Best AM FM Stereo RF Generator for a Newbie

[George S.]

Been playing with and investigating the SDRplay RSP1A receiver and RSP Spectrum Analyzer software.
Results are extremely promising. I am using a 60 dB SMA rf attenuator between the Panasonic AM/FM stereo signal generator and receiver. This is a must.
The receiver is dead on accurate frequency wise with the SDRuno software when scanning for beacons and time signal transmitters on the long wire antenna. Frequency accuracy is yet to be determined with the spectrum analyzer software. Looks like the Panasonic may be appx 300 Hz out of calibration. We'll find out.
I've watched the Panasonic's output drift appx 6 Hz as the unit warmed up this morning.
The photo shows the 19 kHz pilot and 38 kHz harmonic. First time I've been able to actually see these first hand.

 

[George S.]

The Panasonic is 150 Hz off frequency AM and FM.
Made it into a transmitter with a antenna.
Checked calibration of the SDR with the Canadian Time Station CHU, and WWV in Boulder CO. It's right on the money.
Time to hunt up a service manual and get it on frequency.
https://en.wikipedia.org/wiki/CHU_(radio_station)

 

[George S.]

Found the manual. Output frequency is supposed to be within + or - 100 Hz. Easy enough to run through the procedure.
Would be nice to print out the manual, but at 347 pages, isn't going to happen. Sure is nice that someone took the huge effort to scan it into a PDF.
Having a working spectrum analyzer is key. Now I can see and learn about the individual components that make up a FM Stereo transmission and do better alignments of old test equipment.

 

[mlucitt]

I am looking for a spectrum analyzer now that you have placed me on that slippery slope. I will probably invest a ridiculous amount of money in test equipment to do the alignment on one Phase Linear 5000. But hey, that's what we do around here...

I like the HP 8560EC or the 8561EC. I really need the LCD display because there is so much data and not much room. The green monochrome CRT displays are just not as sharp or clear to me.

 

[George S.]

Mark L., I've been pursuing a HP or Advantest SA for over 2 years. I watch every other day and read about them by searching the model name followed by "forums". This gives the results from the Electrical Engineering forums where these guys have vast experience with them. Many comments about certain models are eye opening. Many of the units are now degraded by time and not fixable, or are so specialized as to not work for general purpose.
My solution for audio is QuantAsylum, and SDRplay for RF.
Would love to have a nice HP SA, but can't justify the thousands of $. And I'd really need two plus those expensive attenuators.
The 60 dB SMA attenuator I was using on the SDR is $11 and is accurate from DC to 8GHz.

 

[George S.]

Oh, and a tracking generator can be added to the SDRplay for SA use, it's directly supported. It's Arduino based. Something I've no experience with. Another aspect I'm going to pursue.
But, I'm back at work. Heading to a chemical plant down river from Pittsburgh. So much to do and so little free time.

 

[George S.]

Here's some interesting reading from HP.



The SDRplay SA manual says:
PPM Trim may be used to correct for any slight frequency differences between radios. Resolution
is 0.001 PPM. With a 24MHz crystal, +/- 1PPM will equal a +/- 24Hz error.
Not bad at all, especially for what I'll be using it for. 50% chance It'll be within 12 Hz of accuracy I'm thinking.
Now to read about Arduino. Seems they're using a Analog Devices chip to create the tracking generator signal.

 

[George S.]

Been playing with the RSP Spectrum Analyser. It's amazing! Very easy to calibrate.
Forgot that the HP 5334B is a 100 MHz unit so testing at 90 MHz. If the HP is on frequency, and I believe it is (it has the optional 10 MHz crystal oven), then I believe the Panasonic is off 150 Hz.
Getting there!

 

[Gepetto]

Been playing with the RSP Spectrum Analyser. It's amazing! Very easy to calibrate.
Forgot that the HP 5334B is a 100 MHz unit so testing at 90 MHz. If the HP is on frequency, and I believe it is (it has the optional 10 MHz crystal oven), then I believe the Panasonic is off 150 Hz.
Getting there!

+140.5 Hz to be exact.

 

[George S.]

