'frequency' to hit the true peak of a very narrow RBW filter.

Otherwise zero span mode will be missing the true peak of the RBW filter.

The formal 'Calibration Procedure' should be summarised on the lower pullout tray that summarises all the key functions and gives basic user info about how the analyser operates and how to get started with it. It also lists and describes all the SHIFT key functions.

i.e. look to see if your analyser has the (optional?) pullout info trays under the RF unit. These pull or slide out with a visible tab and are about A3 sized plastic sheets with all the info printed on them.

These are well worth studying

[G0HZU]

Note: The usual procedure is to calibrate the analyser to -10.00dBm using the internal -10dBm CAL OUTPUT signal and [RECALL] [8] first. Then do the [RECALL] [9] adjustment for max peak in zero span mode

There's also a neat and fully automated 'Error Correction' routine you can run that takes about a minute to run by pressing:
[SHIFT] [FREQUENCY SPAN]

You also have to have your short RF cable connected from the CAL OUTPUT to the main RF input ((N type) connector when doing this error correction/test routine.

[bench_knob]

@MarkPalmer & G0HZU,
 
 Thank you gents for the good info....is very much appreciated.
 
 I guessed and chose to adjust the horizontal trace 'line' to
 highest graticle setting, as you say it exhibited a fairly
 'sharp' behavior. I'll go back and see what the dB readout
 shows. If the readout isn't too close to the -10dB calibrator
 then its likely that something else may need calibrating?
 
 All in all the SA appears to be running pretty well, however
 I notice that at times the FREQ COUNTER mode seems to
 jump around quite a bit, maybe its ok, but I'm watching
 it. Unfortunately I have little experience working with
 spectrum-analyzers, so I don't yet have an understanding
 of what's normal and what is not.  I have never been able
 afford one of these tools and to gain the opportunity to
 acquire such a beautiful instrument, well, all I can say
 is WOW! I know several folks who say that the 8568 and the
 8566 may well be the all-time gold-standard of analyzers.
 However that may be, I am very happy with my acquisition
 although I fully anticipated having to expend additional
 time and money restoring it to its original functionality.
 I also bought an unused 'new' old-stock CRT for it. We'll
 see if that is indeed true. :L
 
 @G0HZU,
 
 The pull-out trays are not present; I had planned at some juncture to ask
 the EEVblog community if someone has or might consider scanning
 those sheets so that I get them into a pdf file for perusal. 
 
 The pullout sheets are selling on eBay for the nearly the same
 money I expended on the RF section. 
 
 Thanks again
 
 bench knob
 
 
 

[G0HZU]

Maybe I should sell my HP8568B pullout sheets on ebay 

I can have a go at capturing them with a digital camera but none of my cameras are that great. I'll have a go tomorrow.

They are in excellent condition because they have always been stored separate from the analyser either at my place of work (since about 1990) or since I bought the analyser from the company many years ago. IMO they look a bit unsightly when fitted under the analyser so I keep them against a wall behind a desk.

Your HP8568B looks to be very clean and tidy with a sharp CRT. Sadly mine has seen a lot of use over the years and the CRT has now lost some quality wrt the graticule focus. A real shame but it is still very useable

[alex.forencich]

I have heard that the 8566 is supposed to be excellent as well.  However, I have never seen a comprehensive comparison of all of the old HP analyzers, and they have several different units from different eras that cover the same frequency ranges.  I presume there are slight differences in performance.  Do you know if there is any significant difference between the 8566 and the more portable 8563? 

[G0HZU]

Do you know if there is any significant difference between the 8566 and the more portable 8563?

I also have an 8566 here. At my place of work there are several HP8563 variants but I've never directly compared the two side by side. Apart from the obvious size differences the 8563 has a few advantages.

The 8566 covers 0-22GHz in two separate ranges. 0-2.5GHz and 2-22GHz.
So you can't, for example show a span of 0-10GHz on an 8566. But you can on the 26GHz 8563.

Also, the phase noise of the 8566 isn't as good as the 8563. But the 8566 is still better than many modern analysers in this respect.

The 8566 has fairly modest linearity performance in the 0-2.5GHz range because of the mixer it uses. eg the mixer IIP3 is typically only +8 to +10dBm and is spec'd at something like +5dBm. The 2HI performance is fairly modest as well.

However, the big old 8566 does have a very flat gain response on the 0-2.5GHz range and also a very consistent (flat) noise floor across this range and very low VSWR with the 10dB attenuation setting selected. Truly remarkable in this respect.

I believe there are some reliability issues with the 8563 wrt the YIG and also they can suffer from intermittent front panel button membranes making some of the front panel buttons a bit deaf. This can make the analyser a real chore to use.

