[Collins] 32S-1 grid problem found

[Gerald]

On Sat, 2005-12-03 at 08:20 -0600, David Murman wrote:
> I am not sure of the part number but he is talking about stud mount feed
> thru going from under the chassis in the shield area of the driver plate
> tuning caps and then going to the pa neutralization cap C57 in the final
> shield area. Both 32S-1s I have this feedthru. I cannot find it on the 32S-1
> schematic I have but I'll keep looking.
> 
> 
> 
> 
> David

In all the documentation (two 32S-1, one each 32S-3 and 32S-3A) I have,
C62 is shown on the schematic as an ordinary capacitor, yet in the parts
list its a feed through capacitor. Its part number and specifications
change with time, beginning as a loose tolerance part, later as a 10%
capacitance tolerance part. Its not drawn in the schematic as a feed
through though C137, C166, and C167 which started out as the same part
are. C56 is also a feedthrough (RF feedback) with the same part number
but is also shown as a plain capacitor. The schematics are FAULTY. Note
in the S-1 schematics that there is no connection dot where the wire
from the neutralization capacitor gets to the next variable. There
should be a dot there. That wire crossing just to the left of the coax
cable symbol. Between C69 and C71. That dot helps the understanding of
the circuit a whole lot. I sure did prefer the older schematic standard
that required a loop where wires crossed and didn't connect. Then the
lack of connection was clear.

The use of a feed through in both those locations results in more
predictable performance with the drastically reduced (and controlled)
inductance of the capacitor. Both the PA neutralization and the RF
feedback are bridge circuits where a capacitor going series resonant or
inductive changes the performance over the HF frequency range. And since
these are adjusted bridge circuits, the precision and temperature
stability of that feedthrough is critical to keep the transmitter calm
from cold to hot, Antarctica to Viet Nam. Summer to winter in Korea.

What Glen is seeing as a variation may be that his transmitter came out
of production at a time when they'd just discovered the temperature
instability of the feed through to cause more PA instability and tossed
in a low capacitance feed through for the wire with the separate more
stable fixed disk ceramic. But then I believe they found that the lead
inductance (especially as it varied in production some having virtually
no leads, some having a half inch total) made the neutralization adjust
badly or to show it neutralized on 10m (where adjusted) and oscillating
on 80m. Its hard to convince draftsmen and assemblers that lead
inductance is important on a disc ceramic capacitor. Been there had to
do that. Using a $2.50 feed through capacitor in place of a 5 cent disk
ceramic solves that problem too and other problems.

He should use a tight tolerance feed through in place of the disk and
the feed through. No doubt it wouldn't hurt to check for leakage also.
It would be useful to select for temperature stability also. Many
ceramics have very poor temperature stability. Hence the -80+20
tolerance.
-- 
73, Jerry, K0CQ, Technical Advisor to the CRA
All content copyright Dr. Gerald N. Johnson, electrical engineer