[R-390] Best approach for SSB mod on R-390A

[Drew P.]

On the 2 diode AGC modification for SSB on the R-390A, Don (2002tii) wrote:

<snipped>
"Here are the AGC and BFO mods I've been doing for 35 years:

(1)  Parallel R546 and R547 with solid-state signal diodes (1N914 or
equivalent), with their cathodes toward V509A.  This will produce
asymmetrical attack and release times.  (Attack will be very quick,
release will not change.)  To slow down the attack, if desired, put a
resistor in series with the diode across R546 -- I often use 1 to 5 kohms."

I've seen a similar scheme which put the resistor in series with the time constant cap.  I think that would allow the full AGC voltage to reach the gain controlled stages (relatively) immediately upon reception of a large signal transient, instead of being dragged down while the cap is charging. The effect might be to reduce the "pop" often heard on the leading syllable of a strong SSB transmission.

"(2)  Install a new 20 uF film capacitor with one end to the junction
of C548 and C551 (i.e., the grid of V506A).  Disconnect the plate of
V506A from J512 pin 15, and connect the free end of the new 20 uF
capacitor to J512 pin 15 instead.  Switch the wires from S107,
terminals 9 (ground) and 7 (C551) -- ground to terminal 7, C551 to
terminal 9.  Install two, 2.2 Mohm resistors at S107 -- one each from
terminal 9 (C551) to terminal 7 (ground) and from terminal 8 (new
capacitor) to terminal 7 (ground).  The AGC switch will then add C551
in parallel with C548 for Medium AGC and the new 20 uF film capacitor
in parallel with C548 for Slow.  The new 2.2 Mohm resistors bleed any
residual charge from C551 and the new capacitor to reduce any pops
when switching between AGC time constants.  You can adjust the AGC
release times by changing the values of C547 and 548 (Fast), C551
(Medium), and the new capacitor (Slow).  I often use as much as 50 uF
for the Slow AGC capacitor."

I just use the AGC "SLOW" position without these mods - still works pretty well.  If doing so, it would be well to ensure that the 2 uF oil-filled AGC cap isn't leaky - mine was, and I replaced it with an under-chassis mounted polyester dielectric radial lead cap.

Don, do you see a disadvantage to using the original AGC integrator (tube) circuit with a good cap as opposed to disabling it and using larger caps as you did? Is disabling the AGC integrator tube mainly to facilitate gaining flexibilty in selecting time constants?

"Mod (2) fixes the "audio disappearing" problem when switching to and
from Slow AGC.  (Note that this is a design "feature" [i.e., problem]
-- If your 390A does NOT have this problem, it's because C551 is not
a capacitor anymore.)  If you can live with the "audio disappearing"
problem when switching to and from Slow AGC, you can omit mod (2)."

Ahhhh -  sounds like Dave Wise's "Ending the Moment of Silence" mod.  However, with all the mealy-mouthed preachers on the shortwaves nowadays, an occasional "moment of silence" can be a good thing!

"(3)  To increase the BFO injection, replace C535 (12 pF) with 47-75 pF."

I didn't perform that part of the mod.  If adding capacitance to increase BFO injection, it has been recommended to readjust the neutralization of the last IF amplifier stage to keep the added BFO injection from finding its way back through that stage and into the AGC IF amp/detector, where it would drive up the AGC voltage and desense the receiver.

ISTR where someone recommended replacing the 6BA6 BFO tube with a 6AU6 (12AU6 to all who use the paperclip ballast substitution) to obtain increased BFO injection.  I don't know how the increase (if any) thus gained would compare with that obtained by upping C535.

Drew