Yup, meant to insert "appx" before 150 Hz as I really don't know if the HP is perfectly on frequency. I do believe it's very close
Today I think I'll look very closely at the spectra of commercial FM stations.
Should be able to calibrate the SA to them, then calibrate the Panasonic to the SA.
Not going to touch the HP until I've done some more reading about it.

 

[Gepetto]

Yup, meant to insert "appx" before 150 Hz as I really don't know if the HP is perfectly on frequency. I do believe it's very close
Today I think I'll look very closely at the spectra of commercial FM stations.
Should be able to calibrate the SA to them, then calibrate the Panasonic to the SA.
Not going to touch the HP until I've done some more reading about it.

Everything has a tolerance George

 

[George S.]

Yup. Got the Panasonic on frequency.
Going to shut everything down so it cools off, cut some some grass. Then power back up and watch it drift before buttoning it up. Pretty sure it's good. Looks like + and - 15 Hz. Specs say + and - 100 Hz.
Very touchy adjustment.

 

[George S.]

Buttoning up the Panasonic. The SDRplay was calibrated to the Hz on adjacent commercial FM stations.
The Panasonic was broadcasting through a antenna on a adjacent clear channel.
The HP shows a 100 gate count.
Time to do some reading about the HP. Couple years back I read they were attaching some sort of GPS receiver to these to create a ultra precise clock.
Also, channel A appears to have reduced sensitivity compared to channel B. May be a setting I'm not familiar with or it needs repair. Going to find out.

 

[Gepetto]

Buttoning up the Panasonic. The SDRplay was calibrated to the Hz on adjacent commercial FM stations.
The Panasonic was broadcasting through a antenna on a adjacent clear channel.
The HP shows a 100 gate count.
Time to do some reading about the HP. Couple years back I read they were attaching some sort of GPS receiver to these to create a ultra precise clock.
Also, channel A appears to have reduced sensitivity compared to channel B. May be a setting I'm not familiar with or it needs repair. Going to find out.

That's what phase locked loops are for George...

 

[George S.]

Got it Joe, had to think about it for a moment. The PLL on the tuners or receivers. It's all making better sense now.

 

[Gepetto]

Got it Joe, had to think about it for a moment. The PLL on the tuners or receivers. It's all making better sense now.

Frequency accuracy is specified in Section 15.215 (c). Intentional radiators operating under the provisions in Sections 15.217 through 15.249 must be designed so that the 20 dB bandwidth of the fundamental lobe, must be within the frequency band that operation is permitted.

Normally each channel is 200 kHz (0.2 MHz) wide, and can pass audio and subcarrier frequencies up to 100 kHz. Deviation is typically limited to 150 kHz total (±75 kHz) in order to prevent adjacent-channel interference on the band. Stations in the U.S. may go up to 10% over this limit if they use non-stereo subcarriers, increasing total modulation by 0.5% for each 1% used by the subcarriers.

 

[Gepetto]

You can see that the center frequency can be a good way off and still comply with the FCC regs.

 

[George S.]

Got it. I did check SDR calibration against several AM shortwave broadcasting time stations like WWV at 10 and 15 MHz, and CHU at 14.67 MHz.
The big domestic shortwave stations like WRMI Radio Miami International were also right on frequency, while Radio Havana Cuba seems to always be off frequency as usual.
Many foreign language stations were several hundred Hz off a nice round MHz number.
Was listening to a English broadcast of Radio Romania, and watched the AM center frequency vary and drift due I think to the distance and propagation.
Then went to the commercial FM band and looked at several strong local FM stations. All look to be right on frequency.
I am going to read up on conventional precision frequency sources.
Here is the highest resolution shot of local FM 98.5 WNCX. The center window on the right monitor.

 

[George S.]

Now pursuing a GPS based 10 MHz frequency reference for the bench. Will probably use a old Symmetricom unit. More study required.

 

[George S.]

Made a very low ball offer on this unit from a Riverside CA electronics recycler. They accepted. Former Pan Am Satellite unit. If it works or is repairable it'll be a kick ass GPSDO (GPS Disciplined Oscillator) for a 10 MHz reference standard. Looks to be highly optioned. Only downside is a lack of documentation for these units, but we'll get it figured out. Here's the listing.
https://www.ebay.com/itm/1656998469...NB_S8D7Qg2&var=&widget_ver=artemis&media=COPY