found this on ebay. its from PCS

https://ebay.us/m/l9Qoz5

Yes I can sometimes get them on a portable if the skip is strong enough. And did have a 1/2 wave 20m long dipole on my fence they came in
on a radio super fine on that +9 or more sometimes. I lost a bit of interest for them as I have some humming crud across the MW band
at night some evenings and gave up.

Albert does not seem to like Aareff for some reason, I don't know enough about PLL rigs to be able
to confirm or deny if they are any good. Albert seemed to know the man who made NRG and spoke highly of him.

With very good reason.  The Aareff rubbish is an abysmal attempt to copy the original Steven Moss designs.  Steve was a good friend of mine, and an excellent engineer.  He developed the original Veronica designs - including a PLL version - way back in the 80s.  He made the mistake of employing the clown who now runs Aareff, and that clown tried to sue Steve out of business when Steve fired him.  Steve changed the name of his company from Veronica Kits to NRG, because the clown wanted the Veronica name for his inept copy products, and Steve, Christine and I decided that the "Veronica" name was effectively ruined by the clown's behaviour and inept copies, so the name change to NRG was done.  The name was suggested by another friend of Steve's.

The Aareff stuff tries to copy the 80s Veronica designs, but because the clown  didn't understand the principles that ensured the quality of the output from the NRG gear, he just made appalling sprog-boxes that he still persists in trying to sell, despite their abysmal performance and reputation.  The clown ran away after he ripped off too many people - first to Spain, then to the West Indies, where he languishes to this day AFAIK.

The NRG PLL Pro III was a properly broadband design, capable of >4 Watts without tweaks to the amplifier stages.  It could also be switched down to 1 Watt if necessary, and still retained its broadband ability.  The Kallitron oscillator operated at half the output frequency, and "self-doubled" to the output frequency.  It's crucial that the sides of the oscillator are accurately balanced to prevent half-frequency breakthrough - something that the clown didn't understand.  The clown tried to make his ridiculous exciter "no-tune" by saturating the output stage to give the same RF power output right across the band - which it does, but fills the entire band with noise and it gives as much power on the harmonics as it does on the wanted frequency.

That- basically - is why you should stay clear of the Aareff rubbish.  It's an over-priced way to buy a bunch of components to salvage!

The PCS gear is pretty expensive, but Marco's a great engineer, and his designs work very well.

That's quite similar to the ones we make, but we put screens between the coils.  I compared the filters with and without the screens, and the improvement in the higher harmonic products was significant with the screens in place.

The best filters we use (for higher power gear) use coax lines as resonant stubs, tweaked accurately for the frequency in use.  The higher power gear always uses pairs of output devices, which has the advantage of tending to null even harmonics.  The normally predominant 2f harmonic is several dB lower than the 3f before filtering if the PA is accurately balanced!  This has the advantage of reducing the stringency of the filters needed, especially for the lower harmonic products.

With a conventional filter like the one Zozo shows above, it's crucial to have really low impedance earthing and make sure that the PA is accurately impedance-matched to the filter to minimise losses and to minimise heating problems.  It's interesting to play with the "sense" of the windings - you can get a couple of extra dB of harmonic suppression by winding the first coil clockwise, the next anticlockwise, and so on.  The improvement is marginal but reducing the mutual coupling between coils helps a little.

Use ceramic tube capacitors of at least 500V.

I used these up to about 200W without any problems.
These did have slightly higher losses than silver-plated(Semco)capacitors or ATC type 100B.