UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Other Discussions > Homebrew Equipment

Notices

Homebrew Equipment A place to show, design and discuss the weird and wonderful electronic creations from the hands of individual members.

Closed Thread
 
Thread Tools
Old 2nd Dec 2022, 7:10 pm   #321
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

The 400MHz Marconi 2382 has a 3Hz RBW filter. I'm not sure if this is achieved using crystals or not. My old Advantest TR4172 analyser has the 7Hz RBW option fitted, and this is achieved with crystals. The HP8566/8 analysers use crystals for the narrower RBW filters and the narrowest has a RBW of 10Hz. I've not seen a 1Hz RBW filter made using crystals but maybe someone has managed it.

In practice, all the narrowest (crystal based) RBW filters (100Hz, 30Hz, 10Hz, 7Hz) are a bit ropey in terms of amplitude and frequency response so the best analysers will have an internal automated cal routine that is able to measure and correct for the errors in these filters. The operator usually has to connect a cable from a cal port to the RF input of the analyser and then run the automated routine. It normally takes a minute or two to complete. The accuracy and repeatability of the analyser should be much improved after this, especially when swapping between RBW settings. Usually, the analyser stores the cal corrections in NVRAM, so it isn't necessary to perform the auto cal routine every time the analyser is used.

Analysers with a digital IF don't need to do this anymore. However, my E4440A PSA analyser does have a suite of analogue preselector filters just ahead of the ADC. These are there to minimise/optimise the 'staring' bandwidth of the ADC. If they weren't there, the ADC would always stare at a 10MHz BW and this could make it easy to overload.

The E4440A has numerous LC and crystal preselector BPFs ahead of the ADC. These still need to be calibrated/checked for their response and the analyser does this at bootup and also every so often it pauses and does it during normal use. This can be quite annoying, but it does preserve the excellent performance of this analyser.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 2nd Dec 2022, 7:50 pm   #322
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: 6-gang FM stereo tuner heads

There's a 1Hz RBW in the 8590A, and that uses crystals in its 100kHz IF. "RF" range is only 0-50kH, though.

The 859x analyser family have narrow (below 1kHz) RBWs done with a mix-down to a lower IF. 1kHz and wider is crystal, then LC at 21.4 MHz. Up-down keys will get you to 3MHz. But you can type in 5MHz and get it. Shape is not specified above 3MHz, but it's useful.

The higher performance 856x family can only go to 1MHz. They too switch in an extra conversion for the narrowest bandwidths, but later on, they were changed so the very narrowest IFs put the signal directly into what was otherwise the ADC after the log detector, and did the narrowest bandwidths as well as logging in DSP. This went on into the ESA family. I used the ESA as the basis of the noise figure analyser, but it has a completely different IF hardware and all RBW is DSP, aimed at 4MHz bandwidth for standard Y-factor measurements, but subsequent DSP can take it down to 100kHz if needed, but noise factor gets a bit slow to average down there. It was intended to be able to measure multiple 100kHz buckets at once and speed up measurement, but this didn't make it to the final product chiefly because of software development time for having to handle separate cal factors for each bucket. It was super-critical that the different IF gain stages had steps which combined arithmetically without combinatorial errors which were a limitation in the old 8970. The NFA has an extra mix-down from 21.4 to 6.25MHz. Partly, this helps with the ADC/DSP, but it also breaks up a lot of switched gain all at 20MHz in the 8970, and it's leakage around this chain which spoils the arithmetic combination of the gain steps.

One whole lotta fun (apologies to Robert Plant)

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 2nd Dec 2022, 10:26 pm   #323
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

The classic old HP8566/8 analysers did all the crystal filtering up at 21.4MHz and this includes the 10Hz RBW filter. Because of the ropey nature of the narrowest crystal filters in the 8566/8 (drift and poor amplitude response up at 21.4MHz) the next generation of analysers (eg HP8560A) only attempted to do crystal filtering down to 300Hz RBW. This was done at the third IF at 10.7MHz.

