An UL preamp idea.

[Shoog]

Does anyone think the attached schematic has any potential as a viable preamp circuit.

Its just a rough without any component values.

Shoog

[Bandersnatch]

hey-Hey!!!,
Why not just make a TX coupled circuit? While you're at it, make a few tap %-age options and try a 2:1 step down, or perhaps a 1:1 bi-filar wind. Make the plate load appear just large enough to deliver the needed gain.
cheers,
Douglas

[planet10]

EF86.

Looks while worth a try concepually. I have all those bits in my junk box (and no transformer)

A trick with pentodes with separtely pinned suppressor grids is instead of tying it to the cathode, is to tie it 2-3V below the cathode (a 3.3V 1.2 AA computer battery would work here.

Th etechnical reasons made sense to me, don't ask me to remember what they were.

dave

[Shoog]

I may try building this circuit with the 6AQ5. I have heard people having great success with the 6V6 as a preamp tube, and the 6AQ5 is the miniture 7pin version.
I have a version of this amp at the moment using all PCC88's and a mosfet as the top element of the totem. It sounds very good, but I want to replace the mosfet and run more current through the totem. Also what I really want is for a sound with a touch more bottom end grunt and the 6AQ5 may just bring this. If the UL connection doesn't work as expected I can simple triode it and nothing is lost.

Shoog

[jane]

To me it looks like you are running the pentode with almost 100% g2 feedback (triode mode). The nice thing is that you have included the output stage in this feedback loop.
I cant see why that shouldn’t work well.

Jan E Veiset

[Shoog]

Good to hear a thumbs up.

I was looking at the 6AQ5 and the curves start to look very squashy down at those plate and g2 voltages. However the 6AU6 looks very linear and happy at this sort of operating point.

The only real questions I have are;
1: is the gain going to be excessive (the 6AU6 is 36x in triode mode)
2: is its performance going to be superior over a simple triode

The only way to answer the second question would be to build it I suppose.

Shoog

[Shoog]

I just happened across the data for the ECF83. Looks as if it might be an excellent candidate for this application.
The Pentode section has a gain of 11x in triode mode at -4V bias and 5mA current.
The Triode is capable of 10mA at 80V. This means it could be used as the CCS on the bottom of the totem, and it would perform better than the ECC88 because its grid bias would be 4V rather than the 1.5V or so of the ECC88.

The other main advantage is that this then becomes a two valve design with at least one of the valves been cheap.

Shoog

[Brian Beck]

Shoog,

My first post here in a long time... much too busy...

Interesting circuit. As Jan said, this is really a triode circuit. You can think of it as UL with a 97% (or so) tap, and that's really a triode connection after all. However, there is one difference here: the screen current (both DC quiescent and signal) flows through the CF cathode (V2A), and not through the plate load resistor (R3) as you'd usually see with a screen connected directly to plate. This means that the small contribution to overall transconductance from the screen will be lost. So you'd have maybe 80% of the regular triode-connected transconductance available, with a similarly slight increase in rp. I don't think this a problem, and it's hard to say which connection might be better with regard to distortion and sonics. It's easy to switch back and forth. I think it's worth a try.

If you want the DC quiescent plate voltage to be higher than the screen voltage for a particular pentode, you could slip a zener or a stack of LEDs between the CF cathode and the screen, and readjust B+ accordingly. Zener noise could be problem, but probably not too bad at this point. Bypass it heavily.

Finally, if you want the bottom tube (V2B) to act as a good CCS, then perhaps this isn't the best place for LED cathode biasing. LEDs are used in cathodes to give low resistance and low feedback. Here, you should use an unbypassed cathode resistor for high feedback; its value gets multiplied by mu+1 to add the to the plate resistance, making it a better CCS. You could even put a solid-state CCS in the cathode to really pump it up and to allow for precise settability.

Let us know how it plays out.

[Shoog]

Hi Brian,

Good to hear from you. I'm using your idea for Constant Current Sink differential stages extensively now - to very good result. Just built a headphone amp and used it, and also intend to use it in a Tabor clone. Thanks. The only variation I am trying is cross coupled resistive garter Bias which eliminates the CCS.

Finally, if you want the bottom tube (V2B) to act as a good CCS, then perhaps this isn't the best place for LED cathode biasing. LEDs are used in cathodes to give low resistance and low feedback. Here, you should use an unbypassed cathode resistor for high feedback; its value gets multiplied by mu+1 to add the to the plate resistance, making it a better CCS. You could even put a solid-state CCS in the cathode to really pump it up and to allow for precise settability.

I have used a LED in this position in my current preamp, and it works well. However your explanation clarifies whats going on and leads me to think this is not the best choice, so I will go for a resistor.

I have done some research on the ECF83 and its a perfect fit for this application, with one cavite. That been that the pentode is remote cutoff (not so good). However looking at the curves, if I have it set up with the screen and plate at 80V and the bias point at -4V, the input signal will be about +/- 1V and that is just out of the cut off region. Its certainly worth a try as this operating point is tailor made for the Super Linear buffer. Also both the triode and pentode are happy to pass reasonable current (7-10mA) which is unusual for a combination valve. Its other main advantage is the pentode has a gain of only 10x in triode mode, which is again unusual for anything but a power pentode - and perfect for this application.

I'm just working on ordering a pair of ECF83's.

Shoog

[smoking-amp]

I have been using this type of circuit for some time, it works quite well. Put a CCS in the pentode plate circuit for very low distortion. Some voltage dropping devices for the screen grid will keep its current more linear (which is helpful since it subtracts out from the plate current).

If you use a low capacitance MosFet for the follower part (Fairchild FQP1N50), there is no need for the CCS output follower load (or the top plate follower either) due to the Fet's very high gm.

http://www.diyaudio.com/forums/showt...04#post1172104

It can also be used to make an all SS "Early-effect triode" device by using a Fet cascoded bipolar device in place of the pentode. (The FET gate then becomes the "screen grid" for feedback) Higher Beta transistors give a lower Mu triode as a thinner base region has a larger base width modulation Early effect. Quite linear triodes are possible. I use this technique in a no GNFB SS "Early-effect triode" amplifier.

http://www.diyaudio.com/forums/showt...84#post1258784

Don