Amps
[Top] [All Lists]

Re: [Amps] QST, and class AB/F MOSFET amps

To: amps@contesting.com
Subject: Re: [Amps] QST, and class AB/F MOSFET amps
From: Manfred Mornhinweg <mmornhin@gmx.net>
Date: Thu, 21 Dec 2006 01:58:58 +0000
List-post: <mailto:amps@contesting.com>
I really hope it's not too improper to take this thread for my first 
post to this group:

>      I am becoming more and more disgusted with the lack of technical 
> knowledge QST authors.

If you mean the very usual bugs in the schematics, and omission of 
important information, that's mostly a problem of the editors, not the 
authors. Some years ago I submitted a technical article of some weight 
to QST. It took some work convincing them to actually publish it in QST, 
as it was deemed too technical. Then, they insisted on re-drawing my 
schematics, despite my having submitted them ready to be published. They 
introduced many severe errors. I proofread the article, corrected the 
errors, they made the corrections but introduced some new errors and 
then published the article without further proofreading. Anyway it won 
the cover plaque award, which shows that the readers after all DO like 
to read "heavy" technical articles!

QEX is less "editoriated", so there is a greater chance to get an 
article in without unwelcome modifications.

Now to old amps:

 >  A fine example is the article about reviving old
> linear amplifiers in the most recent QST.   There is a schematic and 
> description of a half-wave voltage doubler which is claimed to have been 
> used on many older amplifiers. But, everyone that I have seen uses 
> full-wave voltage doublers.

You mean full-wave rectifiers?

I own and regularly use, and have repaired, a National Radio NCL-2000 
amplifier. I think it's from 1963 or so. This amplifier indeed uses a 
voltage doubler to make the 2.6kV. In a homemade beast using four 813's 
  I also employed a voltage doubler, just to save me a lot of turns when 
winding the transformer! So that makes two.

 > Where is Wayne Green when we really need him?????

He's alive and well. I had e-mail contact with him a month or two ago. 
But in my opinion he turned too far away from radio. The last issues of 
73 I got were more about esoteric holistic cancer-curing stuff than 
about radio, so I discontinued my subscription. Shortly later the 
magazine bit the grass. It's a pity.

Now to the main purpose of my post:

QEX carried two articles lately that are interesting. One is about a 
class AB/F amplifier using TV sweep tubes. The performance of the AB/F 
principle seems good enough for all kinds of ham use, and the high 
efficiency is very attractivo to save space, weight, but above all heat 
and thus fan noise. Then came another article about a 1.5kW solid state 
amp using a pair of ARF1500 MOSFETs. I had been watching these FETs for 
well over a year, but didn't dare to try them due to worries about 
hotspotting, and cooling issues. The rated 1500 Watts are fantasy 
values, unachievable in practice, with 500 Watts being reasonable and 
700 Watts possible but hard to do. I was thinking about running a pair 
of ARF1505 at 300V, obtained by directly rectifying and filtering my 
220V AC line, and making the insulation in the RF transformers. That 
would be the ultimate in cheap, simple, lightweight solid state 
amplifiers! But the hotspotting problems seem to make reliable operation 
unlikely.

Then it came to my mind that joining the two articles seems almost 
ideal: A class AB/F amp with ARF1500 MOSFETs would be free of 
hotspotting, because in the AB class it would run at very low voltage 
(20 or 30V at most) and above that it would be running class F. At the 
same time, the reduced dissipation goes a long way in helping with the 
cooling problem. The added complexity of the modulated switching power 
supply would be rewarded by the high efficiency. And the whole thing 
would still be small, lightweight and reasonably cheap. Any takers, out 
there?

If nobody else tries and publishes this soon, I might be the first 
volunteer to try it. What scares me is the risk of blowing up a few 
pairs of FETs during development, at $450 per pair (plus shipping to my 
end of the world, plus taxes). Also, time is scarce, but I guess I would 
steal some time off other activities for this project!  I already made a 
crude design, some simulations, and the biggest problem is input 
impedance matching. After all, in class F we need really stiff driving. 
So far it seems that a band-switched LCR input circuit is needed, with a 
separate (but simple) matching network per band, before going into the 
broadband input transformer.

I would be really interested in knowing what you think of this. Am I a 
crazy dreamer (from reading too much 73) and just loosing my time, or do 
you see merits to this approach? Or do I come late, and someone has 
already done it?

Manfred, XQ2FOD.

--------------------------
Visit my hobby homepage!
http://ludens.cl
--------------------------
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>