For RBW below this, the narrowest RBW filters were done in the digital domain. This improved the performance and consistency a LOT. The later HP 8560E and 8563E supported a 1Hz RBW filter in the digital domain.

It's several years now since I last used an HP8560 or 8563 in either the A or E versions. At work, these analysers have mostly been sold or scrapped. At one time there were dozens of them, and I used them on a daily basis. I can remember when the first HP8560A arrived at work and this must have been in the early 1990s. I spotted a dead 26GHz HP 8563EC sat sulking under a bench at work last week. It has a sweep fault and I suspect it will go in the WEEE skip soon. I think it is the last one left in the engineering labs.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 2nd Dec 2022, 10:51 pm   #324
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Going from the extreme low RBW of 1Hz to the opposite spectrum, the YIG-tuned bandpass pre-selector (HP8566 and HP 8411) is like the powerful kryptonite crystal in Superman comics. It can tune much wider than the widest RBW IF and keep the IMD out of the mixer filters.

The YIG oscillator is the pinnacle of RF voodoo, like the microwave ferrite circular, very cool device.
Here is a reverse engineering S21 and power measurement of a RF power amp, it contains a circular which works like traffic going around a circular junction in RF amplifier at 9.30 minutes:

https://youtu.be/9AfuPAmymUU

That guy got all his mobile network base stations and RF amps from the skids.

Quote:
t's several years now since I last used an HP8560 or 8563 in either the A or E versions. At work, these analysers have mostly been sold or scrapped. At one time there were dozens of them, and I used them on a daily basis. I can remember when the first HP8560A arrived at work and this must have been in the early 1990s. I spotted a dead 26GHz HP 8563EC sat sulking under a bench at work last week. It has a sweep fault and I suspect it will go in the WEEE skip soon. I think it is the last one left in the engineering labs.
Many RF devices contain so much silver and gold that they are normally sent to these people to recycle them(duplexers). Zero health and safety protocol:

https://youtu.be/fTOaYsOzh1o
Attached Thumbnails
Click image for larger version

Name:	YIG tubed bandpass preselector HP8411.jpg
Views:	26
Size:	35.5 KB
ID:	269193   Click image for larger version

Name:	YIG tuned filter HP8566.jpg
Views:	22
Size:	65.5 KB
ID:	269194  

Last edited by regenfreak; 2nd Dec 2022 at 11:15 pm.
regenfreak is offline  
Old 2nd Dec 2022, 11:21 pm   #325
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

I doubt the dead HP8563EC will end up like that. Whoever collects the WEEE stuff from the company would probably sell it on to someone else who would then repair it or use it as a spare parts donor.

Yes, YIG filters are a bit special. One of the earliest downconverters at work had a YIG filter in the front end. I think it tuned from 500MHz to 2GHz. It required a lot of power to heat it up to a controlled temperature and we also had to write temperature compensation lookup tables for it vs frequency. The YIG filter also had some tuning hysteresis depending on tuning direction. It also took several milliseconds to tune to a new frequency and this was its main downfall at work. It was too slow to tune so it was the last time we used one. However, nothing could compete with it in terms of filter bandwidth and shape factor. Nothing else came remotely close to the YIG filter in this respect.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Dec 2022, 1:09 pm   #326
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

I had a rummage in an old boneyard box of engineering samples, and I still have one of the early YIG filters. See below. This must be about 30 years old. This should be a 500-2000MHz YIG filter and I think it was about 15MHz wide. I'm not sure if it works though, it might be a dud.

The HP 8566A/B has a PRESEL PEAK button on it that allows the user to optimise the 2-22GHz YIG tuning at a given frequency. However, in my experience pressing this button can cause problems as it can cause dips in the YIG tuning at other frequencies. There is a button sequence that restores the factory default YIG calibration, and the default YIG cal works really well on mine. I try and avoid using the PRESEL PEAK button.

