Fender "UL" twin Adjustable bias question

[goldstache]

Well, I was recapping one of the these beasts recently and when I went to bias it in.  I was shocked that the power tubes are pulling about 12ma a piece.  Well I thought, maybe adjustable fixed bias would be a better option.  I changed the bias balance to more of an adjustable topology.  I changed to a single feed to the bias supply grid resistors to 220K (they were 47k and 68k).  Changed the resistor feeding the diode for the supply tap.


Fired it up and now with the stock 33K bias resistor on the pot i was getting 25ma per tube.  Better?  Well I couldn't crank it, so Im not sure yet.  One thing that I did notice is that the noise floor in the power section (plate and screen filters) was way more noticeable.  Is that why they had it biased so cold?  It being UL, I guess the design goal was to stay clean and loud and that's probably why they boast 135W RMS.  The thing won't want to clip with that bias. Which is fine. I just want it a little livelier.  I don't want to modify the circuit (Preamp, NFB, etc)  I just want to understand why they set it so cold and what my options are. 
I've done adjustable fixed bias with balance before, but my buddy doesn't want extra holes in her.
I tried a few different quartets of tubes and they all pulled within 6-7ma difference per set.


I guess my questions are:


-Should I convert back to stock bias balance with stock values and change the bias feed resistors (off tap and on balance pot) to achieve a better bias level. 
-As I bias hotter in either configuration (balance/adjustable fixed) will the perceivable noise floor always raise to an unacceptable level (IMO)  Ripple is pretty low as measured.
-Anybody got any crafty ideas on how to keep here hum low and the tubes a little hotter and more responsive??
Thanks

[Tony Bones]

By noise floor, do you mean hum or hiss? The only reason I can think of for hiss to increase with current is if one or more of the screen resistors is noisy. Maybe try replacing them all with metal oxide or wirewound?

[shooter]

will the perceivable noise floor always raise

noise floor is typically constant, but each stage will amplify it, so if you increased gain of the PA, you increased the amplified noise

[goldstache]

Good idea Tony.  Thanks
This amp got new tube sockets and all associated on the socket parts (grid resistors, screen, etc.)  Humm is what I'm getting.  With the stock bias balance  it was quiet, probably because the output section was biased cool and output was lower. 


Also, bear in mind that this amp is in UL mode, so the only filter to the OT is the reservoir (most ripple of all).  I do have 470R 3W in there for the screens UL tap>470R>Screen pin. 


So some of the hummmm ripple could be present there as well, since there isn't any additional filtering like a screen supply would benefit from. 


Starting to lean toward going back to stock and letting the 6L6's only eat up 12ma a piece.