BC-348 AVC

[WU2D]

Hi Guys,

I have found a ton of info on AVC/AGC performance and improvements on many other receivers but I can not find anything on the BC-348. I picked a Q model up at Spring Hosstraders and I replaced the suspect caps and any high value resistors and even got the crystal filter working nicely, but I am not convinced that I have the AVC working correctly. It works great on AM with mucho negative volts on strong signals and a nice slow AGC action. On CW however, it just plain don't work.. In MVC it works fine for CW. When switched to AVC however, the BFO seems to phase lock on strong CW signals and the receiver overloads. Turning down the gain control does nothing to help.

The manual claims that the receiver is specifically designed to handle CW in the AVC mode - any personal experience comments before I post the writeup and schematic details of the AVC section? 

Thanks,

Mike WU2D

[WU2D]

No Bites so I will attach the schematics for your look see. This is a an old design, but it is one of the later models of the BC-348 and the circuits are a result of improvements on the series during the war years. There is plenty of gain with two RF stages and three IF stages! Again I am looking for some help on the CW AVC action.

Mike WU2D

[W2JBL]

things get better if you divorce the RF and AF gain controls from the ganged setup they have, but still not very good for CW with the AGC on. might i suggest a product detector? that worked well for me. this is an old buzzard design, and ask any old buzzard how to operate CW it starts with turning the AVC off. great receiver pretty much as is, i have had many. i like to keep them stock as they are fine on AM. my father used one on CW from 1946 to his QRT about 8 years ago. never saw him use AVC.

[WU2D]

Chris,

Thanks for the reply and the comments which come from experience with the old girl. My buddy had one of these as his novice rig and that was my only exposure before picking this one up. You are validating what I thought - that CW AGC was more wishful thinking than reality with the BC-348. No I'm not going to change or modify the radio - It works fine for what I use it for - checking into the OMRN Net once a week!

Thanks again.

Mike WU2D

[w3jn]

Mike, are you sure the thing is wired to factory?  A typical mod was messing around with the RF gain pot, I can't remember what it dealt with, but a couple of the 348s I've had over the years needed de-hamming in this area.

Another area to watch out for is that hams loved to ground the B- from the power supplies they built in.  This prevents any negative bias (IIRC it only affects the audio stage though).  The B- should NOT be grounded,  per the skizmatic.

It sounds to me like the AVC is disabled with the BFO on.   The reason the BFO is phase-locking is that either the IF signal is very strong (as in no AVC action, or someone markedly increased the coupling from the BFO to the IF (also possible, but a high BFO signal usually increases the AVC voltage).  Further, you have a distorted signal which leads me to believe the AVC isn't functional with that BFO on.

[k4kyv]

If the receiver doesn't have a separate agc amplifier, but feeds the agc rectifier directly off the final i.f. transformer secondary, parallel to the diode detector, the agc must be disabled whenever the BFO is turned on.  If not done automatically, it must be turned off manually.  Otherwise, the agc rectifier makes no distinction between the BFO output and a strong signal, and the rectified BFO output generates enough agc voltage to nearly cut off the agc-controlled stages.  Turning off the BFO for CW and SSB is standard procedure with the classic single-conversion superhet where the i.f. feeds a diode detector and the BFO is coupled to the diode detector through a small capacitor.  In some  receivers, the same diode serves as detector and agc rectifier.

If the receiver has a  separate agc amplifier stage, I have found that slow agc is  best for cw and ssb, but for AM, fast agc is preferable, since it allows the agc to better track the carrier under QSB conditions.

Another problem with using agc with  cw or ssb is that under key-up condx or between ssb syllables, the receiver gain recovers and the background noise rises between cw characters or voice syllables.  Hence slow agc for cw or ssb.  If I use agc at all for those  modes, I use just enough to prevent the receiver from overloading on strong signal peaks, by turning down the rf  gain control.  Some ssb receivers attempt to get around this problem by using audio-derived agc.

