Please comment my F2a SE schematic

[Joku]

Hi

I have been thinking about building a single ended triode amplifier with F2a output tubes, using ECC40 as the preamp tube and EL95 as a cathode follower as the driver. So, I'd ask what you think about the following schematic:



I needed voltage amplification of about 60 to the grid of the F2a, so I needed to use both sections of the ECC40. F2a's input capacitance is quite high at 20kHz thanks to the Miller effect, about 150pF so I thought it would be better to use a cathode follower to drive the tube. To reduce the significance of the input capacitance of the second section of the ECC40, ECC40a's anode resistance is only 22k. I divided it to two parts, because otherwise I would have got around 160 voltage amplification which is a bit much. The output transformer will be a 10W Edcor GXSE 2.5k->6 ohms treansformer, which will appear as a 3.3k load to the tube when a 8 ohm speaker is used.

The current of ECC40a is around 3.8mA, and ECC40b around 1.4mA and EL95's current is about 10mA.

And here's the load line of the F2a, I0 and Vb is 80mA and 375V, max power should be about 5.1W at 2.7% second harmonic distortion:

[azazello]

Very long way to great sound....

[Klimon]

I wouldn't insist on using ecc40 because, as you illustrate, it forces you to insert three stages before the finals which is really a bad idea. I'ld start from scratch, look at the requirements a single stage should have to drive the f2a's, and proceed from there. As a side note, why use such high coupling cap values (even presuming you mean 0,47µ)?

Good luck,

Simon

[mach1]

Given the Euro theme, a single C3g (pentode or triode mode) would do the job admirably. Since Vk-g1 for the F2a is only around 14v, a C3m would also work well.

Also, you have shown a 22uF cap in series with a 220uF cap on the F2a cathode. I am not really sure what you are trying to achieve here: a single 220uF would probably be more appropriate.

[Joku]

Well, I just noticed that Vb for the F2a should be 350V, not 375V..

[Joku]

Originally posted by Klimon
I wouldn't insist on using ecc40 because, as you illustrate, it forces you to insert three stages before the finals which is really a bad idea. I'ld start from scratch, look at the requirements a single stage should have to drive the f2a's, and proceed from there. As a side note, why use such high coupling cap values (even presuming you mean 0,47µ)?

Good luck,

Simon

Well, I would need amplification of about 60 and output impedance of no more than 20k. I'm not sure if there is such a tube, at least I haven't found one. Of course, I could use a 1:4 signal matching transformer at input to reduce the need for gain..

Oh, and the 47uF capacitors are probably a bit large, but it shouldn't cause any problems. 4.7uF should probably be fine (impedance @20Hz -j1.7k) but .47uF is probably too small (impedance @20Hz -j17k) On the other hand, it's true that the input impedance of a tube is very large at low frequencies, so 0.47uF should be fine, too.

Originally posted by mach1
Given the Euro theme, a single C3g (pentode or triode mode) would do the job admirably. Since Vk-g1 for the F2a is only around 14v, a C3m would also work well.

Also, you have shown a 22uF cap in series with a 220uF cap on the F2a cathode. I am not really sure what you are trying to achieve here: a single 220uF would probably be more appropriate.

Well, line level audio signal is about 0.5Vpp, and F2a needs 28Vpp at its grid.. so I need about 60 amplification and C3g's aplification factor is only 40..

that 22uF 'cap' is a 22ohm resistor. Anyway, its point is to improve linearity without sacrificing too much gain..

[Klimon]

it shouldn't cause any problems.

Depends whether you consider lousy sound a problem

Well, line level audio signal is about 0.5Vpp, and F2a needs 28Vpp at its grid.. so I need about 60 amplification and C3g's aplification factor is only 40..

6GK5 for one has µ>60 and Rp<<20K http://www.mif.pg.gda.pl/homepages/f...135/6/6GK5.pdf ; then there are the usual "high transconductance - low Rp" triodes or -strappeds like 5842, 6c45, d3a, e180f/e280f, 8233 etc. would comply with ccs or choke load. C3g as pentode (and amongst other pentodes) would work also, e.g. at screen at 100v / 8k load / bias 200v; 13mA; grid -0,75V. I think those f2a's will reward you if driven well.

Simon

[MartinR]

Hello Joku,

I would use the C3m in pentode mode. Operating point as in the datasheet specified. With RL=12K and Av=gm x RL you would have a gain of about 72. Unfortunatley, you would need a higher B+ for the C3m. You could either use a second supply or why not use a anode choke parallel to the RL? At least that´s what I´ve planned until I decided to use the F2a in tetrode mode.

Kind regards,
Martin

[Joku]

Originally posted by Klimon
[B]

Depends whether you consider lousy sound a problem

Why a big coupling capacitor would cause lousy sound? Isn't this a case of bigger is better?

Originally posted by Klimon
6GK5 for one has µ>60 and Rp<<20K http://www.mif.pg.gda.pl/homepages/f...135/6/6GK5.pdf ; then there are the usual "high transconductance - low Rp" triodes or -strappeds like 5842, 6c45, d3a, e180f/e280f, 8233 etc. would comply with ccs or choke load. C3g as pentode (and amongst other pentodes) would work also, e.g. at screen at 100v / 8k load / bias 200v; 13mA; grid -0,75V. I think those f2a's will reward you if driven well.

Simon

I'll have to see if I have access to any of those tubes (I can use a rather large set of tubes for free for this project)..

Originally posted by MartinR
Hello Joku,

I would use the C3m in pentode mode. Operating point as in the datasheet specified. With RL=12K and Av=gm x RL you would have a gain of about 72. Unfortunatley, you would need a higher B+ for the C3m. You could either use a second supply or why not use a anode choke parallel to the RL? At least that´s what I´ve planned until I decided to use the F2a in tetrode mode.

Kind regards,
Martin

Btw, wouln't it be a bad idea to use a pentode in a tetrode amplifier? I mean, pentodes have more distortion and all..

[Joku]

Anyway, I got the idea to use the ECC40 as it seems to be very linear, at least according to this page:

http://www.geocities.com/la1zka/hifi/distortion.html

But it would be still better to use a single pentode rather than ECC40s and a cathode follower?

On the other hand, I could use a 1:2 matching transformer at the input (I'll probably have to use a matching transformer at input anyway to ensure that there is no galvanic connection to the ground), and then a single ECC40 or CV2543 and a cathode follower?