LF: Re: [rsgb_lf_group] Linear Amp for LF

[Andy Talbot <andy.g4jnt@gmail.com>]

Shortly after my original post on this topic I received two PMs, both pointing at a design by W7IUV for 400 Watts output, with a hint more is possible, and that its robustness is proven in practice..    This was for use at 475kHz but its design is broadband, so by choice of ferrite could easily span 137 and 475kHz. I won't repeat the URL here as both responders chose to reply individually with it, but if you Google (*) W7IUV something will show.

One reason this design appeals to me  is the use of degenerative feedback formed from multiple source resistors.   In any amplifier design, using devices substantially below their maximum frequency - which is the case with switching FETs at LF - some sort of feedback should, really,  always be used.  It will help with bias stabilisation, as well as keeping the overall design more stable.   MOSFET amps for HF nearly always apply NFB, often via a dedicated feedback transformer.

The G7IUV design shown doesn't give details of bias circuitry or PSU but these are pretty standard for MOSFETs, although th euse of multiple parallel FETs does ideally require that FETs be all from teh same batch to ensure moderate matching.

Andy  G4JNT

(*) Other search engines are available (allegedly) 

On 9 Jan 2016, at 09:14, "David G0MRF g0mrf@aol.com [rsgb_lf_group]" <rsgb_lf_group@yahoogroups.co.uk> wrote:

 

Hi Terry.

 

Ask away...You have almost read my mind.......

 

I've been looking at the class D circuit in detail over the past few weeks....FETs obsolete....SWR detector not quite right etc.

I've even used it on its limits to make a nice QSO with SV on 630m last weekend.  - Got so hot the snubber resistor on one of the drains fell off

 

I plan on producing around 50 boards if my Chinese manufacturer hasn't gone bust in the stock market crash.... 

However, I'm not sure the linear board has much sales potential as CW / WSPR / JT9 can all be done with non linear amplifiers.

 

Also, the recently introduced power FETs are all optimised for rapid switching in power supply applications. That's very different from 15yrs ago when the Vg and Id transfer characteristics were much more 'linear' in nature.

 

What do you think??

 

David