Shunt Cascode Power Valve Driver - diyAudio
Go Back   Home > Forums > Amplifiers > Tubes / Valves

Tubes / Valves All about our sweet vacuum tubes :) Threads about Musical Instrument Amps of all kinds should be in the Instruments & Amps forum

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 13th October 2009, 06:56 PM   #1
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Default Shunt Cascode Power Valve Driver

Here is a design for driving a Power Valve at the highest level of sound quality!

I have been using a similar circuit since 2005 to drive my 300B-SE. The sound is better all-round than any circuit I could find using triodes or pentodes. Compared to the usual 6SN7 cascade (ac or dc coupled) the sound is much more immediate and articulate (can hear lyrics much easier).

Pentode drivers work well too, but the power supply needs about 10 stages of LC filtering to be quiet enough.

This circuit has been designed to work with the Schade-connected antitriode project that we are working on here:

Spud, Schade, PP, Anti-triode ECL86

but it will work with many others. It will drive loads all the way down to 1K or less, and it takes 1nF to roll the response off at 20kHz with a 2K ohm load.

It will drive our EL84 Schade-connected amp to near full output with only 0.3% THD - almost all is 2nd harmonic.

The circuit is a shunt cascode in architecture.

The TL431 3-pin chip regulates the screen voltage of the output valve, and uses this 210V for a supply voltage. At the same time, the circuit regulates the current source M5, and the cascode voltage via Q1.

The performance of shunt cascode transistor Q1 is important, please try to use the ZTX558 here. MOSFETs have too much capacitance and/or too little gm (remember that hole mobility is lower than electron mobility, so P-channel FETs are much worse the N-channel). The ZTX558 has 400V and 1W handling; flat hfe, but only 5pF Cob. BJTs are the best (upper) cascode device, due to their high gm.

The current source feeds 20mA into the anode circuit, which is pinned to 100V by Q1. the excess current rolls into the output as a class A offset.

The E88CC is designed for cascode duty, and is a perfect fit here. With a fixed 100V anode the valve can operate at its most linear region - hence the very low distortion, and fine sound. A current production valve seems a good idea too.

Other advantages: The input AND the output are referenced to ground, so the loop areas (EM moment) can be reduced greatly. This also means the power supply rejection is hugely better than ordinary cascodes, or pentodes, and is better even than low-Z triodes.

There is no signal current in any electrolytic capacitor (not even the supply cap).
Attached Images
File Type: jpg scfft.jpg (33.9 KB, 938 views)
File Type: jpg schshuntcasc.jpg (85.7 KB, 953 views)
  Reply With Quote
Old 13th October 2009, 07:20 PM   #2
diyAudio Member
 
Join Date: Nov 2007
Location: Dallas
Folded cascode... Neato circuit for sure, but not new.
Putting it together with a shunt regulated servo loop,
that part of it certainly merits attention.

PSRR is indeed excellent, if ref'd only to ground level.
But the circuit you propose to drive is noise ref'd to B+
and floats upon a CCS to GND. How do you propose to
cross the incompatible rail references?

Last edited by kenpeter; 13th October 2009 at 07:26 PM.
  Reply With Quote
Old 13th October 2009, 08:30 PM   #3
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
I tried a few ways of dc-coupling to the EL84 circuit, but they all made a mess of it one way or another.

I imagine in this case that the EL84 anode to grid resistor will be 100K, and have a dc coupled 12~15K resistor from the driver output to ground. cap coupling to the EL84 grid.
The current (signal) source in the power amp circuit is this driver - just remove the 100p/2K from the output of the driver, and connect up.

At least there will then be a low impedance to ground, if not exactly ground reference at the EL84 input side.
  Reply With Quote
Old 13th October 2009, 09:01 PM   #4
diyAudio Member
 
Tubes4e4's Avatar
 
Join Date: Jul 2005
Location: Leverkusen
Hi Rod,

Quote:
Originally Posted by Rod Coleman View Post
Pentode drivers work well too, but the power supply needs about 10 stages of LC filtering to be quiet enough.
10 stages of LC-filtering needed for a pentode driver to meet your demands on what is "quiet", eh?

