[AMPS] Re: Parasitics

[Tom Rauch]

> Date:          Tue, 19 May 1998 01:56:07 +0100
> From:          w7iuv@axtek.com (Larry Molitor)

Hi Larry,

> I hope you don't take this comment the wrong way, as I really am serious
> this time, and I don't want to start a pissing contest. I really have a
> problem with the term "unconditionally stable". To state a design is
> "unconditionally stable" is to assume a position of omnipotance. And while
> you certainly are a talented design engineer, I can't quite credit you with
> omnipotance. If you said " The amp is completely stable at VHF under all
> conditions that I am aware of", I would accept that without question.

That is a much better (although longer) way of saying 
what I meant. Being "crystal clear" is difficult without peer review. 
I appreciate your insight.  

In fairness, here is how I do the test:

1.) No load on input and output, and I use an external L/C circuit or 
stepped transmission line (at VHF) to rotate phase of input and 
output terminations while testing.

2.) I bias the tube into and conduction with AC bias while testing.

3.) I rotate all controls on all bands through all positions in 
small steps.

4.) I apply about 25% more test voltage than operating voltage. (In 
solid state PA's, I test with less than operating voltage as 
well!)

I have never found a solid state PA "unconditionally stable" in my 
test standards.

I have never found a more than three 811A or more than one 572B 
tube amp that was stable at upper HF without being neutralized or 
without being  swamped by a resistor in parallel with the 
cathode.  

Virtually all tubes that offer poor VHF performance (3CX1200A and 
D7's, 811A's, 572B's, 4-1000A's, 833A's, etc) are the most unstable.


> How do you measure grid circuit resonances? It's easy in the anode circuit
> (big parts, lotsa lead), but have never had much success with the grid. I'm
> sure it's because I try to build them all like they were GHz range amps,
> and there is virtually no lead length to couple my GDO to.

Simple. Measure an effect that closely ties to the parameter you want 
to know but can't measure. Measure the coupling from anode to cathode 
using the grid(s)  as a shield.

Pull all the connections, and couple a low source impedance RF 
voltage to the cathode. Measure that voltage at the cathode and at 
the anode (nothing there but a resistor load) and plot 
transmission of power vs frequency.

By playing with the grid connections, you can see the effects of grid 
IMPEDANCE.

Resonance, as measured with a GDO, tells you nothing at all except 
where something external to the GDO couples to the GDO. You have no 
idea whether impedance is high, low or what.....or even if the 
resonance is what the grid actually "sees".

Measuring the transmission loss through the tube will paint a 
precise picture of the grid's RF impedance, because it becomes a very 
poor shield at frequencies where the tube "misbehaves".

Does that make sense to you?
73, Tom W8JI
w8ji.tom@MCIONE.com

--
FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm