2A3 pp biasing schemes compared.

[James D.]

I finally got to do some trials of diode, cathode and fixed biasing schemes for my 2A3 pp amp.

Haven't got all the bits for my design so I used parallel 6SN7s to drive the 2A3s through a SE to PP interstage.

Anyway my results are:

Diode bias - The worst sound. Not bad but a bit one note in the bass and seemed to produce lower resolution than the other ways.

Cathode Bias - bypassed.

Pretty good more open and better detail. More tuneful bass

Fixed bias.

The best result. Nice open sound with deep tuneful bass. Better sense of PRAT and better upper midrange.

I was a bit suprised by this and expected the diode or cathode bias to be best... Any views or suggestions or rude comments welcome...

ciao

James

[dhaen]

Hi James,

I believe that fixed bias will produce the lowest ra, so the available inductance will be most adequate in this mode.
Is the bias arrangement common to both valves?


Cheers,

[James D.]

Hi John,

Is the bias arrangement common to both valves?

For the cathode and diode bias, yes, this was connected to the wiper of a small value pot (47R) that connected the two cathodes together. This pot was used to adjust the balance between the two 2A3s for best sound quality...

The fixed bias was applied to the ct of the interstage transformer with the wiper of the cathode pot applied to earth.

The cathode to plate (damm! sorry...) the cathode to anode voltage was set to 275V for each case, grid to cathode voltage was set to -48V for each case - I assume the current through the valves was the same in each case...

As expected the cathode pot was set with 20R one side of the wiper and 27R the other for best sound. This varied by a maximum of 1R between different bias cases.

BTW the OPT has a load impedance of 5K for each side... wouldn't this minimise the effect of lowered ra? My output transformer also has taps at 3K loading each side and this produced the same results when comparing cathode against fixed bias so maybe it has about the same effect at all sensible loadings...

ciao

James

[Joel]

You'll get no argument from me James! I use fixed bias wherever I can, especially batteries.

Did I mention my button-cell trick? You can make almost any value of bias voltage by stacking the small calculator lithium cells end-to-end, and wrapping them in tape. I've made several 40V batteries this way for my various 71-A experiments and amps. I'm going to make a 30V cell for my #31 push-pull project.

[ThorstenL]

Hi,

I finally got to do some trials of diode, cathode and fixed biasing schemes for my 2A3 pp amp.

Haven't got all the bits for my design so I used parallel 6SN7s to drive the 2A3s through a SE to PP interstage.

Anyway my results are:

Fixed bias has advantages. One key disadvantage is the tendency of the operating point to shift over time and the general lack of long term stability. You can more or less match the sonics of the Fixed bias approach if you omit the cathode bypass capacitor and instead use a high quality, moderate value (20uF+) capacitor from the cathode bias resistor top (aka valves cathodes) to the +B junction feeding the OPT (aka "ultrapath").

If using and IT it may also pay to add a bootstrap circuit from the cathodes to ground connection of the centertap of the IT, as done in the WE Amplifiers...

Sayonara

[Sch3mat1c]

Or use...say...LOW ESR/L CAPS for the final PS cap and the bypass. (As if you could hear 1 ohm of difference over 500 ohms of bias anyway.) Two caps in series...

Tim

[James D.]

Damm! No argument from Joel!!!

Did I mention my button-cell trick? You can make almost any value of bias voltage by stacking the small calculator lithium cells end-to-end, and wrapping them in tape. I've made several 40V batteries this way for my various 71-A experiments and amps. I'm going to make a 30V cell for my #31 push-pull project.

Neat!!! Thanks for the tip
A 31 push-pull amp! Another valve to look up Can't wait to see the design...

KWC said:

Fixed bias has advantages. One key disadvantage is the tendency of the operating point to shift over time and the general lack of long term stability. You can more or less match the sonics of the Fixed bias approach if you omit the cathode bypass capacitor and instead use a high quality, moderate value (20uF+) capacitor from the cathode bias resistor top (aka valves cathodes) to the +B junction feeding the OPT (aka "ultrapath").

and

If using and IT it may also pay to add a bootstrap circuit from the cathodes to ground connection of the centertap of the IT, as done in the WE Amplifiers...

Thank you for the pointers and suggestions. One day I hope to know as much as you have forgotten about valve amps

ciao

James]

[ThorstenL]

Hi,

KWC said:

.......

Thank you for the pointers and suggestions. One day I hope to know as much as you have forgotten about valve amps

Actually, we all have forgotten. Credit is due to the Engineers in those days way back in the 1930's with Western Electric and RCA in the US, Siemens/Telefunken and Klangfilm in Germany, to name those I know. I'm sure France had it's greats as has the UK with unjustly forgotten GEC and Decca and likely many others.

Studying old WE and Klangfilm schematics AND tryingto understand the reasons for given applications can be HIGHLY instructive.

Take for example the output stage of the WE 92B Amplifier....



This illustrates a few neat tricks:

1) While the amplifier works in Class A the cross connected 15pF (15mmf - micro micro farad - a micro of a micro) will offset most of the input capcitance. Especially when using high anode impedance coupling Transformers witha lot of stepup the Miller capacitance can wreak havoc on the bandwidth, these two critters make the problem less. Capacitor Value is appx. equal to the ANode-Grid Capacitance, a little lower is advisable in practice.

2) Note the Output stage cathode bypass. 2 pcs 20uF Capacitors in series for 10uF, NOT from the top of the Cathode resistor to ground, but to +B (aka Ultrapath). 10uF is still rather on the wee side, I must note....

3) Note the RC circuit coupling between the 300B Cathodes? As long as the output stage retains perfect balance this circuit does nothing, but if the cathode starts to show any AC signal this circuits "feed-forward" (s) the error to the Grid circuit, thus especially linearising LF response, but also symetry errors! Here we have in my view the first application of the so calld feed-forward error correction", probably before Harold Blacks "negative Feedback".

Hellfire, these olf cats had really ALL the bases covered. Much learning to be found in these.

It only really gets me that whenever I have come with a REALLY, REALLY neat and genial trick to do something unusual with Valve Amplifiers some smartalec comes crawling out of the woodwork and tells me:

"weeel, WE/RCA/Seimens/Telefunken/TAB/GEC/Decca.... did that 50/60/70 years ago I'll have ye know..."

(Underscore appropriate, cross out those that do not apply)



So, I hope the above gets some creative juices flowing.

Sayonara

[fdegrove]

Hi,

Thanks TL...

Indeed , you can twist you're mind any way you want to come up with something new just to find out someone else has beaten you to it.

What is even more amazing is that all this was achieved with only a fraction of the means we have at hand today.

Cheers,

[ThorstenL]

Hi,

Indeed , you can twist you're mind any way you want to come up with something new just to find out someone else has beaten you to it.

What is even more amazing is that all this was achieved with only a fraction of the means we have at hand today.

Yeah. This is one of the reasons why I have taken up speculative and operative Alchemy. Given that nobody really did anything serious since something like the 18th century and that a lot of knowledge was not only rejected but completely lost, plus the impact of modern Physics I figure there ought to be some milage in that old Lady Alchymia yet...



Sayonara