FM PLL using mixer as prescaler

Hi guys,
i know that this thread is really really old with a cpl of new comments. i have a cpl of comments with using a mixer instead of a prescaler.

in harrys pll pages he said that the pll uses a vco and a phase comparitor in a loop. these components are integrators with gain. this makes a multivibrator so the loop gain has to be reduced to stop the pll constantly swinging up and down the band. the prescaler is also a filter and gain-damper so if the prescaler is swapped then i can see lotsa problems.

the loop gain will increase and it could oscillate
the prescaler will give less fm modulation filtering in the loop

these mean that the design of the pll will have to change. am I wrong? is my thinking sound? i know i am not in the same league as ivan and harry but i didnt see this mentioned in the thread.
lots of fantastic reading here in old posts. been sick for a cpl of months so not been a round.
/D

Hi...since you state that AM IS working, all things considered, it is the prescaler IC.

Basically the manner in which these PLL's work is, they contrast a separated down VCO contribution with a reference recurrence. (the reference is gotten from the XTAL). The yield of this correlation is a mistake voltage that creates a PWM (like) yield that takes care of the charge siphon where it is sifted and taken care of to the Vt. It's an exceptionally high impedance circuit and (it's actual) that a broken varactor diode can makes the Vt be undulated, making it difficult to tune.

As fas as not getting the genuine help manual, check for a Carver tuner circuit, or an Adcom GTP-500 (I think) tuner schematic. Toshiba made the reference plan that most finished result makers followed perfectly.

Ruud wrote:In the days that I was building some transmitters for pirate stations, I used a 'reverse AFC.
This worked as follows. A normal receiver AFC corrects the drift in a receiver oscillator, the transmitter frequency remains constant. The correction voltage is derived from a (10.7 MHz.) discriminator circuit.
I reversed this principle: I used a crystal oscillator that produced a signal 10.7 MHz. above (or below) the desired transmitter frequency. This was mixed with a sample of the transmitter frequency and (again) a 10.7 MHz discriminator circuit delivered the AFC voltage, in this case not for the receiver oscillator, but for the oscillator in the transmitter!
This was before the PLL circuits became popular (grandpa is telling stories...), but at the time it was a great way to build a stable FM transmitter!

THAT was interesting!
A rather clever idea?

>grandpa is telling stories...

 Please continue, grandpa. Such stories are always an inspiration 

BR Harry - SM0VPO

Startek has a Fujitsu prescalar: https://www.ebay.com/itm/280924085134?rmvSB=true

Aliexpress.com also has several prescalars.

73
KB7BB

In the days that I was building some transmitters for pirate stations, I used a 'reverse AFC.
This worked as follows. A normal receiver AFC corrects the drift in a receiver oscillator, the transmitter frequency remains constant. The correction voltage is derived from a (10.7 MHz.) discriminator circuit.
I reversed this principle: I used a crystal oscillator that produced a signal 10.7 MHz. above (or below) the desired transmitter frequency. This was mixed with a sample of the transmitter frequency and (again) a 10.7 MHz discriminator circuit delivered the AFC voltage, in this case not for the receiver oscillator, but for the oscillator in the transmitter!
This was before the PLL circuits became popular (grandpa is telling stories...), but at the time it was a great way to build a stable FM transmitter!

Prescalers were used to:
1 - simplify syntheser construction
2 - allow synthesisers to work at much higher frequencies
3 - give a wide frequency-range

Many older designs, using a relatively small synthesiser frequency range, used conversion rather than down-conversion and I worked on several 27MHz CB units that did just that. I remember one particular design published in PW (UK) (or was it wireless world?) that used a CD4001 as a post mixer amplifier and filter, using the low-frequency responce of the logic gate.