I've dug out my HP 8568B today and switched it on. It must be the first time I've used it in a few years. With the internal correction turned off, the various crystal RBW filters show a slight frequency error and also some amplitude errors (+/- 1.2dB between them), and some are not symmetric anymore. I do have the calibration supplement to the service manual, and this describes how to correct the LC and crystal RBW filters. The last cal sticker on it is dated 2006 and that must be about the year I bought it from the company. I'll have a go at aligning the filters today as I have some free time. It's too cold to go outside!
Attached Thumbnails
Click image for larger version

Name:	YIG.jpg
Views:	32
Size:	30.4 KB
ID:	269250  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Dec 2022, 1:13 pm   #327
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by Radio Wrangler View Post
Spectrum analyser crystal filter design is somewhat contorted to making programmable bandwidths and synchronous tuning shape (not actually Gaussian, though close enough up top)

David Ford's gang did the 3585A over at Loveland, but he'd been at the 'ferry for some time. I think my write-up of the simpler spectrum analyser filter set (in the 3724A) is in the HPJ article (I no longer look like a hairier Harry Potter).

The higher freq analysers use PIN diodes to control stage Q and hence make variable bandwidth.

With so many resistors floating around, loss == noise floor takes a hit.

David
HP did not openly publish the schematics in the service manuals of their spectrum analyzer. I was wondering how variable RBW could be done with crystals.

Conventional filters with Butterworth or Chebyshev filter topology have a rather long settling time in the transient response. The attached example(source:https://2n3904blog.com/butterworth-filter/)
illustrates the time response of a Butterworth filter of nth order. There is overshoot in the damped oscillation making them a deal breaker for spectrum analyzer application in particular RBW.

The bandwidth of a crystal can be affected by the input and output impedance, and the coupling coefficient k (capacitive or inductive). The variable k can be achieved by the use of variable coupling transformers or varactor diodes.The 2nd attachment shows the variable bandwidth crystal filter developed by Telefubken in the 1930s. The crystals are isolated by buffer amplifiers and the tuning is achieved by a pair of LC tanks at the input and output of crystal:

http://lucafusari.altervista.org/pag...er/Filter.html

The third attachment shows a variable bandwidth crystal filter for a homebrew spectrum analyzer. The load impedance is adjusted by a PIN diode which acts like a variable resistor:

http://lea.hamradio.si/~s53mv/spectana/sa.html

Quote:
I had a rummage in an old boneyard box of engineering samples, and I still have one of the early YIG filters. See below. This must be about 30 years old. This should be a 500-2000MHz YIG filter and I think it was about 15MHz wide. I'm not sure if it works though, it might be a dud.

The HP 8566A/B has a PRESEL PEAK button on it that allows the user to optimise the 2-22GHz YIG tuning at a given frequency. However, in my experience pressing this button can cause problems as it can cause dips in the YIG tuning at other frequencies. There is a button sequence that restores the factory default YIG calibration, and the default YIG cal works really well on mine. I try and avoid using the PRESEL PEAK button.

I've dug out my HP 8568B today and switched it on. It must be the first time I've used it in a few years. With the internal correction turned off, the various crystal RBW filters show a slight frequency error and also some amplitude errors (+/- 1.2dB between them), and some are not symmetric anymore. I do have the calibration supplement to the service manual, and this describes how to correct the LC and crystal RBW filters. The last cal sticker on it is dated 2006 and that must be about the year I bought it from the company. I'll have a go at aligning the filters today as I have some free time. It's too cold to go outside!
We are posting at the same time again.
Wish you luck with the calibration. The process looks complicated in the service manuals!

If you find the datasheet any of the YIG filters, you may be able to power it up. I saw this guy powered up a HP YIG tuned oscillator:

https://youtu.be/NTtPTJQGLLY

It seems the YIG tunable oscillators are used by the military often. There are tons of them on ebay.
Attached Thumbnails
Click image for larger version

Name:	Butterworth overshoot settling time.jpg
Views:	30
Size:	49.8 KB
ID:	269251   Click image for larger version

Name:	Kauter crystal filrers.jpg
Views:	38
Size:	70.2 KB
ID:	269252   Click image for larger version

Name:	homebrew RBW tunable crystal filter.jpg
Views:	38
Size:	72.0 KB
ID:	269253  

Last edited by regenfreak; 3rd Dec 2022 at 1:33 pm.
regenfreak is offline  
Old 3rd Dec 2022, 1:21 pm   #328
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

