[R-390] audio deck saga numbers

[2002tii]

W. Li wrote:

>Here is the list of changes:
>
>part     original value                      new value
>
>R612        220K ohms                   N/A  since it was removed
>R615         56 ohms                                24 ohms
>R101         6800 ohms                  zero ohms (shorting wire)
>R102         820 ohms                        820 ohms (no change)
>C609          8uFD 50v                    N/A since it was removed
>Rxxx            N/A                    10M ohm from pin 1 to pin 7 of V601
>C606       45-45uFD/300v                45-45uFD/350v (new)
>C603      30-30-30uFD/300v          30-30-22uFD/350v (new)

I'd be interested in the rationale for making these particular 
changes.  I haven't been following the thread, so it may have been 
explained before and I missed it.  But looking at the modified 
schematic, I'm not sure I understand why one would make these changes 
(aside from replacing C603 and C606).

Removing C609 decreases the gain of the first AF amplifier, 
V601A.  Adding the 10 Mohm resistor increases the bias of the cathode 
follower V601B slightly.  Both of these changes would make sense if 
there were a problem with V106B running out of headroom, but there 
isn't.  The net result is simply that the audio level at the local 
and line gain controls is a bit lower than on a stock radio.

The local AF amplifier comprises V602A and V603.  This two-stage, 
RC-coupled amplifier has two feedback paths: R615 provides some 
positive feedback to increase the open-loop gain of the compound 
amplifier, and R612 provides overall negative feedback that reduces 
the closed-loop gain while extending the frequency response, lowering 
the stage noise, and lowering the output impedance.  Reducing R615 
reduces the positive feedback a bit, while eliminating R612 
completely eliminates the overall negative feedback.  The result is 
increased overall gain, increased stage noise, and a higher output 
impedance.  I did not calculate the increase in gain, but it no doubt 
more than offsets the gain lost by removing C609.

Finally, replacing R101 with a wire increases the level at the 
headphone jack by nearly 20 dB (although this will be reduced 
somewhat by loading from R102).  At the same time, by leaving R102 in 
the circuit, the 600 ohm output ends up with 820 ohms of permanent 
resistive loading (as opposed to the 6800 + 820 = 7620 ohms it 
originally had).  This is exacerbated by the circuit changes 
discussed above that raise the output impedance of the local AF 
amplifier.  Placing a 600 ohm load on the amplifier (either a speaker 
or phones) results in the output -- which was originally designed to 
feed a 600 ohm load, and is now suited only for higher-impedance 
loads -- facing a load of about 350 ohms.  V603 is not a very happy 
camper in this case.  (Note that the impedance of most modern phones 
is substantially less than 600 ohms to begin with.)

At the very least, you might consider increasing R102 to 10 k or 
so.  (It may be possible to just remove it -- since there is no NFB 
taken from the secondary of T601, the presumably rather high leakage 
inductance of the transformer shouldn't cause any mischief even with 
no secondary load.)  But that still leaves the other changes, which 
do not seem to me to be for the better.

I'm not suggesting that no improvements can be made to the R390A.  I 
might re-bias V106B substantially, using a voltage divider on the 
grid and replacing R627 with an appropriate resistor to ground.  This 
would require adding coupling capacitors both at the grid and at the 
cathode, where the feed to the local and line gain controls would be 
taken.  The latter would have the added benefit of removing the DC 
from the gain controls.  But the cap feeding the controls would need 
to be huge, since the load of the two pots in parallel is only 1250 
ohms.  That would be fixed by replacing the 2500 ohm pots with 10 k 
to 50 k pots, which would also improve the linearity of V601B.  (See 
how complicated everything gets when you set out to improve 
things?  Ankle bone connected to the shin bone....)  I don't really 
see any need to change the stock V602A/V603 local audio circuitry, 
unless one wanted to make adjustments to the overall gain by changing 
R612 (say, within the range of 120 k to 330 k).

Best regards,

Don