So there isn't that much between them overall. However, I think I would prefer to have the 8563 (especially the 8563E) because of the portability and it doesn't have the full range broken into two separate bands and also the fact that it isn't as noisy in terms of fan noise . The 8566 and 8568 are very, very noisy in this respect. It's fine in a big open lab but the constant fan noise can get very tiring and annoying in a small room at home. The DSP back end of the HP8563E also offers better quality RBW filters and faster sweeping on narrow spans. But they cost a bit more to buy compared to a typical 8566.

The EC versions of the 8563 have a modern LCD display that replaces the CRT and some people prefer this. Some of our early 8560E and 8563E analysers with CRTs at work are showing signs of CRT ageing and they are beginning to look a bit tired and dated when compared to modern analysers. So I'd expect the prices of these to drop quite a bit in the next few years.

[MarkPalmer]

As the 856X series analyzers age and companies are getting more concerned about recycling, more of these units are thankfully getting in to the hands of hobbyists through E-Bay and the like rather than being tossed in dumpsters.  I picked up one of the 8567A analyzers last year along with its corresponding 8444A tracking generator.  This model is the orphan in the 856X line, basically an 8568B without the low RBW filters that sold for a lower price.  It’s still an excellent analyzer none the less and I got it for next to nothing.  The previous owner kept it very clean, it looks and works like it was made yesterday.  I also have one of the 3585A low frequency analyzers to compliment it.   



-Mark-

[G0HZU]

I had a go at taking some images of the pullout trays for my 8568B today.
Sadly my camera is pretty ancient so the quality is not great.
See attached jpgs.

They should be 5Mpixel (and approx 2Mb in size) and I've uploaded them to the photobucket site. Photobucket gets more and more hideous each time I use it so I can't guarantee you will be able to get the full res images from photobucket but I can guarantee their site will annoy you a lot with adverts.

[G0HZU]

After enviously looking at the CRT quality of b_k and ap's 856x analysers I had a go at tweaking mine internally and had some very good results!

I found some dodgy resistors on the Z circuits and also adjusted the CRT controls internally.

The result is that Ican now get a very, very sharp and fine graticule once more

See the attached image of my 8566 CRT.

It looks a bit gloomy but that is the only way I can get the camera to capture the image without spoiling the sharpness. It looks much better than the image in reality.

Before, the CRT was very fuzzy and it had a very thick and bloomy graticule and the trace was dim.

Now it looks as sharp as the other analysers on here. I'll get some images of my 8568 CRT next...

[bench_knob]

G0HZU,

The tray photos are good....thank you so much!! 

Your tweaking of the CRT circuit looks pretty good too. Did you do that recently? 

I'm curious...what is that gadget (in your photo-Can gallery) thats disassembled and laid out horizontally with a spring part? 

Born of curiosity, do you happen to know what are the differences between the 8568 and the 8566 models?  Can one swap a few modules to upgrade (hah) an 8568 to 8566? Some osc's, processor board and maybe front-panel, type mods?  I'm sure if it were that simple or even doable folks would be talking about it..

Interestingly, in my search for an HP8568A/B range extension method (via InterNet) I discovered a small US company that provides a 'kit' of sorts that one hacks into an 8568 SA to increase the analyzer analysis range up to 22GHz. Surprisingly it turns out that I worked several years with that chap a couple of years ago at a military GPS company.  Should anyone be interested, he's written a book about L2C. About six of us engineered the L2C control-segment GPS DSP receiver together. Eventually when L2C 'Civil-Tracker' is turned on, our receiver will be used to 'control' the SVs in orbit. Just as an aside...

http://www.dkdinst.com/products/DC100description.html

I desire to uprange my SA, as I am fiddling around with 2.4GHz ISM band xcvrs for a fancy weather-station
that I've been hacking together these past couple of years. While I did get the analyzer because...well I wanna learn how to use these tools. At the GPS company we had several SAs but they were continually in use by the RF guys and I really never had any quality usage time with these instruments.     

Thanks for the photos.

bench_knob
=====================


Quote from: G0HZU on September 16, 2014, 02:14:51 AM

I had a go at taking some images of the pullout trays for my 8568B today.
Sadly my camera is pretty ancient so the quality is not great.
See attached jpgs.

They should be 5Mpixel (and approx 2Mb in size) and I've uploaded them to the photobucket site. Photobucket gets more and more hideous each time I use it so I can't guarantee you will be able to get the full res images from photobucket but I can guarantee their site will annoy you a lot with adverts.

[alex.forencich]

The components to extend the range are likely proprietary HP microwave compoents - mixers, RF switches, etc.  And probably a YIG tuned filter.  That's the case on the 8590 series analyzers, anyway - the only differences between the models are the RF section, presence of the counterlock board, and which version of the analog interface board is installed.  The sub-2.9 GHz units have a completely different RF front end and cannot be upgraded.  2.9 GHz analyzers have just the low band.  Everything above 2.9 GHz have an additional high band that uses a different signal path and several additional components.  You would basically have to pull the parts out of another analyzer. 

