Where to find a good clean design / kit?

[Albert H]

Sinus - that's a pretty good take on the old NRG 5 Watt circuit, with a tacked-on PLL. I wouldn't bother with the PCB-mounted SO239 - you'd be better off bringing the RF output to the edge of the board, allowing easy connection of coax to further stages. You'd get much better performance if your upper PCB side was a groundplane - the PLL section could be fully etched, but the RF section would be better with a complete groundplane.

It would also be sensible to add a lowpass filter to the output - it would make it cleaner if you wanted to go straight into an aerial. You could also screen between stages to prevent ½f breakthrough. It might also be worth redesigning the output stage to use one of the really cheap RD06HVF FETs - you might even be able to get away with omitting the intermediate stage between the oscillator and the PA as the gain of the FET is so high. The MRF237 is getting hard to find and is incredibly expensive these days - one is going to cost as much as the rest of the parts!

The SAA1057 is an easy chip to use, but is getting hard to find. You could easily re-design the PLL to use the 5511, which would allow you to use the rig in Band I, Band II and even as high as Band IV (with some changes to the components used).

Stephen and I experimented with a couple of extra varicaps for the modulation. One of these was used to adjust the coupling between the mod varicap and the oscillator, using the tuning voltage, so that you got the same deviation right across the band. We scaled the mod circuit so that 1.5V p-p gave 75kHz deviation - right across the band. This allowed the use of a "brickwall" clipper to prevent over-deviation - it just wasn't possible to get more than 75kHz deviation!

As a first go, this is a good try. However, you can improve it radically at little cost, by taking a few ideas from elsewhere. Stephen and I experimented with a heatsink that went across the PCB, providing both cooling of the output device (and a Voltage Regulator IC) and screening the output filter from the rest of the board. He always insisted on single-sided PCBs, because he found that home constructors frequently couldn't handle double-sided. His boards were always designed with as much earth plane as possible. He also introduced screens made from PCB material around the oscillator.

One weird one he and I found - an early prototype of the PLL PRO III hummed badly. We tried a couple of different power supplies - it still hummed. We tried adding some extra earth paths on the PCB and cutting others - it still hummed. Eventually, trying to identify where it was coming from, I noticed that the hum was reduced if I had my head over the board..... He had fluorescent tubes lighting his workshop, and the 50Hz flicker was being picked up by the glass-bodied varicap diodes! I quickly proved this by scribbling on the diodes with a black marker pen!

[Bton-FM]

Where do you get your 5511’s Albert?

[Albert H]

https://www.ebay.com/itm/50pcs-TSA5511- ... 0911222270

https://dutchrfshop.nl/en/pll/154-270pf-500v.html

Shop around and you can get them quite cheaply.

For simplicity, you'll need either a 3.2MHz or 6.4MHz crystal for the 5511 to get either 50kHz or 100kHz step size. I clock the PIC off the PLL reference - this saves another crystal.

I found a source of 6.4MHz crystal oscillator modules - these are temperature-controlled devices and astonishingly accurate - I divided one down to 198 kHz and beat it against Radio 4 LW - the accuracy was amazing once the module had stabilised. Used as a PLL reference resulted in an output frequency accurate to <1kHz at 100 MHz!

[Bton-FM]

Is that eBay one legit?

[sinus trouble]

Cheers Albert!

Although the PLL PRO III was far superior, The 5Watt VFO has always been one of my favourite kits!
I did not feel worthy to mess with Stephens hard work so I kept it as original as possible.
Also it was meant to be a one off project, Yet I have made 4 of these now! lol
I still have a few SAA1057s and MRF237s in stock, Once these are exhausted that's it!

If I was to re-design? I would deffo go for your two varicap method! The audio response rolls off with increased PLL control voltages. 4V seems to be a happy compromise!

The FET output stage you mentioned is quite interesting? I might knock up a prototype and see what happens!

[Albert H]

This little PA card is available on eBay for £15 - £20. At 12V supply, it'll give 15 Watts for about 40 mW in. As you can see, it's no-tune and this one is a cheap Chinese copy of one of my boards from about 10 years ago. It was designed as a little PA to tack on to the end of a PLL board for a 10 Watt, battery-powered rig for rural use in small towns and villages. It is actually remarkably linear, and a re-coiled version is built into my 2m multimode mobile rig. It handles SSB remarkably well.

