Baffling rf/ground loop issue

[luthierwnc]

Hi all,  I'm rebuilding an old Masco MA 25 into something else (schematic attached).  In doing so I tried to keep the same point-to-point layout.  Ground points have been reduced from a dozen to three.  It took a while to chase down a 6.3v hum which still leaves me with a wicked buzz.  I've mainly made amps using Fender-style or garolite boards and haven't honed my chops on crowded point-to-point circuits.  Using a test probe with a .047/630v cap in series between ground and the tip the buzz is greatly reduced after the first two coupling caps and completely eliminated where they join at the 470k mixer -- along with any signal.  The preamp and PI tubes are on the 6v DC supply and the 6L6s are AC.  I wonder if the first tube is too close to the input jack and the bare wires of the grid resistors are picking up interference.  The hotter .0047 side is slightly noisier.

At any rate, if anyone can spot what I'm doing wrong, it will be a relief.  Also attached is a closeup of the two 6SL7s and the scope at the 470k junction; 5ms, 5mv with no signal.
Cheers, Skip

[PRR]

You have one side of the DC heat grounded, and the center of the AC heat grounded (to cathodes!!!). You can't have it both ways at once. Power tube heat is not a big deal, while preamp heat is fussy.

[luthierwnc]

Sorry, working on the 6.3v system was another headache.  It was originally done as drawn with the power tube cathode tied to an artificial filament center-tap.  It sounded like a garbage truck.  I then tied the cathode and 100R resistors to a chassis ground and it got slightly better.  Then I just lifted the 100R resistors with no ground reference at all and that reduced the filament hum quite a bit -- not completely but definitely the right direction.  Voltages on both the AC and DC wires show 6.6vac and 6.4vdc respectively.

The original schematic for this amp shows one leg of the 6.3v winding going to the cathode but no center-tap one way or the other.  That gets a little more confusing because there are two taps on the power transformer that go the the ground on the output transformer terminal block along with a third wire from that junction soldered to the chassis.  I wondered that one of those might be a filament center-tap not shown on the schematic.  The light green arrows leaving the DC supply only go to the filament lugs on the last two tubes.  They aren't grounded.

Both noises are problems.  Reducing the hum didn't get rid of the buzz.  Never a dull moment.  Thanks for any more ideas, Skip

[d95err]

Can you clarify more about your grounding? I don’t understand from the schematic how the ground points are connected or where they connect to the chassis.

First thing I would do is to loose the DC heaters. They very often seem to cause more problems than they solve (judging by thresds on amp forums). In this type of low-gain amp, I don’t think there would be much benefit anyway.

[pdf64]

I was thinking to ditch the DC heaters too, but 6SL7 are unknown territory for me, so dunno.

[sluckey]

I have two amps that use 6SL7s. Both work very well with AC filaments.

[luthierwnc]

Getting rid of the DC worked -- possibly with other problems fixed.  The power tube cathode is tied to the heater leads with 100R resistors.  It is still a bit noisy but the shop is in a basement with fluorescents and I leave a P90 guitar down there -- in part because if I can get it clean enough, everything else will be better.  Getting the chassis in the case should help too.
I tried it with a 6SN7 in the #1 hole too but it didn't make much difference.'As always, thanks for your patience and expertise.  sh