Still banging away on the pentode ..
It has a gain varying from around +32dB in full pentode mode, down to around +24dB in full triode mode.
Now that's a lot to throw in the path, especially when you have some pots'n'switches hanging off the plate and screen - hiZ = lotsa noise!
After trying a few things, I ended up using a pair of 2-core shielded + foil as hookup for the controls.
I have the shield+foil grounded at both ends, in this case.
Now that's made a very big positive difference in the noise levels related to the control wiring.
Something like a 15dB improvement in headline noise, mostly at hum freqs but with some hiss reduction too.
With the pentode stage switched out, I now have my baseline noisefloor of around -82dBu or so.
With the pentode in +32dB I have a noisefloor of around -72dBu
So that's pretty good and should improve another dB or two when I have the controls mounted on the alu face plate rather than lolling about on the table
So - when dealing with hiZ, high gain -> shield + foil grounded is best!
---
That's a big milestone because I've always had big issues at the pentode gain stage with noise.
Of course, it's quite microphonic as one would expect! I may try some tube dampening rings.
----
The final checks are with pentode in, Redd 47 at unity gain with -10dBu line input
+32dB gain -> 'Drive' pot -> AU7 and line amp -> T-pad
with the T-pad at min attenuation and the 'Drive' set for unity gain at output
I get my nominal baseline noisefloor of -81dBu
Same thing but with the t-pad turned way down for 'strong atten' and the 'Drive' way up, even better at -82dBu.
So that's a wrap.
----
Now I have a working 'max pentode gain' scenario, I can go forward and figure what fixed atten to put after the pentode.
THEN, re-connect the EQ circuit and that's the hard part done.
It has a gain varying from around +32dB in full pentode mode, down to around +24dB in full triode mode.
Now that's a lot to throw in the path, especially when you have some pots'n'switches hanging off the plate and screen - hiZ = lotsa noise!
After trying a few things, I ended up using a pair of 2-core shielded + foil as hookup for the controls.
I have the shield+foil grounded at both ends, in this case.
Now that's made a very big positive difference in the noise levels related to the control wiring.
Something like a 15dB improvement in headline noise, mostly at hum freqs but with some hiss reduction too.
With the pentode stage switched out, I now have my baseline noisefloor of around -82dBu or so.
With the pentode in +32dB I have a noisefloor of around -72dBu
So that's pretty good and should improve another dB or two when I have the controls mounted on the alu face plate rather than lolling about on the table
So - when dealing with hiZ, high gain -> shield + foil grounded is best!
---
That's a big milestone because I've always had big issues at the pentode gain stage with noise.
Of course, it's quite microphonic as one would expect! I may try some tube dampening rings.
----
The final checks are with pentode in, Redd 47 at unity gain with -10dBu line input
+32dB gain -> 'Drive' pot -> AU7 and line amp -> T-pad
with the T-pad at min attenuation and the 'Drive' set for unity gain at output
I get my nominal baseline noisefloor of -81dBu
Same thing but with the t-pad turned way down for 'strong atten' and the 'Drive' way up, even better at -82dBu.
So that's a wrap.
----
Now I have a working 'max pentode gain' scenario, I can go forward and figure what fixed atten to put after the pentode.
THEN, re-connect the EQ circuit and that's the hard part done.