The crystal RBW circuits in the HP 8566/8 are very similar to your third image, except a BJT is used in place of the phase inverting transformer. The compensation trimmer should help with filter symmetry, and I think that section is what needs tweaking in some places in my HP 8568B. The cal procedure is quite detailed and lengthy. I've got the gear here to do it all, but it might take me a long time...
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 3rd Dec 2022, 3:32 pm   #329
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by G0HZU_JMR View Post
The crystal RBW circuits in the HP 8566/8 are very similar to your third image, except a BJT is used in place of the phase inverting transformer. The compensation trimmer should help with filter symmetry, and I think that section is what needs tweaking in some places in my HP 8568B. The cal procedure is quite detailed and lengthy. I've got the gear here to do it all, but it might take me a long time...
I have found the HP8566 crystal RBW schematic attached. It is like a maze inside, an enigma machine.

I need to man up and go outside in spite of the cold now. Enjoy your cal.
Attached Thumbnails
Click image for larger version

Name:	HP8566B crystal filter.jpg
Views:	38
Size:	56.6 KB
ID:	269261  
regenfreak is offline  
Old 3rd Dec 2022, 11:51 pm   #330
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

I had a go at the HP8568B filter alignment with mixed results. Looking through the cal/adjustment manual, I'd overlooked that within the 21.4MHz IF, the analyser briefly mixes down to 3MHz and back up again to 21.4MHz to do the very narrowest crystal filters. The rest of the crystal filters and the LC filters are done up at 21.4MHz

There are crystal filters in various modules to adjust and some special RC jumper links are needed to bypass the crystals during the adjustment process.

I've aligned the 21.4MHz crystal filters and achieved good symmetry, but I now have the 3MHz crystal filters to do. Hopefully, I'll be able to centre the 10Hz and 30Hz RBW responses on the CRT display. These filters are already symmetrical so I hope this will be easy. I've also worked through the logamp and video processor adjustment section, and this has improved the display accuracy.

I'll have to revisit the 21.4MHz crystal LC filters after doing the 3MHz filters as I think it's best to do it all in the correct sequence. So, lots left to do.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 4th Dec 2022, 1:21 am   #331
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by G0HZU_JMR View Post
I had a go at the HP8568B filter alignment with mixed results. Looking through the cal/adjustment manual, I'd overlooked that within the 21.4MHz IF, the analyser briefly mixes down to 3MHz and back up again to 21.4MHz to do the very narrowest crystal filters. The rest of the crystal filters and the LC filters are done up at 21.4MHz

There are crystal filters in various modules to adjust and some special RC jumper links are needed to bypass the crystals during the adjustment process.

I've aligned the 21.4MHz crystal filters and achieved good symmetry, but I now have the 3MHz crystal filters to do. Hopefully, I'll be able to centre the 10Hz and 30Hz RBW responses on the CRT display. These filters are already symmetrical so I hope this will be easy. I've also worked through the logamp and video processor adjustment section, and this has improved the display accuracy.

I'll have to revisit the 21.4MHz crystal LC filters after doing the 3MHz filters as I think it's best to do it all in the correct sequence. So, lots left to do.

It seems the 3MHz filters' bandwidth is controlled by the simple switching on and off the 5 crystals (attached)?
Attached Thumbnails
Click image for larger version

Name:	HP 8566B 5 pole crystal RBW.jpg
Views:	33
Size:	83.5 KB
ID:	269278  
regenfreak is offline  
Old 4th Dec 2022, 11:03 am   #332
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

I can see a pattern. Every crystal reduces the RBW by about a third. So a cascade of 7 poles of crystals may bring it close to 1 Hz RBW. I havent found an example of 1Hz RBW filter schematic yet.
regenfreak is offline  
Old 4th Dec 2022, 12:47 pm   #333
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

I think the 5Hz to 50kHz spectrum analyser that David was referring to earlier might be the HP3580A. This audio spectrum analyser dates back to 1973.