[G0HZU]

Your tweaking of the CRT circuit looks pretty good too. Did you do that recently? 

Yes, very recently. I did it after seeing how good your CRT looked and I was fed up seeing so many other people with decent CRT traces on their HP856x analysers 

So I dived in and found a few issues and made some adjustments and the results are very good as you can see! I now think that my 8566 has a new CRT because when I compare it to my 8568B the CRT in the 8566 looks spotless inside with none of the classic dust and soot anywhere. It also looks to have a new PSU section at the rear because the parts look new. It was an ex rental machine and I'm now guessing that the CRT was replaced not long before the analyser was retired and sold to me

Born of curiosity, do you happen to know what are the differences between the 8568 and the 8566 models?  Can one swap a few modules to upgrade (hah) an 8568 to 8566? Some osc's, processor board and maybe front-panel, type mods?  I'm sure if it were that simple or even doable folks would be talking about it..

I think the IF/Display unit is the same but the RF units are very different inside. Your best bet would be to just buy an RF unit from a HP8566B as these can be bought quite cheaply now although you would have to test it out thoroughly.

I'm curious...what is that gadget (in your photo-Can gallery) thats disassembled and laid out horizontally with a spring part? 

It's part of the insides of this rather ancient (and very large!) R&S 2GHz 140dB switched attenuator. My first ever bit of Rohde and Schwarz test gear 

I bought a couple of them a few weeks ago for a low price and (once stripped and serviced) these attenuators give remarkable performance up to 2GHz. Definitely lab grade

[bench_knob]

Hi there alex.forencich,
 
 Thanks for the nice informative reply.
 
 Hah, in my defense however, I knew it was hee-haw question, ok?    + having taken a peek-see inside my 8568 RF section, well its a complicated gadget inside there and the 8566 is no different in that regard, I've seen photos of their interiors too.  However, after seeing the gadget that my ex-workmate builds and installs (DK Inst) into 8568 models, well it still seems possible that module swapping might be doable using only 8566 HP parts instead. As for pulling parts out of an existing 8566 RF section, well there are bnches of 8566 modules for sale on eBay for $50ea.  DK Inst just replaces one card to facilitate the range upgrade and he uses the existing internal RF relay to switch range bands.  So that seems to suggest that most of the components between the two models (HP8568 & HP8566) are likely the same parts!!!  <sigh> 

Cheers,
 
bench_knob
 
 
 
 ===============
 
 Quote from: alex.forencich on Today at 06:37:40 AM

The components to extend the range are likely proprietary HP microwave compoents - mixers, RF switches, etc.  And probably a YIG tuned filter.  That's the case on the 8590 series analyzers, anyway - the only differences between the models are the RF section, presence of the counterlock board, and which version of the analog interface board is installed.  The sub-2.9 GHz units have a completely different RF front end and cannot be upgraded.  2.9 GHz analyzers have just the low band.  Everything above 2.9 GHz have an additional high band that uses a different signal path and several additional components.  You would basically have to pull the parts out of another analyzer.

[bench_knob]

G0HZU,

Your tube looks good and bright.  Although you say mine looks ok, its a bit dim I darken my lab lights to view it and the intensity knob is 75% rotation, 100% rotation offers no additional (that I can perceive) brightness.

I'm swapping out the CRT today with a 'new' tube that I bought on eBay to test and verify that it actually works.  I wanna see if the 'new' tube imparts any substantial visual improvement. Since eye-balling & memory is so subjective, I'm gonna setup a test screen display using the same intensity adjustment setting along with before & after photos (tripod) + using an ExeTech lux photometer measurement to provide info for a quantitative comparison.


>>G0HZU said
> ...think the IF/Display unit is the same...

I beleive that this is true.

So, what did you find amiss in the R&S module that you had disassemble and better how did you discern it needed fixing?

And now that HP seems to be spiraling down into the bad-managers can (like Danaher-Tektronix) is doing, i your opinion, how do you rate R&S in comparison with the HP of olden days?

I've included a few photos of the HP 2090-0225 - Panasonic 170XB31 CRT for y'alls perusal..

 bench_knob
 
 ===============

Quote from: G0HZU on Today at 07:58:06 AM

Quote

Your tweaking of the CRT circuit looks pretty good too. Did you do that recently? 

