[NLRS] Thoughts on VHF-UHF beacon hardware.

[Dr. Gerald N. Johnson]

On 7/17/2015 9:46 AM, Doug Reed wrote:
>
>
> Hi Jerry, David, Lloyd.
>
> Yup, many of the newer (1980's&  90's&  up) crop of surplus commercial
> 2-way radios are synthesized. But in most of them the minimum step is
> set by the intended channel spacing. Many are limited to 5 or 6.25KHz
> spacing. Depending how old or new the radio is, they usually had no
> better than 20PPM frequency stability, although newer radios that have
> digital modes might be better than that, down to maybe 2PPM.

The 20 ppm probably mostly comes from a sloppy reference for the PLL. 
There's no way a locked PLL can wander that much long term though short 
term e.g. phase noise might allow that. At the same time the reference 
could be warped to move the frequency in smaller steps than the 
programmable minimum steps.
>
> The problem is that those radios are often more highly prized because
> they ARE synthesized. It is the crystal radios that are too expensive
> to use and they are probably free for hauling. I either have such
> radios or know where to look, and the only expense is hauling them
> home. It is the synthesized radios I want to keep and use, mostly for
> APRS, but many of them are too old to use as well.....

With the forced narrow banding last year, many a synthesized radio is 
now not usable commercially. So if it won't narrow band legally its junk.

Crystals are not low cost, but places like ICM still make them and have 
the formulas and circuit needs for many a radio.

Some of us have a large collection of crystals from various sources. I 
have many of mine in small holders sorted in little paper envelopes to 
speed up searches. I find Surplus Sales of Nebraska has hundreds of 
frequencies. http://www.surplussales.com/Crystals/Crystals-9.html
As do Mouser and Digikey. And at reasonable prices. The Mouser and 
Digikey microprocessor crystals are very cheap, and have a wide 
tolerance on frequency so may have to be adjusted or selected to get a 
precise desired frequency. Axman has a few dozen bins of crystals too. 
Mendelsons in Dayton has hundreds of frequencies, but not listed on line 
I think. With crystals in those collections being fundamental, 3rd, 5th, 
and some 7th overtone, maybe even 9th overtone the selection is 
increased by using the fundamental or a different overtone than the 
marked crystal frequency. The rub is that the crystal is sure to 
oscillate, but the frequency won't be exactly 1/3 the marked third 
overtone frequency, or the overtones won't be exactly 3, 5, 7, or 9 
times the marked fundamental frequency. They will be close not not 
exact. I have used fundamental crystals on various overtones and 
overtone crystals on different overtones successfully.
>
> I suggested using the RX front-end as a RX pre-filter and mixer for
> the SDR RX to reduce out-of-band signals or get down to the HF bands.
> I didn't think of taking the signal off after the IF filter because I
> wasn't thinking of reducing the RX bandwidth that far. But there is no
> reason not to, just have to stay ahead of the IF amp stages because
> the SDR doesn't need the extra gain or the limiting.

The limiting is the fundamental problem. The bandwidth after the filter 
would be probably at least 15 kHz in recent radios, wider in older 
radios. I don't know about modern radios but when I started on FM about 
1961, the first IF selectivity was only about a half MHz wide. Why the 
initial population of 2m repeaters was limited to 146.6 to 147.4 because 
the Motorola and GE receivers of that era would only move that far 
without realignment. They often use one crystal at different harmonics 
for both mixers. The transmitters mostly went from 146 to 148 without 
retuning. That relatively narrow first IF helped receivers work better 
in a crowded RF environment like most cities.
>
> I was going to say you couldn't step the original synthesizer output
> frequency in small enough steps for the beacon modulation, but that
> would be wrong. It might be more difficult, but I expect something
> could be done, possibly via the CTCSS input. Same with CW keying the
> middle multiplier and amplifier stages. And if you want to improve the
> PPM, I'm sure you could program an Arduino, PIC, or Ras-Pi to adjust
> the synthesizer LO to keep it on frequency against GPS timing pulses.
> That is what the Ultimate 3S does.
>
> But all of that seems like a lot of work compared to using a
> controller and replacement frequency source like the Ultimate 3S to
> generate the beacons and drive the transmitter. The PI4 beacon mode is
> a full power beacon mode where information is sent by frequency
> hopping (FSK). The PTT line from the controller just activates the PTT
> on the transmitter. There is no modification beyond figuring out where
> you want to inject the RF. The built-in software in the Ultimate 3S
> can handle various types of CW, QRSS, WSPR, and other modes. The hard
> part is figuring out how to program it to send the data you want.....

