Sinus 1watt PLL

[sinus trouble]

The carrier noise on that type of circuit is very low. You have to make sure that your power supply is clean (a cheap Chinese wall wort won't give good results!). If you use a 3-terminal voltage regulator to feed the board, be certain to fit bypass capacitors (10n - 100n) to both the input and the output, both down to ground. I used to use the 7815 regulator, and actually solder 100n SM capacitors to its legs - this keeps RF out of the regulator, and considerably quietens things down.

Another trick I used frequently was a "gyrator" circuit (just a single transistor one) in the varicap bias supply, and sometimes in the supply to the Oscillator. This also reduces carrier noise.

I also used to supply the Vcc3 pin on the SAA1057 with 30V, to allow a wide tuning swing. I could (just) get rid of the trimmer capacitor in the oscillator, and have the circuit tune right across the band. Stephen Moss and I tried out that approach, and we were considering it for the next version of the NRG board. We derived the 30V from a simple oscillator and rectifier circuit (I've put it up here before), and it provided a worthwhile improvement. The next version of the PLL PRO was also going to have a more sophisticated phase comparator (but still made from discrete ICs). I have a nifty little circuit that I've used for years, which uses a 74HCT74 and a 74HCT00, and gives a really reliable locked / unlocked indication too. It has much lower phase noise than the 74HC86 EXOR gate that Stephen used for years.

Albert is correct!

The carrier noise is very low! Infact external audio sources (Cables, Mixer Etc) caused more noise than the exciter itself!

Also! Now that you mention the Vcc3 pin? With the oscillator trimmer centred (6.5V tuning voltage) and a supply voltage of 15V! It is possible to tune the full band without touching the oscillator trimmer!!

An unexpected lucky outcome!

However, this method leaves uncomfortable margins for frequency errors! I prefer some headroom, which the trimmer provides when set at 5V tuning voltage!

[Albert H]

On the NRG PLL PRO III the trimmer was adjusted so that the voltage at the emitter of TR1 was 3.2 Volts. This ensured a fast lock-up when switched on and minimum distortion. I came up with a simple circuit with just 1 IC, driven from the 5V regulator on the board, and you just had to tweak the trimmer (when locked) until both LEDs went out.

It would be trivial to modify the circuit, to drive it from the 9V regulator on your board, and indicate your optimum 5.0V....

Here's the circuit:

Tuning_Indicator_Sinus_PLL.png

[sinus trouble]

On the NRG PLL PRO III the trimmer was adjusted so that the voltage at the emitter of TR1 was 3.2 Volts. This ensured a fast lock-up when switched on and minimum distortion. I came up with a simple circuit with just 1 IC, driven from the 5V regulator on the board, and you just had to tweak the trimmer (when locked) until both LEDs went out.

It would be trivial to modify the circuit, to drive it from the 9V regulator on your board, and indicate your optimum 5.0V....

Here's the circuit:
Tuning_Indicator_Sinus_PLL.png

Correct again Albert!

The 3.2V you are referring to is the optimum voltage for a single varicap to produce audio quality whilst keeping track of tuning simultaneously!

However, The "Sinus PLL" uses dual varicaps! The audio side is biased @ 3.5V to 3.7V via a potential divider!

The second varicap is independent of the audio and serves as a tuning purpose only!

From what i have observed? Tuning can be anywhere between 1V and 14V (supply voltage limited) without degrading the audio quality significantly!

[Albert H]

Whoops! Not paying attention. Of course you're separating modulation from tuning.

[reverend]

3.5V seems a little low to me. It's sometimes easy to forget that Varicap diodes are still diodes and, when forward biased, will rectify any signal passed to them. Assuming that the oscillator is producing a few volts peak-to-peak then the varicaps will rectify this and produce a bias voltage of their own. On the Rev FM exciters, if you disconnect the feed from the PLL, and measure the voltage as rectified by the varicaps, this is in the region of 3 Volts. As such, if you tried to bias the varicaps below this level, they would conduct for much of the time. This produces all manner of spurious outputs and definitely does not lead to clean modulation.

The simple test is to remove the PLL or modulation feed from the varicaps and see what voltage is produced without any feed. You might be surprised!

[sinus trouble]

Whoops! Not paying attention. Of course you're separating modulation from tuning.

No worries Albert!

I am sure you have much more important things to do than study my Projects!

