The cabinet arrived 3/23/06

This was achieved after "soaking" the PTO, leaving the filaments on for several days.  It may be slightly different with less "warmup".

I've given the PTO cover can a light buffing, ready to reinstall it and the PTO.

Cleaning this week

dirt on rag from one side of the KC dial

panel before cleaning

  panel after 2 passes, need to clean up a bit still.  The wrinkle had been partially filled with dirt, nicotine, etc.  Note, no red "7" any more.

grunge near power transformer, very sticky, not sure if it's tar from the transformer (don't think so, and hope not), or something that was spilled inside that end.  Decided to completely remove the side panel in order to do a better cleaning job, only about 2 dozen screws holding it & parts mounted to it.  May not be able to do too much down between the filter unit & xfmr.

  I took the IF cans off & found the whole problem ;-)    - the engineer was dead.   

After cleaning the grunge.  Last pic is the bottom of the power transformer, I don't think the stuff came out of it.

IF cans back in place, after buffing.  Starting to look pretty.

  I had noted tape over wiring at the meter earlier.  Apparently the 6.3vac line to the dial lamps had been shorted to ground at one of the lamps (I've heard of that problem before), and the wire had burned it's insulation off.  A previous owner had spliced a new wire on at the point near the meter where the wire comes out of a cable, and covered it with tape.  I took it loose and put heat shrink tubing over the splice.  

  Before re-installing the panel I re-installed the PTO and installed the cables necessary for switching in a product detector, which will be added.  The black wires are small dia. coax. cable for the new lines.  All knobs were cleaned and waxed, as was the front panel dial bezel.  I had noted earlier that all the dial lamps were dim, that's because they are 14.4v lamps and the supply is 6.3vac.

   All tubes were tested, one 6BE6 was very weak and was replaced.  All tube pins were straightened and DeOxit applied as they were re-installed.  I replaced the 3 9-pin tube shields that were missing with IERC heat dissipating shields like the ones that had been installed on many of the 7-pin tubes.  Several 7-pin shields are still needed, I do not have any more IERC's of that size.

  A 3-wire power cord was installed, along with proper 250vac rated X-Y bypass capacitors.

   A 10 watt wire wound resistor can be seen standing up from the panel just to the right of the new orange caps.  It's an addition by a previous owner, and is faulty, open near one end where a tap is for a bias/AVC line.  The bias voltage was measured when the rcvr was powered up after re-installing the front panel and all tubes, and was found to be -70 vdc.  It should be between -60 and -56.  This is a common problem on the R-388/51J-3,4 receivers, partly due to the increased AC line voltage we now have.  There are a couple of minor changes that can be made to bring it more into line, plus I will have to trace out the wire that is on the extra resistor tap, and remove or replace the resistor.  The low sensitivity is at least partially due to the high negative voltage on the AVC.  The calibrator signal is strong on the lower bands.  The sensitivity changes abruptly after a few minutes of warmup, and does change back and forth some for another few minutes before just staying poor.

   There is an intermittent "crackling" or noise, which I can't make to be related to anything, might be a tough one to chase down.  Before I removed the bathtub caps, shown below, I did replace C-204 & C-226, 100pf micas, in the AVC section, hoping that might cure the intermittent stuff, it didn't.  I've seen similar symptoms which were due to a bad C-204.

C-204 is a mica coupling capacitor at the AVC detector, it is a common failure item.

  C-204 & C-226 original & replaced.  The black loop is a cable for the prod. det. installation, not connected at this end yet.

  5 bathtubs removed and cleaned out      5 of 7 re-stuffed

  All 7 back in   There are 7 bathtubs, with a total of 11 sections; 4 electrolytics, 7 bypasses, paper originally, now polyester.

   I was hoping that replacement of the bathtubs would result in 2 things, better bias voltage, and no more crackling/change in sensitivity.  The bias voltage is developed by voltage drop over resistors in the center tap of the B+ winding of the power transformer.  That is dependent upon the B+ current draw, which hopefully would come down some with new bypass capacitors.  It did slightly, from -70 to -68. Still not low enough, and that will be fixed.  It still has some crackling and change in sensitivity, but in preliminary alignment adjustments it may be that some of that is caused by loose or broken wiring at L-115, in the first IF.  That may be the one that had the slug which had fallen out.  Overall sensitivity is still a bit low on band 1, low BC band, and probably the others.  I may have to resort to signal tracing with the scope to chase these down.