That's true and important.
>
> I keep thinking in terms of feeding the original crystal oscillator
> stage as a buffer, or perhaps feeding into one of the later multiplier
> stages so the SI5351A doesn't have to go above 50MHz to work, but that
> depends if I can adjust the PI4 frequency hops to meet specs after
> multiplying up to the final frequency..... I'm not on-off CW keying
> any of the stages, so no key clicks.

Rapid frequency changes make broad band clicks too, especially of the 
waveform is interrupted. Its best to preserve the sinewave, and change 
the slope gently at the frequency change. And abrupt change in slope of 
the waveform also makes some wide band noise.
>
> But you do have to worry about raising the noise floor near the beacon
> frequency due to phase noise. I'm sure a lot of the old-timers in the
> metro could tell you how the 6M noise floor rose after KTCA channel 2
> replaced their tube transmitter with a solid-state transmitter.... I
> don't remember if all of these oscillator modules are phase continuous
> during frequency changes, but that is another potential source of
> transmitted interference.
>
> If you use this method, any of the common oscillator modules will
> work, assuming you can improve their frequency stability and their
> phase noise doesn't trash the output signal. If you are going to write
> your own beacon controller program, that means you will be programming
> the oscillator too. If you read the hanssummers.com web site articles
> about the "ovenized SI5351A oscillator" and the "GPS stabilized
> Ultimate 3/3S transmitter" you will get some good information and
> clues to how you can achieve 10e-9 accuracy with your beacon while
> using a (relatively) garden variety AD9850/SI570/SI5351A
> oscillator..... And the various web pages and links often include
> sample code or at least flow diagrams to help you write your own code
> if you used a different CPU for the controller.
>
> Using a commercial radio as the basis for the beacon works if they
> made one for that band. So 2M and 432 bands are covered, 6M is a
> little harder. 222MHz, 902MHz, and 1296MHz are all harder because
> equipment is limited, but there are RF power modules available for all
> those bands and you can use the multiplier stages from a commercial
> rig to get to up to the band of interest.

A rig tunable to 432 drives a tripler to 1296 decently. All my 1296 
contacts were made that way. And a rig tunable to 452 ought to multiple 
to 904 nicely.
>
> That's about all I've got. I tend to think of the beacon as something
> at a remote unattended site, so ideas about control are important. I
> didn't care for the beacon-at-home idea because they usually get shut
> off during contests due to local overload, just about the time people
> want to listen to them....

I'm very slowly working on a 10 GHz beacon I'd like to put it about 
halfway between here and the cities to have a signal for testing 
potential contest roving sites. I plan on using solar panels and a 
relatively small lead acid battery for night time power. I figure on a 
micro doing the keying and monitoring the battery voltage and charging 
and to shut it down when the evening load has not quite killed the 
battery. I do plan for only CW and to key as many multiplier stages as 
possible to reduce the average power consumption. I'll have to keep 
crystal oscillator and maybe heater and a buffer or two powered up to 
prevent chirp, but when it shuts down those will be powered down too.

I kind of like PIC chips, but my assembler programming was mostly Z80 
and peripherals and the PIC chips have so many port options and 
frequency options that setting those is more work than all the assembler 
or C. I'm studying a book I picked up at Radio City on PIC chips with C 
but its only about one particular 40 pin chip and my projects could 
often use an 8 pin PIC chip.

I submitted one article to CSVHF that could be done with that simple PIC 
chip, but it needs simultaneous serial in and out. Easy to do with a 
UART, possible by bit banging with the most simple PIC so long as it has 
enough ram and rom to hold the code. It sends and receives serial CAT 
commands to make an Icom R7100 mute for transmitting periods, then 
recovers it. RS-232 like but TTL levels. Easy to program in basic or C 
in the other alternatives.

73, Jerry, K0CQ
> .
> 73, Doug Reed, N0NAS.
> ______________________________________________________________
> NLRS mailing list
> Home: http://mailman.qth.net/mailman/listinfo/nlrs
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:NLRS at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>