Maybe you could find two pots mounted with concentric shafts to separate the af and rf  gain controls.  That used to be common with older tube type TV sets which often had concentric knobs to accomadate dual functions.  Just make sure you use pots with the proper taper.  The rf  gain should be linear taper, while the af gain should be audio, or logarithmic taper.  If a linear taper pot is used for the af gain control, all the audible signal variation will be squeezed into about the first 25 degrees of rotation of the pot.

[w3jn]

If the receiver doesn't have a separate agc amplifier, but feeds the agc rectifier directly off the final i.f. transformer secondary, parallel to the diode detector, the agc must be disabled whenever the BFO is turned on.  If not done automatically, it must be turned off manually.  Otherwise, the agc rectifier makes no distinction between the BFO output and a strong signal, and the rectified BFO output generates enough agc voltage to nearly cut off the agc-controlled stages.  Turning off the BFO for CW and SSB is standard procedure with the classic single-conversion superhet where the i.f. feeds a diode detector and the BFO is coupled to the diode detector through a small capacitor.  In some  receivers, the same diode serves as detector and agc rectifier.

THat's one of the reasons BFO signals are very lightly coupled to the IF.  You can improve SSB reception by increasing the coupling, but at the expense of decreased sensitivity due to the AVC action (resulting from the BFO signal) decreasing the gain of the receiver.

The other reason, as I noted, is to prevent the BFO from phase-locking to the incoming signal.

[W1RKW]

The other reason, as I noted, is to prevent the BFO from phase-locking to the incoming signal.

What is the behavior of a BFO phase-locking to an incoming signal? What does one hear when this occurs? I've been messing around with an old buzzard rcvr where the BFO injection is extremely light but I get strange behavior when I increase the BFO coupling. One thing I'm definitely getting is reduced sensitivity.

[w3jn]

You can't smoothly tune the BFO thru zero beat and up the other side.  The BFO will jump in to zero beat (or nearly so) at low audio frequencies.  The BFO could also try and phase lock itself to the audio creating an irritating FM effect. 

[WU2D]

I'm still in sunny California and do not have the skizmatic with me. One trick that the BC-348 uses is to inject the BFO into the last IF stage instead of into the detector. That last IF has no AGC applied - it is the only fixed gain IF stage.  That stage amplifies the IF and also acts as a BFO amplifier. I am not sure how they isolate the BFO from getting rectified by the AGC detector. Perhaps they keep it low enough to not cause trouble with that trick.

As far as mods go, I did not see anything fishy and the unit never had a ham power supply - it had a dynamotor inside that looked original.

Thanks for the comments. I obviously have some troubleshooting to do with my generator and a voltmeter.

Mike WU2D

[W2JBL]

clamp your HS-53 cans real tight to your head. tune in WSM on 650KC. key your interphone and say "captain i've got some great music coming from the states- do you want me to patch it over the interphone?" last time the touring B-17 and B24 came by the local airport (a few months ago) i stopped while touring the plane, sat down and cranked that BC-348 to 3885... then while nobody was looking i also slammed the 80meter tuning unit into the '17's festering BC375. if anybody out there tours the Collings Foundation's B-17 "909" and finds the rig on 75AM it was only me...

[k4kyv]

THat's one of the reasons BFO signals are very lightly coupled to the IF.  You can improve SSB reception by increasing the coupling, but at the expense of decreased sensitivity due to the AVC action (resulting from the BFO signal) decreasing the gain of the receiver.

The other reason, as I noted, is to prevent the BFO from phase-locking to the incoming signal.

The first problem can be addressed by turning off the AGC.

I discovered the second phenomenon while living in Houston in the late 70's.  The HRO Sr. had too little BFO injection with that little fabricated coupling capacitor consisting of a chip of bakelite separating  two small rectangular metal plates.  It worked ok for weak-signal cw, but there was not enough injection for strong cw signals, and it would not demodulate SSB very well.  So I used the injection  capacitor as a terminal strip and soldered a larger cap - something like 50 pf in parallel with it.

