Tubes Asylum Questions about tubes and gear that glows. FAQ

---

[ Follow Ups ] [ Post Followup ] Thread: [ Display  All  Email ] [ Tubes Asylum ]

---

[ Alert Moderator ]

---

RE: Bogen DB110 etc AGAIN

Posted by ivan_terrible (P) on February 13, 2016 at 11:14:29

In Reply to: Bogen DB110 etc AGAIN posted by ivan_terrible on January 27, 2016 at 23:38:25:

I thought it was a good idea to post the basic redesign idea I had been working on.

The whole plan was to keep the above chassis parts count for a stereo amplifier to the 3+4 format.-
ie. 4 output valves for the 2 stereo channels + 1 each for each stereo channel and 1 split between the 2.

Of course the other idea was keep the cost down and the quality up.

I had intended to move to an input pentode stage.
This was the perfect moment to use an almost unknown "ultralinear" stage which Mr Broskie illustrated so well, combined with some other astute tricks.
The EF86 stage illustrated by Svetlana has astonishingly low distortion and high gain, but for me it's physically WAY too big, so that will have to "GO UNDERGROUND", ie. sub chassis on a little board holding lots of sub-miniature stuff.

A clever guy in Russia decided to measure quite a bit.
He used the full size version triode connected and found the anode impedance dropped to 27k.
(http://klausmobile.narod.ru/testerfiles/6j32p.htm)

In effect introducing stage by stage negative feedback, linearising each stage, making global feedback unnecessary.

The next obvious way to proceed would be to be able to adjust the input stage sound signature "on the fly".

I also introduced the clever old "NEVE" and Mullard idea which is to parallelise multiple devices to make a quantum leap (or drop) in noise as well as anode impedance.

This means we're going to put some very serious current through the entire input stage while at the same time making each valve the size of the cap of a biro pen...

The anode impedance of an EF86 according to Mullard is roughly 2.5M as a pentode, and quote about 16.5K as a triode.

In this case we can say it's somewhere say 100x impedance lower as a triode.
I don't want to do that, because it defeats the whole point, but the pentode circuit would need to have a following triode circuit which loads it very lightly.
That would defeat the whole point too, noise, hum, miller effect, and other nasties being the danger.

The normal current for an EF86 is 1-2.5m/a.
It's quite noisy & prone to being microphonic as a pentode.

The good news,- there's as a tiny wired ended Russian version of it which operates much better in the range of 4-6m/a.

Here it is,-

At 2V bias it runs 2-3x more current than an EF86 with a fraction of the microphonic noise, so I decided to run 4 in parallel to drop the noise by another 6dB.

This effectively creates an input stage with incredibly low distortion, huge gain and very high bandwidth,while using rather a lot of current.

The current requirements go from a few 100s of microamps of a 12AX7 stage to a range more commonly seen in a driver valve (12BH7?).

With such low impedance, there's never going to be a problem driving a few grid circuits down the line, so I connect the phase splitter directly coupled to it as well as the Ultralinear feedback circuit to drive the grid via a seperate triode.

Broskie's clever piece of design showed how easy it is to make a potentiometer to change the ultralinear ratio on grid 2 of the 6Ж32Б to make whatever you fancy sounds best.

The idea is therefore to put a pot with a switch on it.
Snap it on, and the potentiometer,allows you to go from pure PENTODE, to PURE TRIODE in the input stage.

The stages that follow only serve as phase splitter, and driver using valves that are being used with no gain, and low impedance (cathode followers).

This means they are far less critical in the signal path,and because of degenerative feedback are also very low distortion.

The only smart idea here is to lock everything to a rock solid power supply via a regulator to remove all hum then allow only mV of ripple through to both the AF amp and the output valve Grid 2 circuit.

As we know, the anode supply is far more immune to supply voltage change than the G2 circuit, whereas only a few volts of instability on the G2 is enough to introduce significant distortion, especially if we start to drive G1 into grid current.....

(C4 in fact is unneccessary, because the next stage being a "concertina" type phase splitter is fed directly coupled to the grid.)

In effect this bit of the circuit is the only vestige of the old design remaining, but another nice idea is waiting.

Lots of Bogens used the 12DW7/7247.
Inside a 6U8 compactron is EXACTLY the same valve combo but with an extra 12AU7 triode.

Here's what it looks like,-

Not very big is it?!

They seem to look pretty nice lit up too!

All we need to make proper cathode follower drivers for the 2 output valves, is there, 1 X 12AX7 and 2 X 12AU7.

Let's just remember now, the plan was to move from the 6V6 simply to suit a rather better HT supply.
The 6V6 struggles at anything over 65-70m/a with the Anode maxed out at 12W.
A hop to the hybrid 6L6-7591 type Russian design is nothing at all.
This valve easily manages 90m/a with Anode diss of 21W, (nearly double).

Here's the rough plan which moves from some sort of maxed out 12AX7 phase splitter to a cathode follower arrangement to drive the output valves direct, into grid current if you like (!)

The way to do this is to run a -ve bias rail of about 25V, running a partial cathode bias, partial fixed bias arrangement.

SO far out parts count is putting 5 tiny valves under the chassis and so far 2 compactrons above. (that's 6 triodes instead of 4 in a stereo amp).

That leaves ONE more space to use up.
This one is the space for our tube regulator (A2293), which can be conveniently be fed from the shunt voltage regulators we're already using for the AF amp

The nice thing about tube regulators is they take all the stick for the warm up, so there's no HT on the G2/AF amp circuit even with a solid state rectifier until the regulator warms up.

So what's the final line up & price for it all?

2 GE 1973 year 6U10 compactron, (about 20USD)

1 MO-V GEC 1957 year regulator, UK MADE (about £5)

8 Russian Siberian sub min valves 1970-80. (about £5)

4 Output valves from Uzbekistan 1960-1970 (about £20 max)

3 sub min regulators (150V) from Siberia 1970(about £5)

Transformer base model about £70, or a really nice Toroid for about £80.
Transformer original model (ZERO)
Driver chokes (about £30), but not really essential.

Original power output approx 10W at 4% distortion
Target power output approx 20W at 0.1% distortion.

I hope you enjoy this interesting project.
I'm waiting to try the stereo chassied version out this spring.

This post is made possible by the generous support of people like you and our sponsors:

Parts Connexion

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2024, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---