Hallicrafters SR-150 Operation And Maintenance Manual page 37

FILTER AND BALANCED MODULATOR DISCUSSION CONTINUED.
Thanks Jim.
So if I am following this correctly, if the filter is swept for a 2.7Kc pass band the chart should look like this:
This makes good sense for better performance. My only question now is what effect will it have on carrier suppression in
SSB? I'll have to test that. Carrier suppression in the SR-150 just does make spec now. However some of that is from not
using match diodes in the balanced mod ckt. Considering all this, the direction the factory manual leads us is truly miss-
leading.
You have it right on this one. The sideband suppression is slightly lower in USB mode in the SR-150/400/400A because
they had to move the USB carrier frequency far enough into the filter pass band to generate enough CW and Tune drive.
On your drawing, add the LSB carrier and it will make more sense when you release it because it will be farther down the
skirt than the USB carrier. The SR-2000 used the 30pf cap across the USB crystal in CW and tune mode to move the carrier
from 1.651550 to 1.651450.
One other note, I believe the flat portion is more than 2.1Kc. May be different from filter to filter but believe it to be closer
to 2.3Kc. Sounds like you are having a great time with the spectrum analyzer and sweep studies
That is what it is supposed to do. But in some rigs the best you can get is about 54db. If you adjust for min carrier cold
then it is about 45db when hot. Of course you should balance it out hot, but then you are out of spec for the first 30
minutes of operation. The balance trimmer cap (5-25pf) is a ceramic NPO cap. It is in parallel with a standard mica cap,
18pf. I have, on a couple radios replaced the 18pf with a mica COG (NPO) and put in a matched set of 1N87's (yes I can
still find them) and get 58db hot, that also maintains at least 52db when cold. In some cases I have replaced the 470
resistors and improved rejection and balance. I have developed a simple test. If you put the balance pot (250ohm) in its
mechanical center then adjust the balance trimmer cap for min carrier. Then if it takes more than 1/3 of the pot to get min
carrier either the diodes or the 470 ohm resistors need to be replaced. Oh, well I do believe you could write volumes on
every little ckt in these rigs.
Thanks again for the input and suggestions. But alas, every time we talk I get 2 or 3 more things to research.
Walt, there you go again passing to me more information that I was totally unaware of. I have never paid any attention to
the balanced modulator, but now, l will have to add another regression test to a very substantial list. What is your source
for the matched 1N87's --- based on your comments, I will need a bunch over time. Would never have guessed the
balanced mod would be temperature sensitive but now will have to check all radios that visit my qth including my SR-
2000. Walt, every time I think we are done another revelation appears. This sounds like a good one though. I've heard
folks frequently that sounded like their radio had a lot of carrier bleed through and this may be the answer.
I can't thank you enough for bringing it to my attention.
I have found sources for matched pairs but they were $18.00 to $24.00 per set. RF Parts has individual diodes for $1.00. I
bought 30 and used my curve tracer to match them up. Out of 30 I ended up with 12 perfect matches.
http://www.rfparts.com/diode.html. The application in the 150 does not approach any spec limits on the diodes. The most
stress they see is in the CW/Tune function with CR10 is biased off and CR11 is biased on to pass the USB osc signal
straight thru. At that it is only running 1.6ma. It is good for 50ma.
By Walter Cates WD0GOF (UPDATE 10/30/2013) 37