Potentiometer value for 6SL7 input

[mahleriana]

I have a ALPS 100k log input pot then a 6SL7 with a plate load from 100k to 330k (but usually 100k)

The sound quality really increases when i pass the first third of the pot course (more open, better bass) but then is quite loud even with 90dB full range (2A3 output).

Is the pot value too low for the 6SL7 ?

thanks in advance.

eric

[Eli Duttman]

Eric,

Changing pots. will not help. Ratio, not absolute value, governs in a resistive voltage divider. Padding the pot. with 100 KOhms of fixed resistance will attenuate the I/P signal by 6 dB., when the control is set for max. volume.

[leadbelly]

Well no psychics have responded yet so I guess you're just going to have actually give us information.

What does the circuit look like? Do you have a schematic? Is the grid DC coupled? Is there a grid resistor? The pot is wired correctly? The source is OK with other amps?

EDIT: Ooops, beat me to it Eli.

[Miles Prower]

Originally posted by mahleriana
I have a ALPS 100k log input pot then a 6SL7 with a plate load from 100k to 330k (but usually 100k)

The sound quality really increases when i pass the first third of the pot course (more open, better bass) but then is quite loud even with 90dB full range (2A3 output).

Is the pot value too low for the 6SL7 ?

If anything, I'd say that a 100K pot is too high. Here's the spec sheet data for the 6SL7:

Cgk= 2.15pF
Crt= 3.5pF

Given the plate loads you give, you're looking at a voltage gain of ~40 -- ~50. If Av= 40, then:

Cmiller= 3.5(40 + 1)= 143.5pF

Ci= 143.5 + 2.15 + 10= 155.65pF (Adding an additional 10pF for Cstray -- this might be even higher, but you probably can't get it much lower)

Since:

w= 1 / RC

w= 1 / (100E3 X 155.65E-12)= 64.25E3 rad/sec= 10.23KHz

That costs you quite a bit of your highs. So the result you describe is quite understandable: you have better bass since you've made an accidental treble cut circuit in addition to a volume control.

That's the big problem with those high gain triodes: big Cmiller's. I usually avoid the problem by making volume controls from 100K linear pots with the slider paralleled to the ground end with a 16K resistor to make a log pot function that way. It not only gives a guard resistor in case the pot goes open, it also gets that resistance down so that you get the low pass pole out of the audio band. I've also found that log taper pots go scratchy pretty quick, whereas linear pots do not.

[kevinkr]

Actually the maximum resistance of a 100K pot relative to the grid assuming a reasonably good voltage source driving it is at electrical mid point, the thevenin equivalent source resistance of said pot at mid point is then 25Kohms giving a -3dB at about 40kHz. (Still not that spectacular.)

Still I think Miles is right about the pot, however it may be a quality issue as well as a parametric one. Pots have lots of stray capacitance all on their own, and depending on how it is distributed you may get quite poor frequency response over 1/3rd to 1/2 of its rotation with the best measured performance happening over a range very close to maximum volume, and at the other volume extreme near maximum attenuation. This neglects other sources of stray capacitance like long wires, etc.

I would recommend a 50K pot or less if the driving source is solid state, if tube 50K/100K would be reasonable.

I prefer stepped attenuators to any pots I have heard or can afford.

[mahleriana]

Effectively, the best sound is achieved with the using the 100k ALPS pot at maximum volume.
Amplifier is fed directly from CD player.
I have also in stock a good quality 50k TKD KO-ON pot i was using with a trends TA 10.1.
I tried it a few months ago and was not convinced then, but it is worth a new try...the amplifier is not exactly the same.
Steeped pot with high quality resistors could certainly be a good choice, but as they are not cheap i need first to be sure of the value and the kind of resistors i like most (and since resistors matching and mixing is haute cuisine, it could be the most difficult point).

[Nihilist]

Read this thread and try it out. Works pretty good.

Passive Preamp


.....................Blake