The notes in the link below suggest it has a 1Hz RBW.

https://www.hpmemoryproject.org/wb_p..._b_page_04.htm

If you scroll down the link a bit further, you can see some stuff about the HP8568A. Back in 1978, this 1500MHz spectrum analyser really represented a huge step forward in terms of RF performance and features and it could be fully controlled via GPIB in an ATE system. I think a lot of the 8568 analysers were purchased for use in ATE racks. The high purchase price was presumably worth it, because it was able to speed up production testing and improve the quality/repeatability of the testing.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 4th Dec 2022, 1:28 pm   #334
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

Back in the early 1990s my HP8568B analyser was the company's gold standard analyser. In those days the company had a design team that did a lot of work with satellite modems. This meant that the company had to be able to measure noise power very accurately at 70MHz in order to do BER curve testing.

This HP8568B went to HP every year for a basic calibration check but I don't think the RBW filters or the logamp would have been adjusted in any of these visits to HP. The cal costs would have been huge if this was requested.

Due to the limitations of the logamp and the analogue RBW filters and the internal 10dB and 20dB gain stages, it was possible to see a subtle variation in noise power (using the 1Hz noise marker) depending on RBW setting and reference level and attenuation setting. This obviously had an impact on the BER testing so the company adopted a neat solution in order to achieve repeatable noise measurements with this analyser.

The two tubular 70MHz BPFs below were purchased from Telonic in the USA and these were designed to be very rugged and stable with very low VSWR at 70MHz. One is about 12MHz wide and the other is 3MHz wide. The filters were then sent to the National Physics Laboratory where the noise bandwidth of the filters was measured very accurately. For example, the narrow filter is stamped 66.12dBHz.

A precision high-level 70MHz noise source was then passed through the 70MHz filter and the filter output was then fed to a thermocouple power meter. The power meter might read -21.3dBm for the average power of the noise. The power meter is able to measure this very accurately.

The filter is then connected to the spectrum analyser with the 1Hz noise marker enabled. If correctly calibrated, the analyser should show -21.3 - 66.12dBHz = -87.42dBm/Hz on the noise marker.

You can see how old and worn the filters look today. The 3MHz version shows a lot of handling wear on the BNC connectors as it was the favoured filter. These filters were used a lot back then. I borrow them from work now and again to play with. I've got them here now and I have my own precision noisecom noise source to use with them. When used with the Agilent E4440A analyser, the results for the 3MHz filter are still spot on today. The 3MHz wide filter was the favoured choice as it was easier to maintain noise flatness across a 3MHz BW. However, my Noisecom based noise source is very flat across 2MHz to 180MHz so I can use either filter.
Attached Thumbnails
Click image for larger version

Name:	Telonic 70MHz BPFs.jpg
Views:	40
Size:	44.2 KB
ID:	269309  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 4th Dec 2022, 6:04 pm   #335
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by G0HZU_JMR View Post
I think the 5Hz to 50kHz spectrum analyser that David was referring to earlier might be the HP3580A. This audio spectrum analyser dates back to 1973.

The notes in the link below suggest it has a 1Hz RBW.

https://www.hpmemoryproject.org/wb_p..._b_page_04.htm

If you scroll down the link a bit further, you can see some stuff about the HP8568A. Back in 1978, this 1500MHz spectrum analyser really represented a huge step forward in terms of RF performance and features and it could be fully controlled via GPIB in an ATE system. I think a lot of the 8568 analysers were purchased for use in ATE racks. The high purchase price was presumably worth it, because it was able to speed up production testing and improve the quality/repeatability of the testing.
Thanks, I have found the schematic. The HP3580A has 5 pole crystals switching on and off RBW down to 1Hz. The RBW does not change with temperature. The crystals were individually matched for the temp coefficient and the inflection points with slope = 0 of their temperature curves.