Most of the "crackling sound" was found to be caused by a 6BA6 tube in the filter module, V-302.  It had tested good, but had not been tested for noise.  Some of the changes in sensitivity, and some crackling was also due to a bad internal ground connection to the external ground post in the filter module.  A previous owner had added an external jumper wire from the top of the module to the top of the xtal filter can, as seen in one of the early pix.  I installed a hidden ground strap below, to avoid the difficult job of removing the filter module.  There is no secure mechanical connection between the filter module and the chassis, as originally designed.  The mechanical filter module was originally designed by Collins as an "kit" addition to the R-388/51J-3 model, to add mechanical filters to a receiver that did not have them originally.  When they produced the 51J-4, to include the mechanical filters, the merely used the additional box, as if it was an afterthought modification, and did not re-configure the chassis.

A product detector is to be installed, basically the circuit of Dallas Lankford, similar in many ways to those of Orr & Lee.  The 6BA6 BFO tube, V-114, is replaced by a 6BE6 which is used as the BFO oscillator and a product detector.  The original diode detector is retained, so AM reception is not affected.  

Components stripped out for product detector installation, one wire from T-104 has been relocated on V-110.

Product detector almost complete, at V-114.

It was necessary, as noted by Lankford and others, to custom select values for the coupling capacitor and shunt resistor that feed the IF signal to the product detector.  This is made necessary due to differences in individual receivers.  He had found that a 3mmf coupling cap and 8200 ohm resistor were optimum for him.   5mmf & 15k were necessary here, not inconsistent with what others have done.  This is done to match the audio output levels when using the prod. det. vs. the AM diode det.  Audio level was insufficient using Lankford's original values.

Some modifications to the AGC are also needed, the attack time is reduced to eliminate "popping" on CW reception, and the hang time is increased to eliminate, or reduce, "pumping".

  some of the AGC mod. at the AGC amp. tube, V-110, next to V-114, the prod. det./BFO, before addition of the 1.5mf "hang time" cap.  Lee's attack time mod. was used, with a 820k ohm resistor replacing R144, and bridged with a 2.2k res. & diode.

  Close-up and overall shots of the completed AGC & Prod. Det. installation.   It pretty much covers up V-114, AGC amp., and V-110, prod. det/BFO tube sockets. 

Here's the Prod. Det. & AGC info  

With things all back together and operating well, I still needed to chase down the low sensitivity problem.  Compared to the "good" 51J-4, broadcast band signals, for example, were some 20-30 db less on this receiver.  I replaced the 1st RF tube, a 6AK5, with a new one, and it helped on higher bands, but only slightly.  Following the schematic, and checking components, I finally suspected the standby relay, which switches the antenna from the 1st RF stage to ground when on stdby.  After cleaning the contacts with a strip of paper soaked with DeOxit, the receiver came alive.  This relay is never operated unless it is connected externally.  Occasional operation would keep the contacts clean.

I  have put the new aftermarket 6kc filter in, had to put solder on the pins.  The provider didn't do a real nice job for pins - he used small dia. threaded brass screws or rod, abt #2 or #3.  Just too small.  He had already put a little solder right at the tips, just abt enuf to fill in the threads.  Needed a little more, in dia. and further down. It does make BC signals sound nicer than the 3.1kc stock filter.

I started to do some touch up on the front panel scratches, but my small bottle of paint had dried up and is gooey.   The next day I used some of my St. James Gray cabinet paint, much nicer.

Have put the bottom cover back on, with all screws, will do a final alignment with it in place. 

Then I think this one's done.

------------------------

The above have been done, touchup looks much better.                                                      and as a reminder of earlier stuff,

        before                    after                             before                   after                                before                   after

Here's the minor mod. needed at the back of the cabinet to let the rcvr fit in.  Notch out skirt to clear the studs for the top cover.  I think the cabinet is for a 75A3 receiver.