Mosaic
Well-Known Member
I built the EBAY special 45w RF amp ($14) ....but signal purity is very poor under around 20 meters.
I've played around modding the amp and gotten some improvements based on front biasing and feedback and a bit of output transformer tuning with small primary capacitors.
https://hackaday.io/project/26999-45w-rf-ssb-cw-amp-ebay-diy-build-55w-upgrade
But although the inductance of each of the final centre tapped push pull output windings is 1 uH or about 18Ω XL at 3Mhz, the purity is terrible unless the frequency is raised to around 14Mhz. It becomes kinda sine wave like at 7 Mhz. At 14 Mhz (passable sinewave) the push pull windings are over 80Ω each, XL, yet this seems to work better, IRF530 NFET finals Rdson is less than 0.5 Ω, but I suppose some parasitics can be messing with their actual HF output impedance . The secondary of the final BALUN transformer is 6.5 uH.
You can find all the relevant files and details here:
https://drive.google.com/drive/folders/0BxARZnl2qhssYTNqMFBVQmN4eGM
Note the LTSpice mod#2 file for the latest tweaks. They improve purity but cost amplitude past 10Mhz or so.
So my question is why does the performance seem better at >10Mhz?
I found this interesting viz the transformers:
https://www.electronics-tutorials.com/basics/wide-band-rf-transformers.htm
I've played around modding the amp and gotten some improvements based on front biasing and feedback and a bit of output transformer tuning with small primary capacitors.
https://hackaday.io/project/26999-45w-rf-ssb-cw-amp-ebay-diy-build-55w-upgrade
But although the inductance of each of the final centre tapped push pull output windings is 1 uH or about 18Ω XL at 3Mhz, the purity is terrible unless the frequency is raised to around 14Mhz. It becomes kinda sine wave like at 7 Mhz. At 14 Mhz (passable sinewave) the push pull windings are over 80Ω each, XL, yet this seems to work better, IRF530 NFET finals Rdson is less than 0.5 Ω, but I suppose some parasitics can be messing with their actual HF output impedance . The secondary of the final BALUN transformer is 6.5 uH.
You can find all the relevant files and details here:
https://drive.google.com/drive/folders/0BxARZnl2qhssYTNqMFBVQmN4eGM
Note the LTSpice mod#2 file for the latest tweaks. They improve purity but cost amplitude past 10Mhz or so.
So my question is why does the performance seem better at >10Mhz?
I found this interesting viz the transformers:
https://www.electronics-tutorials.com/basics/wide-band-rf-transformers.htm