Regards,

Tom Schlangen
__________________
If in doubt, just measure.
  Reply With Quote
Old 13th October 2009, 09:13 PM   #5
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Quote:
Originally Posted by Tubes4e4 View Post
Hi Rod,



10 stages of LC-filtering needed for a pentode driver to meet your demands on what is "quiet", eh?

Regards,

Tom Schlangen

Tom, OK - I'm making an allowance for the G2 supply filter!

but seriously, it is not so much noise as the sound quality overall. Sound improved very much when the pentode driver mains trafo was upgraded from 150VA to 400VA. (even with 4 stages of LC/CMC filtering, and 6X4 rectifier). The pentode supply was not shared with the power valve - only 2 x 10mA pentodes.
  Reply With Quote
Old 13th October 2009, 10:52 PM   #6
R.I.P.
 
Join Date: Dec 2001
Location: Schaffhausen Switzerland
Rod,
Why do you call this circuit topology a "Shunt cascode"?

Heard of regular cascodes, hybrid cascodes, and inverted cascodes (which can be hybrid) but never a shunt cascode.

To my eyes, it's a hybrid inverted cascode.

Regards, Allen (Vacuum State)
  Reply With Quote
Old 14th October 2009, 02:49 AM   #7
работник
diyAudio Member
 
Rod Coleman's Avatar
 
Join Date: Jul 2004
Location: Warwickshire UK
Allen, well, I haven't heard the name 'inverted cascode' until now!

It's a shunt cascode because the cascoding transistor (Q1) shunts the excess current into the output - as opposed to the regular cascode, where the corresponding device is a series element.

Or, to think of it another way, the output device (Q1) is shunting the input device (the valve) in action. In the normal cascode, the device relationship is series.

The shunt action is what distinguishes it from normal cascodes in terms of performance - especially supply noise immunity (PSRR), and referencing the output to ground (rather than B+).

The name is an attempt to illustrate its action.
  Reply With Quote
Old 14th October 2009, 02:59 AM   #8
diyAudio Member
 
smoking-amp's Avatar
 
Join Date: Dec 2001
Location: Hickory, NC
Re: shunt cascode
I've seen similar type SS cascodes referred to as folded cascodes.

Re: Ken: "PSRR is indeed excellent, if ref'd only to ground level.
But the circuit you propose to drive is noise ref'd to B+
and floats upon a CCS to GND. How do you propose to
cross the incompatible rail references?"

How about using an attenuating current mirror in the Schaded "triode's" plate circuit (at the B+ terminal) instead of the Schade feedback resistor. The grid summing node is summing currents anyway. Then the grids and gates can be ground referenced in the anti-triode stage. I guess this would make the feedback sensitive to load Z, giving a high output Z.

Guess one is back to a hum bucking resistor somewhere and using a normal Schade resistor.

Don
__________________
I want a Huvr-Board!!

Last edited by smoking-amp; 14th October 2009 at 03:09 AM.
  Reply With Quote
Old 14th October 2009, 05:34 AM   #9
diyAudio Member
 
Join Date: May 2007
Rod,
How does it mate to the EL84 with the 15k/100k feedback? It would be interesting to see if it matches the output-stage to give distortion-cancellation. Do you have simmed figures for the complete amp with, lets say, 14V ptp at the output.
How much current through Q1?
  Reply With Quote
Old 14th October 2009, 06:35 AM   #10
diyAudio Member
 
Join Date: May 2007
Did a quick sim of the folded cascode before going to work. Not so encouraging, but my sims involve some guesswork. The folded version has lower sensitivity and a lot higher THD then the CCS´d single tube. And the BJT is random choice.......

So maybe more current throught the BJT? Now is 40mA through the CCS with 15 of them through the E88CC.

Click the image to open in full size.
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
WNA Headphone Amp III and Cascode power supply PCB's for sale. stonedowner Swap Meet 1 18th October 2009 09:55 AM
Smart Folded Cascode power amp, SFCamp roender Solid State 15 20th August 2007 04:50 PM
Driver valve discussion garbage Tubes / Valves 57 7th December 2005 03:03 AM
HV Valve Shunt reg. James D. Tubes / Valves 14 31st January 2003 08:29 PM


New To Site? Need Help?

All times are GMT. The time now is 01:20 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2