Surprising volume change when changing SELECTIVITY on Phantom FM

[Eddie Watts]

Thanks to Jim Sargent (of Jim Sargent Auctions), I acquired my first Scott, a Phantom Deluxe FM (model BFM, sn ZZ455) receiver and power supply which became available in Lubbock TX where I live. 

After making initial repairs so I could make "as found" measurements - I discovered obvious volume changes when I change SELECTIVITY.  Tuned to noise between stations, using SHARP as my 0db volume reference at the output transformer, the noise is +17db louder when switched to MEDIUM, then drops a little to +8db when switched to BROAD.  On a local station, the volume is only +10db louder when switched to MEDIUM, and is +3db louder in BROAD - still an obvious change in volume.  Another Lubbockite with a model BFM says his does about the same.  I wouldn't have expected this from a Scott - is this normal?  I saw a You-Tube of a Phantom which seems to hold volume steady as you switch SELECTIVITY - but that could just be his recording AGC.  

I did not want to alter alignment until I have time to sweep the IF (maybe I'm going overboard on how much "as found" data I want).  I did do a 1kHz plot of the IF BW, but still want to sweep it.  BY THE WAY - I got a spare HP141T spectrum analyzer from a Jim Sargent auction to complement my '141 outfit - mine is waiting  for time to address an HV repair.  

Here is my progress for those who are interested:  

November 2020 I acquired the radio.  Replacing a blown 3A fuse and running it up on a Variac, it wouldn't maintain B+ at all.  The power transformer had been replaced with an oversized one, with a 143 ohm dropping resistor in the HV CT (before the bias dropping resistor).  The resistor was intermittent.  Long story short - I replaced it with a 125 ohm 20W R selected to maintain proper B+ based on the schematic on this website, and a little data in the "Dean of DX" book.  The Candohm had been replaced with a large multitap Ohmite, which I slightly readjusted to the correct tap voltages after getting proper receiver chassis B+:  230/150/125/75 VDC.  

I was relieved to measure reasonable resistances on all the chokes and transformers, including the IF tertiary windings.  

My most appropriate speaker (so far) is a Jensen 12" PA speaker with 1k ohm FC with no tap for the 6L6G screens.  I installed a temporary 10k ohm 20W resistor for the 6L6G screens, selected to keep them safe below 270V max rating.   

CHROME CLEANING:  I discovered a safe way to lift the grime off the chrome - I lightly "flood" the grime with Fantastic, then lightly brush that off with a deep microfiber washcloth, leaving no visible scratches.  I did the can tops - will get to the chassis later.  

I tested all the tubes and replaced any bad or very weak or badly mismatched ones - and a couple of missing FM tubes, then logged initial ~~"as found" tube socket voltages.  

Noted a possible error on the Phantom schematic - the VR150 CANNOT run enough current fed from the "150" tap (even if it fires), it must use full +230V B+ (which mine does), and even then it runs below normal minimum VR150 current with its 6k ohm supply R, with the LO taking about 6mA.  

I was able to power up the radio and confirm it is basically operational - great relief.  I found the SCRATCH FILTER to function well.  Both rotary switches had to be cleaned, especially the bandswitch, so I carefully dabbed Caig R100 onto the contacts with a tiny foam Q-tip, trying to keep the Caig mostly off the phenolic.  Yes - - it took awhile - but years ago I'd had a bad experience flooding a bandswitch with R5 spray, then powering it up before completely drying.  

(The bandswitch experience was on a Hallicrafters S-40A.  Several minutes after flooding the bandswitch with R5, it began "ticking" increasingly rapidly.  Later I saw an arc growing on a wafer which had B+ on it.  I was lucky to find it had an identical wafer for the antenna transformers, so I scraped the arced wafer down to >1M ohm leakeage, then swapped those 2 wafers to save the radio.  Nowadays if I flood a bandswitch with R5, I allow it to dry 2 weeks before powering it up, if it has B+ on it.  BUT I DIDN'T FLOOD MY NEW SCOTT BANDSWITCH!!!)

My SENSITIVITY control is dirty and will require more aggressive Caig cleaning on the inside.  Meanwhile I have to keep bumping it to make sure it's really at maximum.  I might have to respring the wiper.  