Quote:
The two tubular 70MHz BPFs below were purchased from Telonic in the USA and these were designed to be very rugged and stable with very low VSWR at 70MHz. One is about 12MHz wide and the other is 3MHz wide. The filters were then sent to the National Physics Laboratory where the noise bandwidth of the filters was measured very accurately. For example, the narrow filter is stamped 66.12dBHz.
At work we have a 50-year-old, big wind tunnel from the National Physics Laboratory NPL.The fan motor is still going strong but the transducers of the 6-axis force balance are out of calibrations and no one at work knows how to recalibrate them.

Quote:
I have my own precision noisecom noise source to use with them.
The calibrated noise source costs an arm and a leg. Even the cheaper alternative with calibrated ENR values costs a few hundred pounds. Without the noise source with known ENR, it is very hard to measure accurately an amplifier with low NF and low gain.
Attached Thumbnails
Click image for larger version

Name:	HP3580A.jpg
Views:	31
Size:	41.7 KB
ID:	269327   Click image for larger version

Name:	HP3580A 5-pole crystal.jpg
Views:	41
Size:	64.1 KB
ID:	269328   Click image for larger version

Name:	HP3580A 1Hz.jpg
Views:	33
Size:	128.5 KB
ID:	269329  

Last edited by regenfreak; 4th Dec 2022 at 6:29 pm.
regenfreak is offline  
Old 5th Dec 2022, 12:24 am   #336
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

I did manage to finish the RBW filters in the HP8568B late last night and it's been on all day on a soak test. Here's an image of it showing a 5kHz span.

Sadly, my mobile phone can't capture how crisp and sharp the CRT is now. I also tweaked the CRT when I had the covers off. It looks quite vibrant and really sharp now. This analyser was at its very best when looking at the close to carrier phase noise on VHF and UHF frequency synthesisers as it has very low close to carrier phase noise on a 5kHz span like this.

I had a go at making the TSC-2-1 splitter this afternoon. I added the RC network, and it does give about 70dB isolation at 100MHz. However, the isolation null is very termination sensitive so it would have to be adjusted for peak isolation with the DUT attached at the sum port.
Attached Thumbnails
Click image for larger version

Name:	HP8568B__30Hz.jpg
Views:	40
Size:	49.1 KB
ID:	269377  
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 5th Dec 2022, 3:42 am   #337
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 22,800
Default Re: 6-gang FM stereo tuner heads

The HP3708A is a high power noise source targetted on 70 and 140MHz with noise-bandwidth defining filters AND a thermal power meter all in the same box. It's used for carrier-to-noise ratio testing of telecoms demodulators, but it covers 20-200MHz if you do your own filters (BNCs for external filters) but for the telecomms market it's all 75 Ohms. Its party trick is that if you feed a clean signal into it, it measures true RMS power and will add noise to create a programmable signal-to-noise ratio. Quite handy for checking spectrum analyser IF filter NBWs. It isn't a thermocouple power meter, it's actually a matched pair of resistor-diode power sensors with signal on one side and DC on the other... the DC servo controlled to make the diode Vbe's match and track.

Blackest magic is involved. The thing is sufficiently fast and accurate, that with parameters measured under favourable conditions, it can change a noise level so fast that you cannot measure the result with low enough uncertainty to prove it works in spec!

The top box of the 8566 is the same as the top box of the 8568, so a lot of the final IF stuff applies to both. I did check earlier, the RBW knob on my 3580A does go down to 1Hz. The 8566B ans 8568B have a lot faster microprocessor system tan the A versions and this solved a few issues where the thinking machinery would run out of steam. There was some general modernisation and tidying up in some of the RF sections.

The R&D project which resulted in the 8566/8568 pair was called "The Doomsday Box" The 8590 project was "Hornet" and its ESA replacement "Mosquito". The NFA nicked the ESA cabinet, Rx front-end, CPU, etc. of Mosquito and was itself called "Bluefin" The matching smart noise source being "Wasabi". A huge 'portable' cell base tester was 'Dagger' and the software suite for it was 'Cloak' They didn't make it through to production simply too heavy and expensive. Would have been easier to carry the base station to a test facility.

We had fun with project names.

