[Premium-Rx] Reaction Instruments 685 AM-FM demodulator

[Peter Gottlieb]

Last I left off I had changed a couple of IF filters in my unit and moved on to 
the FM demodulator to improve performance for narrow BW reception.

The demodulator assembly is a densely packed PCB which on first glance didn't 
make a lot of sense to me.  There are two sets of four plug-in filter PCBs, each 
with trimmer caps.  Why are these there? Isn't the IF filtering done by the 
Miniceptor-style filters?

The first thing I did was to make a fixture for my VNA to look at the response 
of these filter assemblies.  Unfortunately that didn't tell me much, they are 
not very selective, pretty flat, around 21.4 MHz.  So, OK, I need to make a 
diagram of the demodulator to see what they're doing.

This has taken me a bunch of days.  I only have an hour or so each evening to 
work on this and family events took me away for a week or so.  However, I am 
slogging through and now think I know how FM detection works.  I'm not an RF 
engineer so PLEASE if someone can offer any further insights, add to this 
discussion.

Detection is done with a MC1496 balanced modulator/demodulator chip.  One input 
is fed raw IF and the second is after going through one of the four selected 
filter sets.  It is not frequency that the filters are there for, it is phase 
difference.  The 1496 detects the phase difference and produces a voltage 
proportional to the deviation of frequency from center.  If the filters have a 
linear phase response zeroed at IF center frequency then this should be a nice 
way to recover FM information.  One of my next steps will be to put filters back 
on the VNA and look at phase response but I bet that's what I'll find.  Has 
anyone else ever seen such a scheme?

My next immediate step is to slog through the audio/video switching and 
amplifier circuits to finish my diagram of the assembly.

I did notice two repairs done to this PCB.  They were both neatly done so not 
immediately obvious.  The first I noticed when under microscopic examination I 
saw excessive oily black particulate contamination near the AM detector 
transformer.  An inductor from the power supply rail was of a different type 
from the others.  It fed a tantalum capacitor.  Apparently at some point the 
tantalum fried, creating the contamination I saw (most of which was cleaned up) 
and taking with it the inductor.  This is just a warning to others that this 
receiver contains tantalums which are subject to shorting and burning.

The second repair I believe was done wrong.

Audio levels will be different for different BWs and there is a means of 
adjusting gain in two spots.  Different resistors are switched in to effect the 
control.  There are two sets of transistors to ground and they are fed in 
parallel so that of the 8 transistors, two are conducting at any given time and 
the two controlled circuits are set appropriately.  Here's the problem:  one set 
consists of 2N4447 FETs and the second set are 2N2222A bipolars.  You cannot 
parallel the inputs to these two as the bipolar B-E junction will saturate at 
about 0.65 volts which is well below the 3-4 volt threshold for the 2N4447, so 
the 2N4447's will never turn on.  Looking closely, I see that the 2N4447's are 
original and the 2N2222A's were replaced.  It seems to me this was a mistake as 
it cannot possibly work right.  There must have been a problem with gain 
selection and a tech replaced four 2N4447's with 2N2222A's to fix the problem.  
It did, at least in that section, but now the gain control of the other 
controlled section won't work. Perhaps that is why I ultimately ended up with a 
receiver in such pristine condition?  Perhaps it didn't work quite properly 
after the repair and the tech gave up and shelved it?  That's a guess.  In any 
case I'm going to try and find some 2N4447's and restore it to a design which 
will work as intended.  Again, comments?

One of the group members offered to host scans of my diagrams once I'm done on 
his site but if anyone wants partials to look at and comment on send me a PM.

Peter