While studying the VR150 and LO regulation, I noticed fluctuations of a few volts with some frequency fluctuations.  On BAND D the LO made occasional jumps up to >200 kHz.  Later, I also noticed BAND C occasionally jumps up to 80 kHz - but BAND A and B were absolutely stable.  I tried 3 different 6J5G's for the LO, but the problem remained.  For a long time I couldn't find any mechanically sensitive parts in the LO by tapping with a wooden dowel, and was about to lift the LO can and the tuning capacitor shield for a look.  ALSO - I had noticed "crackles" when tuning 10.8 - 12.8 MHz on BAND D, and near the same place on the dial on BAND C. making me wonder if the tuning C grounding clips needed cleaning - another reason to look inside.  AND the background noise gets dramatically quieter near that part of the dial.  BUT BACK TO THE LO:  I had found a pair of mica capacitor leads intermittently shorting (one involved the LO grid), so moved them - but that was only a small part of the problem.  I finally evoked a big frequency shift by hitting a lead of the BAND D tank coupling capacitor, a mica, just exactly right.  Previously, it had seemed stable no matter how I prodded it.  I temporarily replaced it with a same size glass capacitor, and ordered a same size mica (just to keep it more original).  I also replaced the tank coupling C for BAND C, and now both bands are stable (except for a little slow drift the first half hour of warmup - might have to look into that).  By the way, BOTH these mica C's are about half the value shown on the old Phantom Deluxe schematic:  .0013uF instead of .0025uF for BAND C, and BAND D is .0025uF instead of .006uF.  

Someone had tried to align the RF on BAND D with that bad coupling C, even to the point of loosening some turns on the LO coil and spreading them quite a distance (per Scott service sheets on this site).  With the LO repaired, and RF touch-up on just BAND C and D, it had very poor tracking - so I'll have to undo that coil-stretching and tack the turns back down again with Q-Dope - if I can still find any.  I found I can restore good tracking by pressing the turns back near to where they were - but I might have to remove the LO can to reach far enough.  What glue do y'all use if you don't have Q-Dope?  THIS heavy wire will take more than candle wax to hold it locked in place.  

I had also discovered a badly bent bandswitch contact on the 2nd wafer from the front (LO), and found it is used to short an unused coil when switched to BAND D.  It looks like it was done deliberately, as if someone changed their mind on whether that unused coil should be shorted or left open when using BAND D - perhaps even from the factory(?).  I stopped worrying about it - the contact had been bent completely out of the way as if to deliberately make sure it couldn't make contact, and there was no other damage.  There was no obvious change in BAND D when I shorted those contacts together.  

Quick-check of FM, after cleaning dirty pins on a couple of Loktal tubes, shows FM "sensitivity" of about 0.3mV (takes 0.3mV to hear the audible tone modulated at +/- 75kHz) - probably some work to do there.  AND - I was happy to discover this Scott FM tuner evidently can receive the lower half of the current FM BC band - same as I have noticed on a simpler Philco 42-390 - which means I can receive the local classical station on this radio without a converter.  

After all that, I plotted IF response at 1 kHz intervals and IF -6db bandwidth to be 3.8kHz in SHARP, 9.2kHz in MEDIUM, and 12.7kHz in BROAD.  I figured BROAD would be wider - this only gives just over 6kHz audio BW.  Any comments?  (I've heard some 1930's AM stations were allowed to broadcast more than a 10kHz wide channel.  I know the Philharmonic boasts 16kHz IF BW.)  

That's where I am now until I have more time to work on it.  Besides the 2 chasses, I do have the front panel with the correct escutcheons, but still hunting appropriate knobs (especially coarse tuning) and correct speaker.  Do I understand that the correct standard speaker is a 15" non-pedestal with a "treble disperser" that looks like a whizzer cone?  (I'm aware there was an optional 4-piece hi-fi speaker.)  And I still need 3 chrome upper tube shields.  And I might like to find a Warrington cabinet for it someday ...

[David C. Poland]

HI Eddie. Welcome to this Scott Enthusiasts site. And congrats on your first Scott. Post a photo.

I can't help you on the FM band having avoided the pre-war FM sets.

Knobs are walnut with an "S" embossed on the knobs, except for the higher speed coarse round walnut that sits behind slower speed "S" tuning knob. I can supply the coarse knob ([e-mail])

Scott supplied either a 12 inch or 15 inch standard speaker, depending on the cabinet selected. FYI, the output transformer will be mounted on the basket and the terminal cover will say Magnavox. The standard speaker VC impedance is 38 ohms. The single field coil is 925 ohms and tap at 515 ohms.

The standard speaker is not electrically suitable to use with the optional 4 unit high fidelity speaker system. 

Scott offered a number of cabinets besides the Warrington that fit the FM Phantom Deluxe. For example the Accousticraft and Braemer.