[AMPS] Is screen potential important?

[Steve Thompson]

-----Original Message-----
From: measures <2@vc.net>
To: Steve Thompson <rfamps@ic24.net>; AMPS <amps@contesting.com>
Date: 26 March 2000 14:41
Subject: Re: [AMPS] Is screen potential important?


>
>>
>>
>>-----Original Message-----
>>From: measures <2@vc.net>
>>To: AMPS <amps@contesting.com>
>>Date: 25 March 2000 11:54
>>Subject: [AMPS] Is screen potential important?
>>
>>
>>>
>>>In a G2DAF amplifier, screen potential wanders from zero to approx. max.
>>
>>
>>It doesn't wander. It follows the rf envelope.
>
>The ''envelope'' consists of RF cycles.  However, RF can not exist across
>the screen's RF bypass capacitor.  In other words, the grid sees RF and
>the screen can not.


Like I said, it follows the *envelope*

>>
>>>Some say that screen potential can wander about with no serious
>>>consequences.  Please consider the constant-current curves for the 8171
>>>posted at the beginning of my Web site.  One set of curves is for a
>>>screen potential of 500v.  The other set is for 1500v on the screen.
>>>Assuming a peak grid potential of +10v, with1500v on the screen, peak
>>>anode current is c. 12a.  However, with +10v on the grid and 500v on the
>>>screen, .peak anode current is 3a.
>>
>>But screen voltage is set by rf (=grid) voltage. If +10 grid gives +1500
>>screen it won't suddenly or randomly give 500V screen instead.
>
>The charge on the screen RF-bypass cap. can not go from 0 to 1500 volts
>in one input cycle.

No, but neither can the rf drive from a SSB exciter go from zero to full in
one cycle. The screen charging circuit can take the screen voltage up as
fast as the rf envelope can change.

>After a sequence of +10v-peak drive cycles, the
>screen/cathode bypass cap.  eventually becomes charged to a level of
>1500v.  The peak anode current is then 12a.  As the screen cap. was
>becoming charged, it reached a level of 500v.  At this point, +10v peak
>drive cycles would have produced 3a anode current peaks instead of 12a .

The rf drive goes from zero to +10V in, say, 100us. At what point is the rf
drive +10 and the screen 500V?

>Does this sound like linear amplification?

That wasn't my point.

>>constant in the screen system is in the order of usecs, the fastest edge
in
>>a SSB signal is in the order of 100us - to all intents and purposes, the
>>screen will follow the rf envelope.
>
>All ''edge''s in a SSB signal are individual RF cycles at the fundamental
>frequency.  The variables are the amplitudes of RF cycles and their
>offset from the carrier frequency.

yes
>
>>I don't think you could get +10 grid and 500 screen at the same time with
a
>>normal SSB drive signal.
>
>The charge on the 8171's screen, RF bypass starts off at 0 volts and ends
>up at 1500 volts.  At some point, the charge on the RF bypass cap. had to
>have been 500v.

Yes, but what was the rf grid drive at that point - it wasn't +10 peak for
sure

>The fix is keep 1500v on the screen at all times.  RX/TX bias is handled
>by an electronic grid bias switch.  (Figure 4 on my Web site).  This
>arrangement provides a constant gain of c.20db.   Simply put, the DAF
>circuit only provides constant gain when the input signal is constant and
>maximal.

That's not how the measurements for the QBL5-3500 amp look.

>When the screen's RF bypass cap. is not fully charged, the gain
>is not what it should be.  //  note - In a typical HF, 8171,
>grounded-screen amplifier, the total cathode/screen RF bypass capacitance
>is c. 0.05uF.  .

With that much C it's probably not a good candidate to try in a DAF system.

I'm not arguing for or against the issue. The key thing is whether the rf
output envelope follows the rf input envelope faithfully. G2DAF's ideas are
not conventional practice, but I see nothing in there that says it *can't*
work if things come together right.

Steve


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