Yes, very recently. I did it after seeing how good your CRT looked and I was fed up seeing so many other people with decent CRT traces on their HP856x analysers 

So I dived in and found a few issues and made some adjustments and the results are very good as you can see! I now think that my 8566 has a new CRT because when I compare it to my 8568B the CRT in the 8566 looks spotless inside with none of the classic dust and soot anywhere. It also looks to have a new PSU section at the rear because the parts look new. It was an ex rental machine and I'm now guessing that the CRT was replaced not long before the analyser was retired and sold to me
Quote

Born of curiosity, do you happen to know what are the differences between the 8568 and the 8566 models?  Can one swap a few modules to upgrade (hah) an 8568 to 8566? Some osc's, processor board and maybe front-panel, type mods?  I'm sure if it were that simple or even doable folks would be talking about it..

I think the IF/Display unit is the same but the RF units are very different inside. Your best bet would be to just buy an RF unit from a HP8566B as these can be bought quite cheaply now although you would have to test it out thoroughly.
Quote

I'm curious...what is that gadget (in your photo-Can gallery) thats disassembled and laid out horizontally with a spring part? 

It's part of the insides of this rather ancient (and very large!) R&S 2GHz 140dB switched attenuator. My first ever bit of Rohde and Schwarz test gear 

I bought a couple of them a few weeks ago for a low price and (once stripped and serviced) these attenuators give remarkable performance up to 2GHz. Definitely lab grade

[MarkPalmer]

There were two types of CRT's used in this series of analyzers, the earlier one made by HP and later ones sourced from Panasonic.  The two tubes don't directly interchange as the power supply potentials differ.  Just something to be aware of when you see CRT's available for these, one needs to know which one their analyzer has.

These tubes have become very expensive.  There are a few places making color TFT display replacements for it, but here again, big $.

The CRT's, especially the Panasonic one, respond very well to rejuvenation from CRT tester/rejuvenators like the Sencore CR70.  They will give a dim tube back much of its brightness and crisp focus, and around another 500 hours of life.  That might not be much for 24/7 lab use, but for a hobbyist only running the analyzer occasionally, it can be many years of additional use from the CRT, so I wouldn't toss a tube until it runs at least one rejuvenation cycle. 

-Mark-

[G0HZU]

So, what did you find amiss in the R&S module that you had disassemble and better how did you discern it needed fixing?

I tested it on a VNA up to 3GHz and it had a few issues with some of the attenuator sections. Sometimes they didn't select properly. Also the various contact surfaces needed cleaning because there were a few subtle dips in the frequency response when in the 0dB/thru position.

It now has about 0.6dB insertion loss at 1GHz in the 0dB position which is very good indeed. The accuracy of the attenuator is remarkable and these must have cost a lot of money when new.

And now that HP seems to be spiraling down into the bad-managers can (like Danaher-Tektronix) is doing, i your opinion, how do you rate R&S in comparison with the HP of olden days?

I generally just use RF test gear and (at my place of work) I'm happy to use any of the big names like HP or R&S or Anritsu. R&S gear tends to be more solidly built compared to anything else and it usually has a performance edge but this often comes at the expense of size, power and weight and cost.

[bench_knob]

Quote from: MarkPalmer on September 19, 2014, 11:18:34 PM>

There were two types of CRT's used in this series of analyzers, the earlier one made by HP and later ones sourced from Panasonic.  The two tubes don't directly interchange as the power supply potentials differ.  Just something to be aware of when you see CRT's available for these, one needs to know which one their analyzer has.

These tubes have become very expensive.  There are a few places making color TFT display replacements for it, but here again, big $.

-Mark-

Mark,

Hah!  I am now the proud owner of a perfect brand-spank'n new...wrong type tube, heh heh, but its in really good shape. I must've had cobwebs in my brain to have missed the fact that the tube in my spectrum analyzer (obviously) has the deflection plate pins sticking out the side of the neck.



So, I'll wait awhile and then put it up for sale. Somebody out there would like to have a new bright tube in their scope and I'll let it go for the same price I paid for it, we are all in this together.

bench_knob

[wybren1971]

Hello G0HZU,

i saw you're pictures in the forum for the pullout trays for the 8568B. I recently am the proud owner of this kind of spectrum analyzer. But mine came without the pullout trays. I can see the pictures on photobucket but when i download them they are only 130Kb jpg and the resolution isnt good enough for me to read them.
Could you please mail me the originals? or do a repost?

Thank for the trouble..

Greetings Wybren

[PaulAm]

Ko4bb has scans in this directory:

http://www.ko4bb.com/getsimple/index.php?id=manuals&dir=HP_Agilent/HP_8566_Spectrum_Analyzer/HP_8566A_Reference_Card

[0culus]

Those are the 8566 reference cards. The 8568(A) ones are here: http://www.ko4bb.com/getsimple/index.php?id=manuals&dir=HP_Agilent/HP_8568_Spectrum_Analyzer

They should work for the B front panel. I didn't see scans of the ones that would have come with a B model.