Our tiny little PLL exciter module is stereo, with optional RDS as well, based on the QN8027 IC and an Atmel processor, and has an LCD frequency display on the end of a flat ribbon cable. Frequency and RDS programming is done with either a rotary programmer, push buttons, or a USB input. The settings are retained in non-volatile RAM in the Atmel processor. It delivers about 40mW (or 300mW) with a 5V supply as it has a BFG135 wideband amplifier stage after the IC output. I found it necessary to put a lowpass filter at the output of the PLL board, and added a couple of audio buffer stages to isolate the RF end of things from the audio input sockets. The audio stages then made it easy to add an optional balanced audio input.

The whole little rig fits nicely into a ½-width 19" module box, so we can have two of them in a 1U 19" case. We include a small SMPSU which each exciter and PA which can work on any AC supply from 80 - 290 V AC at any frequency from 35 - 75 Hz. We've seen these things being run from waterwheels and wind turbines!

They're cheap as chips and easily thrown together. Their specification is good enough for British RSL use, and there's a version of the PA with a pair of 15 Watt FETs to give 25 Watts for the usual maximum permitted power. The little rigs have found their way all over the world in their various configurations, and I know of several pirates using them as small, convenient rigs with reasonable power. Unfortunately, we don't make these any more, as the Chinese undercut any price point we can manage, so there's no real market left for them. If I can find a complete rig around here, I'll snap a couple of photos.

[Maximus]

Having had a good look over the different options, I'm looking at giving this one a go https://www.amateurradioshop.nl/webshop ... -2018.html
Has anyone had a play with it yet and if so, what's it like?

Yes it’s easy to build and has a clean output.


Sent from my iPhone using Tapatalk

[Albert H]

Having had a good look over the different options, I'm looking at giving this one a go https://www.amateurradioshop.nl/webshop ... -2018.html
Has anyone had a play with it yet and if so, what's it like?

Yes it’s easy to build and has a clean output.

Unfortunately, there's an issue with the adjustable power output. If you run it flat out, it's very clean. If you turn it down a bit, it gets dirty. There are a couple of "sweet spots" on the power setting, where the output is clean, but you're going to need a spectrum analyser to see where they are! Basically, the inter-stage and output matches go awry when the supply voltage to the driver and final is fiddled with. It would be much better if there were just two power settings - the full 15W and perhaps a 2W setting (if the conjugate match would work out there).

You're going to need a good, clean power supply too.

[Albert H]

Incidentally - the NRG PLL PRO III worked out at 4 Watts and 1 Watt for the two points at which the match was OK. This was very convenient as Stephen had a 10W PA that needed 1W drive and a 45W PA that needed 4 Watts...... I found that an additional earth bond beneath the output device and a couple of extra bypass capacitors would get the output up to about 5 Watts with a 13.5V supply. The BLW60C PA would do exactly 50 Watts on any frequency on Band II with this amount of drive and at this voltage. Stephen "sloped" the gain of the final stage, so that my modified driver board output varied from 4.75 Watts at 88MHz to 5.15 Watts at 108 MHz. The BLW60C would always deliver exactly 50 Watts (assuming the trimmer was adjusted correctly).

[Ignite]

Albert, thank you for the info about the issues with the 15W Dutch board. As I was looking for the ability to alter the output power, this makes the board a no no.

The amp that you mentioned above looks interesting. I do however have a question about the QN8027 board that you mentioned. I thought that there were issues with the QN8027 and how clean the output was. Some of the boards based on this chip that I've seen result for look to have issues with harmonics and noise. Is this the case with the board you are using, and how do you get around this enough to put it through an amplifier? Is it possible to obtain one of these boards and where would I look to buy one?

Thanks as always

[Albert H]

I was really surprised by the output spectrum from the QN8027. I expected it to be like the Rohm chips - noisy! They're surprisingly monotonic, with the 19 kHz pilot at exactly 9% of maximum deviation. There's little or no carrier noise when run at 3.3V, but the noise floor rises as the voltage goes up towards the upper limit of 4.2V. We found that there was more noise from the Atmel chip we were using to programme it! We had serious problems getting the digital crud from the logic out of the 8027 carrier. The solution turned out to be a combination if physical separation (we put the SM ICs on opposite sides of the board) and two ground planes - an RF one and a logic one.

The IC puts out a -50 dBc second harmonic. The little PAs we used would amplify this and gave us more second than we would have liked at the output despite the lowpass filter. The fix was to add a two stage fixed lowpass filter after the post-IC amplifier stage. At 40mW output there were no harmonics or spurs visible at any frequency chosen in Band II. This was enough to drive the PA to just over 10 Watts. The output spectrum was superb once we added the screens around the output filter coils.

We had a batch of the boards made, and had the components loaded by a company in Hong Kong. The finished boards were easily fixed to fit into the cheap metal boxes we'd had folded for them. They made for very cheap 10W rigs!