Loveland was an earlier HP division and had an unusually wide remit whch they retained. There was a bit of treading on each other's toes with the RF and telecomms divisions, with a bit of (almost) concealed heat. John Wastle's lifestory write-up is in the HP memory project web area and tells a lot of the business of interdivisional relationships and empire-building. I know the real names behind some (most) of the characters given aliases in it. I also share John's opinions in most cases. and could add a couple more.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline  
Old 5th Dec 2022, 9:40 pm   #338
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Radio Wrangler the top box of the 8566 is the same as the top box of the 8568, so a lot of the final IF stuff applies to both. I did check earlier, the RBW knob on my 3580A does go down to 1Hz. The 8566B ans 8568B have a lot faster microprocessor system tan the A versions and this solved a few issues where the thinking machinery would run out of steam. There was some general modernisation and tidying up in some of the RF sections.

The R&D project which resulted in the 8566/8568 pair was called "The Doomsday Box" The 8590 project was "Hornet" and its ESA replacement "Mosquito". The NFA nicked the ESA cabinet, Rx front-end, CPU, etc. of Mosquito and was itself called "Bluefin" The matching smart noise source being "Wasabi". A huge 'portable' cell base tester was 'Dagger' and the software suite for it was 'Cloak' They didn't make it through to production simply too heavy and expensive. Would have been easier to carry the base station to a test facility.

We had fun with project names.

Loveland was an earlier HP division and had an unusually wide remit whch they retained. There was a bit of treading on each other's toes with the RF and telecomms divisions, with a bit of (almost) concealed heat. John Wastle's lifestory write-up is in the HP memory project web area and tells a lot of the business of interdivisional relationships and empire-building. I know the real names behind some (most) of the characters given aliases in it. I also share John's opinions in most cases. and could add a couple more.
I always associate Loveland with free love and hippies. They smoke psychedelic herbs and chant along 1Hz binaural beats by the campfire.

In the world of aircraft, "Hornet" = F-18; Delta "Dagger" = F102, "Mosquitos" = de Havilland DH.98 fighter planes.

The HP3708A can do so many things in a box; temperature compensation, self-calibration, generation of thermal noise of Gaussian distribution, flat-top noise over a range of frequencies..etc

The circuit of the HP8566's 10Hz crystal RBW filter is similar to the that of HP3580A 1Hz RBW. The crystals may not behave themselves at 1Hz RBW at RF frequencies.

The crystals may age over long period of time when the spectrum analyzer was in storage without on and off thermal cycles. In addition, the passive and active components, as well as the solder joints are susceptible to aging so the spectrum analyzer would need re-calibration after long term storage. Probably dusts, siders and moisture can accelerate aging process. I would imagine that if one crystal goes bad, it has to be replaced from an equivalent crystal in a donator's board.

Quote:
G0HZU_JMR I did manage to finish the RBW filters in the HP8568B late last night and it's been on all day on a soak test. Here's an image of it showing a 5kHz span.

I always associate

Sadly, my mobile phone can't capture how crisp and sharp the CRT is now. I also tweaked the CRT when I had the covers off. It looks quite vibrant and really sharp now. This analyser was at its very best when looking at the close to carrier phase noise on VHF and UHF frequency synthesisers as it has very low close to carrier phase noise on a 5kHz span like this.

I had a go at making the TSC-2-1 splitter this afternoon. I added the RC network, and it does give about 70dB isolation at 100MHz. However, the isolation null is very termination sensitive so it would have to be adjusted for peak isolation with the DUT attached at the sum port.
Cool, perfect symmetry!

I have no idea how your RC network work improve the TSC-2-1 port isolation..I think my chinese-made combiner is fine for 100MHz range.

I am yet to construct the spectrum analyzer protector based on your design photo. I bought some 125mA SMD fuse (very expensive!) and 1N4148W SMD diodes. My 1KV 100nf and 10nF 1210 SMA capacitors have arrived today after a few weeks. There will be two low pass filters based around the diode junction capacitance of the four 1N4148W, and also the stray inductance of the fuse so the cut-off frequency may be around 200MHz. I can see you use a loop of wire of low Q and inductance to form the first LP. The 125mA fast fuse has resistance 1.6 ohms and my DE-5000 struggles to measure its inductance due to its low Q. I will fly home to see my parents on this Wednesday so there will be a few weeks that I wont be experimenting with nerdy electronic projects. It got to be iterative guess work to find the component values of the two LPF using the NanoVNA.

