[R-390] SP-600 IF and IF harmonic leakage [WAS: Urban...]

[Charles P. Steinmetz]

Wes wrote:

>I will go back and re-do the IF bypass capacitors with ceramics.  I 
>like to solder.

As long as you are going back in there, it is worth your while to 
look at the 455 kHz IF output follower circuit, which is what causes 
the largest part of the IF and IF harmonic leakage.

Here are some extracts of previous posts from the list archives 
(February 2008) that explain the problem and what to do about it (not 
my posts, but I have found the same thing):

>Some versions [of the SP-600] have IF output amplifiers that run the 
>tube into severe nonlinearity, which generates bad spurious 
>responses at 910 kHz and other harmonics of 455 kHz.
>
>*   *   *   Put a scope on the 455 IF output [when you have a nice 
>strong CW signal tuned in] -- if it is a nice [clean] sine wave, it 
>doesn't need fixing.

Otherwise (if it looks distorted), here's the fix:

>on units that have the problem, it isn't in the IF itself, it is in 
>the buffer stage that drives the 455 kHz output jack on the rear 
>panel.  The harmonics of the 455 kHz IF signal that are generated in 
>the overloaded buffer amp are then received by the radio as spurious signals.
>
>There were (at least) three IF output schemes used by 
>Hammarlund.  The older ones give very high output (10V or more p-p) 
>with lots of harmonics (clipping on the negative swing), and the 
>newest one gives a very clean 1 or 2V p-p.
>
>In the oldest version the buffer ampifier -- a cathode follower 
>(V16a) -- is fed straight (through C145) from the tied plates of the 
>IF Driver (V11) and BFO Buffer (V12).
>
>The newest version uses an additional connection inside T5 to feed 
>the follower (through C145 and some shielded cable) from the top of 
>the secondary.  This has two advantages -- it reduces the feed level 
>by the gain of the IF Driver (V11), and it picks off the IF output 
>feed before the BFO injection.  The newest version also uses a 
>tapped coil (L53) as an autoformer on the output (the cathode of 
>V16a feeds the top of L53 through C147, and the IF output jack 
>connects to a tap a ways down on L53).
>
>The middle version has the old-style feed plus L53.  I believe this 
>is the most comon configuration.
>
>If one has an old-style or middle-style unit I highly recommend 
>changing it, because the distortion in the follower due to the high 
>signal level radiates harmonics of 455 and causes spurious signals.
>
>To change it, one can go into T5 and add the connection for 
>C145.  In this case (assuming one already has L53, or adds it) you 
>end up with exactly what Hammarlund built later.  Or, you can take 
>the feed without going into T5 by putting C145 on the junction of 
>R55/R56 (the output of T5).  I prefer the second method, because it 
>comes out at a bit lower level and does not unbalance the load on 
>the secondary of T5.  In either case you need to feed the signal 
>from T5/V11 through shielded cable, grounded at both ends, to the 
>grid of the follower (V16a).  C145 goes at the T5/V11 end of the 
>coax to minimize the capacitive loading on T5 (the capacitance of 
>the shielded cable ends up in series with C145).  You don't need to 
>try to chase down an appropriate coil to use for L53 -- the follower 
>can just drive the IF output directly.  If you want, you can use a 
>resistive voltage divider in place of L53 to lower the signal level some more.
>
>Alternatively, if you don't use the 455 kHz output, you can reduce 
>spurious responses even further by disconnecting C145 from the tied 
>plates of V11 and V12 and connecting it to ground instead.

 From the same post, information regarding two caps you may not have 
replaced, which can benefit from being replaced with ODs:

>There are only two audio coupling caps -- C143 & C149, both 5100 pF 
>mica.  You can change them to .005 or .01 uF Orange Drops if you 
>want, and will realize some improvement in the audio.

Best regards,

Charles