[sinus trouble]

3.5V seems a little low to me. It's sometimes easy to forget that Varicap diodes are still diodes and, when forward biased, will rectify any signal passed to them. Assuming that the oscillator is producing a few volts peak-to-peak then the varicaps will rectify this and produce a bias voltage of their own. On the Rev FM exciters, if you disconnect the feed from the PLL, and measure the voltage as rectified by the varicaps, this is in the region of 3 Volts. As such, if you tried to bias the varicaps below this level, they would conduct for much of the time. This produces all manner of spurious outputs and definitely does not lead to clean modulation.

The simple test is to remove the PLL or modulation feed from the varicaps and see what voltage is produced without any feed. You might be surprised!

Very interesting Rev!

I never really thought of it that way? Certainly something i will look into!

From initial testing? I found the varicaps to be seriously unstable locking and audio below 2V bias!

Above 3V, stability increased but the audio was rich in high frequency and kinda splattery?? If that makes sense?

Stability and audio quality increased dramatically approaching 4V! Beyond that, audio seemed to degrade rapidly!

[sinus trouble]

Also just to clarify if i am on the same understanding as you Rev?

The Varicaps are reversed biased in respect to the PLL tuning voltage? Yet forward biased by the oscillator!

Therefore the differential between these two potential voltages will determine how the varicaps perform?

Correct me if i am way off?

[Albert H]

No - you're right. I made an oscillator that ended up with the varicap bias being down around 1.9V, and the oscillation of the oscillator applied a voltage swing of about 2V. The mess that came out of the oscillator was truly horrible!

[jvok]

Another thing you'll see with too low bias is the output level starts to drop off. This is because when the diodes go into forward bias they'll start to load the oscillator down. Some oscillators will stop completely of you don't have at least 1-2v reverse bias.

[jvok]

Side note - the circuit with two varactors in reverse series (cathodes tied together and to dc bias input) is better for this than the single varactor circuit because at least one of the diodes will always be reverse biased.

[sinus trouble]

Some excellent points Jvok!

Most other "On frequency" designs i have looked at, The Varicaps are tied to ground unlike the NRG doubler oscillator?

Using dual Varicaps theoretically would present the oscillator with an equal load from an AC perspective?

But yeh, I put a lot of time into getting the audio Varicap potential divider right for best response!

[Albert H]

The whole point of that oscillator circuit is to get a really clean first harmonic signal from it. To that end, it has to be accurately balanced, with identical value components on each side. (This point is completely missed by Hollings in the "Veronica" rip-off rubbish - he radically unbalanced the oscillator, and got huge amounts of ½f, 1½f and further harmonics, with little or no attempt to get rid of them.)

The beauty of the oscillator is that it will produce lots of very clean Band II from a ½f beginning, without the need for critically tuned doubler stages. The six 2-turn coils have to have a reasonably high Q for best results (again missed by Hollings - he's now using low-Q Surface-Mounted Inductors (in a misguided attempt to look "up to the minute")). You can make the circuit using one, tapped, 12-turn coil (some of the originals were done that way), but the hassle of tapping every 2-turns makes that largely pointless.

[jvok]

The NRG circuit still has the same advantage of the usual dual varactor circuit. One of the diodes will always be reverse biased no matter how high or what polarity the RF is. So you never get a direct short across the tank coil. The worst that will happen if one of the diodes gets forward biased is you get a path to ground through the 6k8 feed resistor. This is still a high enough resistance that it won't load the oscillator very much, so you still get high Q.

[rigmo]

This can solve all?
https://www.facebook.com/k19radio/posts ... Q8w11cCm5l

[sinus trouble]

This can solve all?
https://www.facebook.com/k19radio/posts ... Q8w11cCm5l

Nice oscillator Rigmo

However, We are talking about the behaviour of varicaps in AC/DC situations!

Each have their merits and downfalls!

The "Sinus PLL" is not perfect? But functions the way it was intended!

[rigmo]

The "Sinus PLL" is not perfect? But functions the way it was intended!

please, the term evolution, redesign.. something.. stings my eyes, when I see it I have nightmares. total repulsion. to be GOOD, it must also be beautiful..
Albert? can be a smaller, smd version be made??? it's like a dinosaur! it sucks terribly.

[jvok]

Brutal...

[rigmo]

https://www.hamtronik.com/2016/06/veromini-lcd.html
much smaller...

[rigmo]

Brutal...

Yes thank you for eyes... to large
what i find is for coils is
L1.1, L1.2, L1.3, L1.4, L1.5, L1.6 27nH (1210-27n)