With increased  capacitance it demodulated SSB almost as well as a product detector, and with AGC off, the sensitivity on SSB/CW did not decrease.  Even  with the stock coupling capacitor as I recall, that receiver never worked well with the BFO and AGC on at the same time, so I always used the classic procedure of turning the BFO on, AGC off, audio gain near maximum, rf gain near minimum, and advancing the rf gain for the most comfortable copy.  The result was a crystal clear cw or slopbucket signal with absolutely no "pumping" sound.

One night I noticed that with a stable AM carrier, I could turn on the BFO, set the carrier to the centre of the passband, and zero-beat.  With careful tuning, I would  hit a point where the AM audio, especially from a marginal station, suddenly seemed to jump right up above the QRM/N and out  the speaker.  I had found the point where the BFO phase locked with the AM carrier.  This was  my first experience with a crude, but effective SYNCHRONOUS DETECTOR!

I used the receiver in that mode to copy weak AM signals until I replaced it with the 75A4.  Of course, when copying SSB or CW, if the signal was too  strong, the BFO would FM, resulting in gibberish or chirpy cw.  Reducing  the rf gain always solved that problem.  But I found, that with a stable AM carrier, carefully zero-beating would result in phase-lock, even with the weakest AM signals.  Usually there was enough drift that I  had to touch up the tuning every 30 seconds or so to restore lock.  But what else could be expected with this crude, brute-force method?

There is a construction article in one of the old GE Ham News flyers for building a phasing-type "sideband slicer."  For USB and LSB and AM it uses a product detector.  For SSB and CW the phasing method is used to reject the unwanted sideband.  For AM, a (tube type) phase-locked loop circuit phase locks the BFO to the AM carrier, but the sideband rejection feature is turned off, so that true DSB reception is achieved.

The advantage of the synchronous detector is that both sidebands are received simultaneously, with a 3 dB gain in audio output due to the vector addition of the two sidebands.  The output of the heterodyne  detector is simply the product of the BFO and the two sidebands (or one sideband with SSB).  With a conventional envelope detector you have the heterodyne product of the AM carrier and the sidebands, but also of the QRM and the sidebands, the AM carrier and the QRM, and every discrete bit of QRM heterodynes against every other discrete bit of QRM.  Basically, everything within the passband intermodulates with everything else within the passband.  Together, all these intermodulation products decrease s/n ratio and intelligibility.  Of course, the intermodulation between the AM carrier and the two sidebands is how the detector functions.

Most the apparent "superiority" of SSB over AM is due to the advantage of the product detector at the receiver, and is not inherent to the modes themselves.

With the synchronous detector, DSB retains the primary advantage of SSB, reception with the product detector.  The problem with receiving AM on a product detector is that the BFO signal must be not only exactly zero-beat with the AM carrier, but it must be precisely in phase or 180° out of phase with the original carrier.  Thus the injected BFO signal is synchronous to the AM carrier.

I have seen the description of "synchronous detectors" on the market within the past decade or so that not only lock onto the carrier, but reject one of the sidebands of the AM signal.  These are  not synchronous detectors at all; they are nothing more than classic sideband slicers.  Disregarding possible differences in audio fidelity, you can do equally well by simply copying the AM signal in conventional USB or LSB mode if the receiver has a product detector.  A true sync  detector takes advantage of the vector additive effect of the two sidebands plus the advantages of the product detector.

[w3jn]

GE marketed that sync detector commercially as the YRS-1.  It's very interesting in that it will sync up to a DSB suppressed carrier signal - ituses the phase differences between the sidebands to phase lock the BFO so it doesn't need a carrier.  It has 2 separate detectors, and both I and Q channels which are then added/subtracted to select USB or LSB.

I have one, but I haven't been able to get it to phase lock.... another one of the million projects on the bench.