Message-ID: <3F61F04F.8070703@virgin.net>
Date: Fri, 12 Sep 2003 17:11:59 +0100
From: "Stewart Bryant" <stewart.bryant@virgin.net>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.4) Gecko/20030624 Netscape/7.1 (ax)
MIME-Version: 1.0
To: rsgb_lf_group@blacksheep.org
References: <000001c3793a$5d068bc0$fce8c593@rsch15>
In-reply-to: <000001c3793a$5d068bc0$fce8c593@rsch15>
Subject: Re: LF: RE: Inverted tube amplifier
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 8bit
Sender: <owner-rsgb_lf_group@blacksheep.org>
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Jim

See inline:

James Moritz wrote:
> 
> -----Original Message-----
> From: owner-rsgb_lf_group@blacksheep.org
> [mailto:owner-rsgb_lf_group@blacksheep.org] On Behalf Of Stewart Bryant
> Sent: 12 September 2003 09:20
> To: rsgb_lf_group@blacksheep.org
> Subject: LF: Inverted tube amplifier
> 
> 
> I came across an interesting concept yesterday that I thought
> worthy of further consideration for an LF active antenna. The
> so called inverted vacuum tube amplifier...
> 
> Dear Stewart, LF Group,
> 
>>>From what I could figure out over lunch - As well as reducing the value
> of mu, the value of transconductance gm will also be reduced by a factor
> equal to the "normal" mu of the valve. For a preamp, this has serious
> consequences for the noise. The noise voltage referred to the input of a
> triode is proportional to the square root of the "effective noise
> resistance"; as I recall, the ENR of a triode is approximated by about
> 2/gm. Therefore, low gm will result in high noise, as well as low gain -
> or if you prefer, the same noise but less signal at the output compared
> to normal operation.

I have no (recent) experience with valves, but with the amplifiers
we use at the moment we are not device noise limited at LF, so I
am not sure how much of a problem the increased excess noise would be.

> 
> The linearity is good because a given signal input only drives the valve
> over a small part of the characteristic curve due to the large bias
> voltage and small gain - 

Actually I think that it is a bit more subtle than that. The collegue that
pointed this out has spent a long time studying the behaviour of valves
and claims that the physics causes the transfer characteristic to be more
linear, rather than simply just relying on a smaller part of the
transfer curve.

> but this leaves you back at square one, because
> you then need to follow this input stage with enough gain to bring the
> signal level up sufficiently to drive the receiver, while maintaining
> linearity. 

But once you have done the impeadance transformation you can then put
in a filter and reduce the linearity problem to more managable proportions.

> If you could make an amplifier which would do this
> satisfactorily, you could just replace the "inverted tube" stage with an
> attenuator, with the advantage of lower noise, and even better
> linearity. The low gain and noise could be overcome by increasing the
> size of the antenna, but the increased input would then increase the
> distortion products again, cancelling the linearity advantage.
> 
> It is certainly an interesting circuit, if for no other reason than it
> enables useful output from the valve with extremely low supply voltage,
> thus allowing a low-impedance load to be driven relatively efficiently,
> as in the audio amp at the URL you gave - but I think the low gain
> negates any advantage as a preamp.

The thing that I find interesting is that it was invented by Fred Terman
who was the leading light in his field, but was never used for any
application. I think that it would be instructive to read his paper.
One reason to cast this aside in the past would be the difficulty of
getting power out of the grid structure, but we don't need any real
power, we just need to do the impedance transform into a selective
small signal stage.

73

Stewart G3YSX