The first practical sysntesiser I built used an 8Ω loudspeaker as a microphone, positioned over a (286) PC speaker to generate the reference frequency. The PLL divide was fixed and the reference frequency was varied, using the PLL as a fixed frequency multiplier. The reference frequency was generated using GW-BASIC command (sound <freq>,<duration> but low <duration> value made it infinite duration). With this method you could have a very wide frequency range since the reference could be from something like 100Hz to 32.768kHz (? from memory ;-).

The first systhesiser I worked on professionally was several banks of crystal oscillators at 100kHz intervals, 1MHz intervals, and 10MHz intervals. There was also a reference oscillator and tuned circuits selected the correct harmonic before mixing with the crystalk banks. This was used by the millitary (all-valve) for millitary aircraft communications (225MHz to 399.9MHz). See ARC-52 for more info. I think I still have a "guard receiver" (243MHz milliatry distress frequency) receiver, which was a self-contained receiver inside the ARC-52.



With synthesisers you can do just about anything, it only requires a little thought, with reference to what you are trying to achieve.

Very best regards from Harry - SM0VPO

joy444444 wrote:
...It eliminates the need of a prescaler since they are hard to find these days. Let me know your opinion on the schematic. A single BC 547 has been used as a mixer to bring the VCO frequency down to 12 MHz...

If you look at the schematics of frequency synthesizer in the early days of PLL, you see it was the norm to use mixing instead of a prescaler:
http://www.rigpix.com/otherusefulinfo/pll02amfl.htm

-

It is even possible to build the frequency synthesizer with a virtual stepsize of approx. 100Hz - build of two frequency synthesizers (frequency-reference; stepsize approx 10kHz: 10MHz/995 and 10MHz/1005 - or 10.1MHz/1000 and 10.1MHz/1010) and a crystal oscillator for the last down mixing (the loading of the web.archive.org backup page can take some time - please be patient):
https://web.archive.org/web/20170809140205/http://www.pe1jpd.nl/index.php/70-mhz-all-mode-trx/
Quote: "...
The mc145145 can only handle frequencies up to some 15 MHz, so the frequencies were mixed down. Also, this way I can use two pll’s with a relatively high reference frequency of 10 kHz to get to a stepsize of only 100 Hz! This is how this works:
..."

Frequency synthesizer schematic:
https://web.archive.org/web/20180127225924/http://www.pe1jpd.nl/wp-content/uploads/2015/07/70MHzPLL.jpg

Joy, 

Interesting idea to use a mixer rather than a divider, may be useful for a PLL based VFO to get small steps without resorting to dual modulus. I think I have seen a circuit somewhere, probably in the 'Technical Topics' section of Practical Wireless magazine, where a TBA120 is used to lock a VHF oscillator to a free running oscillator running at about 2-3MHz. From what I can remember the circuit was quite simple.

Regards,
John

Got it Harry. I'll experiment with the mixer and pll. One more question. See the attached schematic. The coil matching the oscillator output to 50 ohm is 270 nH. It's reactance is 80 ohm at 50 MHz. Is the tapping wrong? It's 8T coil.. shouldn't the tapping be at 6T from ground for 50 ohm match? In the schematic it's at 2T from ground. Please clarify it for me.

Hi Joy,:
I see absolutely no reason why it should fail. It should work perfectly :-)

The only point to remember tis that there is no signal divider, so high frequency modulation will still be compared with the reference. This means that your loop filter will need to have a very long time-constant (greater than 5ms).

You will also find that a "normal" PLL is exactly phase-locked, but in this case it may slip a few RF cycles. You will get a good "frequency lock" but not a precise "phase lock".

As regards frequency stability then this is absolutely not a problem. I did something similar with the PLL FMTX I designed for Kits-R-Us.

BR Harry

Hi Harry,

Do you think this circuit will work? I haven’t built it yet.
It eliminates the need of a prescaler since they are hard to find these days. Let me know your opinion on the schematic. A single BC 547 has been used as a mixer to bring the VCO frequency down to 12 MHz. I’ve ordered 80 MHz can oscillator from eBay. The 12 MHz output from the mixer may need additional filtering and/or amplification.