Last edited by regenfreak; 5th Dec 2022 at 9:51 pm.
regenfreak is offline  
Old 6th Dec 2022, 12:08 am   #339
G0HZU_JMR
Dekatron
 
Join Date: Sep 2010
Location: Cheltenham, Gloucestershire, UK.
Posts: 3,077
Default Re: 6-gang FM stereo tuner heads

That 3708A looks like a neat and versatile instrument. Amongst its tricks it looks like can do a similar thing with bandpass filters, a power meter and a noise source.

Quote:
I have no idea how your RC network work improve the TSC-2-1 port isolation..I think my chinese-made combiner is fine for 100MHz range.
I can't remember what is inside a TSC-2-1 but I think it may be a magic tee combiner with a 25R to 50R matching transformer at the sum port.

If so, there will be a 100R resistor between the input ports A and B. If there was nothing else but the resistor inside the box (not even a sum port connection) then there would be a 6dB loss between port A and B with 0deg phase shift. This is what you would see with only a resistor in the box and nothing else.

However, there is obviously other stuff in the combiner. If you were to remove the 100R resistor and just leave an ideal version of the other stuff there including the correctly terminated sum port then you would also see a 6dB loss from A to B, but this time there would be a 180deg phase shift.

When you add the 100R resistor then these two A to B leakage paths cancel as they are in anti-phase. Therefore, an ideal combiner would have perfect isolation across ports A and B.

In reality the transformers inside the magic tee combiner won't be perfect. The 100R resistor will usually be fairly close to perfect. However, there will be a sweet spot external termination impedance at the sum port that can re-establish high isolation across the A and B ports. It's just a case of changing the sum port termination slightly away from the normal termination by using a small amount of resistance and reactance. This only works over a narrow frequency range.
__________________
Regards, Jeremy G0HZU
G0HZU_JMR is offline  
Old 6th Dec 2022, 1:44 pm   #340
regenfreak
Heptode
 
Join Date: Nov 2018
Location: London SW16, UK.
Posts: 655
Default Re: 6-gang FM stereo tuner heads

Quote:
Originally Posted by G0HZU_JMR View Post
That 3708A looks like a neat and versatile instrument. Amongst its tricks it looks like can do a similar thing with bandpass filters, a power meter and a noise source.

Quote:
I have no idea how your RC network work improve the TSC-2-1 port isolation..I think my chinese-made combiner is fine for 100MHz range.
I can't remember what is inside a TSC-2-1 but I think it may be a magic tee combiner with a 25R to 50R matching transformer at the sum port.

If so, there will be a 100R resistor between the input ports A and B. If there was nothing else but the resistor inside the box (not even a sum port connection) then there would be a 6dB loss between port A and B with 0deg phase shift. This is what you would see with only a resistor in the box and nothing else.

However, there is obviously other stuff in the combiner. If you were to remove the 100R resistor and just leave an ideal version of the other stuff there including the correctly terminated sum port then you would also see a 6dB loss from A to B, but this time there would be a 180deg phase shift.

When you add the 100R resistor then these two A to B leakage paths cancel as they are in anti-phase. Therefore, an ideal combiner would have perfect isolation across ports A and B.

In reality the transformers inside the magic tee combiner won't be perfect. The 100R resistor will usually be fairly close to perfect. However, there will be a sweet spot external termination impedance at the sum port that can re-establish high isolation across the A and B ports. It's just a case of changing the sum port termination slightly away from the normal termination by using a small amount of resistance and reactance. This only works over a narrow frequency range.
I see. It is one of those things that you have to understand how it works and add modification to get it working better at a certain frequency. The phasing part sounds complicated.
regenfreak is offline  
Closed Thread

Thread Tools



All times are GMT +1. The time now is 9:19 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2024, vBulletin Solutions, Inc.
Copyright ©2002 - 2023, Paul Stenning.