[1000mp] Re: Quadra Warranty

Sun, 19 Jan 2003 22:31:28 -0800

Hi Billy,

Well, personally, I would not dare disconnect the ALC line on the Quadra. I
assume that by "quality of signal", you are referring to IMD. The following
might be a safer method for conducting the type of IMD test you describe:

1. Calibrate the ALC for 1kW CW output.

2. Run a 2-tone test at 1kW PEP. Measure IMD3 on the spectrum analyser.

3. Back down the drive to a point just below the threshold at which ALC
starts to level the output at 1kW CW. This will typically be approx. 925 ~
950W.

4. At this power level, re-run the 2-tone test. Take a new IMD3 reading.

I have not tried this particular test, but did run a 2-tone test at 1kW PEP.
Using an HP853A/8558B spectrum analyser, I measured IP3 = -33dBc (0 dBc =
composite 2-tone PEP). This is well within Yaesu's spec (IMD3 < -30 dBc),
and quite acceptable in the context of my HF operations.

Concerning a solid-state exciter driving a tube amplifier, the purpose of
ALC is to prevent overdrive of the amplifier. If the ALC is adjusted to
level the output at a reasonable value consistent with good linearity, and
the drive control of the exciter is set to a point where the output is just
below the previously set ALC threshold, the system will exhibit acceptable
IMD3.

If the ALC circuit is designed with appropriate time constants (fast attack,
decay sufficiently long to integrate the syllabic rate of speech) the ALC
should not degrade overall system linearity. The circuit elements which
determine the ALC time constants are usually in the exciter, as far as I am
aware; it is desirable that the ALC detector in the amplifier should not
materially alter these time constants.

There is a discussion of tube-amplifier ALC usage on my friend Matt, KK5DR's
website.

http://www.qsl.net/kk5dr/goodamps.htm

There is a detailed article on solid-state amplifier design on my site,
including a discussion of ALC, protection etc.

http://www.qsl.net/ab4oj/quadra/sshfamp.html

On warm-up times: If I may claim bragging rights, the Quadra's POST cycle
takes less than 5 sec. This is one of the nice things one gets for $3 to $4
per watt (vs. $1 per watt for a simple manually-tuned tube amplifier). I do
concede, though, that directly-heated glass RF power tubes warm up in a few
seconds.

Best 73,
Adam, VA7OJ/AB4OJ
North Vancouver, BC, Canada
http://www.qsl.net/ab4oj/

-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Billy Cox
Sent: January 19, 2003 15:19
To: [email protected]
Subject: Re: [1000mp] Re: Quadra Warranty

>Sorry...you are right. I got my wires (grids) crossed, thinking that the
>3CX1200A7 had a 4W grid-dissipation rating. I stand corrected.

Hi Adam ... not a problem, and I learn much from your posts on here! TU ...

>Here are the grid dissipation ratings for some representative Eimac
>and Svetlana tube types, from the data sheets:
>
> Eimac: 3CX1200A7: 50W
> 3CX800A7: 4W
> 3CX400A7/8874: 5W
> 3CX1500A7/8877: 25W
>
> Svetlana 4CX1600B: 2W (ouch!)
>
>One can work around the problem of grid damage in a tube amplifier
>by using a tough tube such as the 3CX1200A7or 8877 (25W is not bad).

Very true, the 8877 (and many of the others in the above list) also require
a 2 - 3 minute "warm up" time. That may seem like eternity for some ...

>In a solid-state amplifier, protection is not an option.
>Neither is ALC - it is the amplifier's first, or only, line of defense.

Adam, I know you have measured the output verses drive on the
Quadra there ... Have you also made any measurements as to
the differences as to with and without the ALC as to the quality
of the signal?

Seems like we have gotten away from using ALC with many
of the solid state to tube PA setups <as compared to years
ago, when we always seem to use ALC with tube exciter
to tube PA setups.>

Just wondering what happens (outside of losing some much
needed protection or insurance against PA damage) when
the solid state exciter does NOT use ALC with a solid state PA?

73 Billy AA4NU