What splitter for this design ?

[Jaap]

I like to try this design (see picture, comes from Glass Audio). The question is what splitter will perform good at the front of it. I need some gain because my tweaked cd-player has a weak signal (less than 2 volts). I am not good at tube theory so I need something I can copy easily.

I like to use the ecc40 (is like a e80cc) but other options are also possible (ef86, 12hg7, 5842, 6c45, 6h30, 2c51, 76, 6n1p, 6n6p, D3a, etc.).

I am thinking about transformer splitting (I have a cinemag) and then a differential stage. Or perhaps a LTP with ccs under it or a tube followed by a concertina splitter. Any suggestions or suitable designs that fit before this:

[Shoog]

It takes a very good (read expensive) transformer to perform step up duties. This is because they are intriniscally high capacitance items. Fortunately an input transformer is about the easiest position to drive so you may get away with a 1:1+1 input transformer.
A CCS LTP driver stage, driven from only one input can only ever achieve half the Mu of the input valve. However you could put a transformer infront and drive both inputs of the LTP and so achieve the whole Mu of the triode.

Another thing to consider is creating a Pentode LTP at the front which would give you massive gain.

Unfortunately you really need to work back from your anticipated output power to the front end and establish how much gain you actually need. You may be surprised that you might need very little and all options are open to you.

Shoog

[ray_moth]

I think Shoog is right about not needing much more gain, and I suspect that an LTP splitter, using your ECC40 with CCS in the tail, would do the job well.

The CCS could be made up of a couple of PNP transistors in cascode but you'll need a negative supply for that. However, since you'll need a negative bias supply for the OP tubes anyway, maybe it can serve both purposes.

[SY]

It looks like the gain to the transformer primary will be around 8 or 9, so you'll need a gain of 20 or 30 out of the input stage/phase splitter. A very well balanced input transformer feeding a diff amp will indeed do nicely.

[Boris_The_Blade]

Driving those 6CB6s should be really easy. Why not just use a simple pentode (or triode) driving a cathodyne splitter? ...especially since this is a learning exercise as well. This should have better bandwidth (and distortion if done right) than any transformer too.

[JLH]

Originally posted by Boris_The_Blade
Driving those 6CB6s should be really easy. Why not just use a simple pentode (or triode) driving a cathodyne splitter? ...especially since this is a learning exercise as well. This should have better bandwidth (and distortion if done right) than any transformer too.

The reason not to use a cathodyne phase spliter is it has asymmetrical output impedance. The phase taken from the cathode has a much lower output impedance than the phase at the plate. Under load, the higher impedance plate phase losses gain which results in asymmetrical drive voltages to the output tube grids. The LTP splitter is the better way to go.

Rgs, JLH

[ray_moth]

Correction to my post above: the CCS should use NPN transistors, not PNP (but I bet you knew that anyway!)

[SY]

Originally posted by JLH



The reason not to use a cathodyne phase spliter is it has asymmetrical output impedance. The phase taken from the cathode has a much lower output impedance than the phase at the plate. Under load, the higher impedance plate phase losses gain which results in asymmetrical drive voltages to the output tube grids. The LTP splitter is the better way to go.

This is a canard that has been exploded over and over. If the cathodyne is loaded symmetrically (as would be the case here), the frequency and phase response at both outputs will be identical. The apparent source resistance at both outputs will be identical and low.

[ray_moth]

Detonated ducks notwithstanding, I would tend to support the idea of an LTP splitter on the grounds of its far better PSRR.

[ja2dhc]

Originally posted by Jaap

I am thinking about transformer splitting (I have a cinemag) and then a differential stage. Or perhaps a LTP with ccs under it or a tube followed by a concertina splitter.

I am using a differential amplifier by a pair of Hi-gm of C3g. The
schematic is seen in my Web below.

http://ja1cty.servehttp.com/6080/60802.html

Because of Hi-gm, a pair of C3g can drive -70V biased 6AS7 to
the full swing where the input voltage is less than 1V.

As you can see, it resembles to yours but the cathodes are
connected and pulled down to -160V via 22Kohm as is intended
to be a quasi CCS. If you will use transistor CCS with about
-15V, -160V is not necessary. -15V can be obtained by
AC 6.3V and voltage doubler rectifying.

If your will use Hi-gm pentode such as C3g, D3a, E180F, 12HG7
and E280F, enough voltage gain will be obtained to drive EL34
without input transformer splitter.