High NFB OTL headphone amp - 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
Thread Tools Search this Thread
Old 10th June 2015, 02:33 PM   #1
diyAudio Member
Join Date: Jan 2014
Default High NFB OTL headphone amp

Hi all

I've been pondering over a headphone amp project for some time now and, everytime shortly after I thought I found some design or topology that suits me, I noticed it had shortcomings in areas I wasn't aware of before. It's apparently a tough job to create a headphone amp design that works universally well with most kinds of cans.

This is what I want
  • Must work with phones 32 - 600 ohms.
  • Low output impedance over the whole audio range for good damping but esp. for bass.
  • Reach 110db SPL with most headphones without clipping for clean dynamics.
  • Low distortion, obviously.
  • Flat response 20-20k min. with low phase shift / group delay.
  • Good PSRR for pitch black backgrounds.
  • Must not consume more power or be more expensive - component wise - than a small power amp (for speakers).

Please take a look at the attached schematic I came up with so far, and comment/confirm (or not) I'm not going over the top here

I chose a OTL topology to keep down cost and help with frequency response. Output stage is a WCF using the 6N6P, which provides decent power for most cans (and hits in the other departments as well). Which leaves us with low Zout and low distortions for the wildly variable load range of 32-600R, which will be tackled by throwing lots of NFB at it (very contrary to my speaker amp designs).

The input stage I came up with is a cascode mu stage with a pentode as active load, which provides ~58dB of gain and good PSRR, while maintaining very good frequency response and minimal phase shifts, which are mandatory for a high NFB ratio. The input stage is dc coupled to the WCF to preserve the same, and for the same reasons.

The feedback loop is stabilized at the low end by cutting of sub bass frequencies at the input before they enter the nfb loop, and by means of C1 and C9, which shunt and dampen out the HF inside the nfb loop. The two caps are tuned to each other to quickly dampen out any overshoots (which do happen in the input stage when fed by rect signals) while maintaining overall frequency response well into the supersonic region (>100kHz).

Critical or noteworthy components are the output coupling cap C8 (for dc protection, so, safety critical) and feedback resistor R19, for which 16kOhm is right for 32 Ohm load and 4 dB overall gain, but should be increased to about ~120kOhm for 600 Ohm cans.

Even though the remaining parts count is quite high (3 tubes, 9 caps and 19 resistors in total per channel), they are all cheapies (except for a ECC88 or 6N23P, look at alternatives) and the circuit is very resilient against parts tolerances.

Note that heaters must be elevated to remain within Ukh limits, and two heater circuits may be required thanks the the E180F (russian 6J9P is better), but the good thing is that the 300V B+ can easily be provided with a cheapie 230V/230V isolation transformer.

Some other figures:
  • Power consumption ~9W per channel (w/o heaters)
  • Output power is mostly limited by a fairly constant 42mA rms (120mA p-p) output current, except for top end loads, where it drops to about 29mA rms, which is 450mW rms into 600 Ohms.
  • THD <0.03% at full power for all loads 32-600 Ohms. (Using ECC88/6N23P cascode, other tubes give slightly higher THD)
  • Zout is ~0.2 Ohms thruout the audio range and above (but not below).

Any striking errors or safety issues in the circuit I missed due to limited experience with such designs?
Attached Images
File Type: gif 10.06.gif (29.2 KB, 208 views)

Last edited by goldenbeer; 10th June 2015 at 02:39 PM. Reason: added schematic :)
  Reply With Quote
Old 12th June 2015, 01:53 PM   #2
diyAudio Member
Join Date: Jan 2014
Added a small Schottky for symmetrical clipping. Easier on the circuit and especially the 'phones.

2nd Schematic is a suitable PSU for use with a 230-230 isolation transformer. R4 should be 2 Watts, the BD136 doesn't need a heatsink.
Ripple < 1mVpp and Zout ~0.13R, 20Hz-100kHz.
Attached Images
File Type: gif 12.06_amp.gif (30.4 KB, 118 views)
File Type: gif 12.06_psu.gif (20.8 KB, 116 views)
  Reply With Quote
Old 8th July 2015, 02:03 PM   #3
diyAudio Member
Join Date: Sep 2010
You may have more success and a bit less "waste" using a series-pass power MOSFET than a BJT. I built myself a very simple series-pass filter similar to yours using that and achieved very good ripple rejection as well.

Keep in mind that you're after both common and differential mode noise suppression on your DC supply, and if you're after more suppression it may make life easier to use DC on your heaters, which will allow you to keep all the nasty AC at one end of the chassis away from all the signal lines.

Got to love those old Soviet tubes! I have a box of them, dual triodes and small power pentodes, shiny NOS from the 60's. Can't wait to put them to use.

  Reply With Quote


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
6N1P and 6N6P OTL headphone amp orson1956 Tubes / Valves 40 8th June 2015 10:23 AM
High Pass Filter for OTL headphone Amp supernut Tubes / Valves 5 1st May 2015 03:44 PM
Adding NFB makes headphone amp sound AWFUL!!! dludingt Tubes / Valves 32 21st April 2014 11:49 PM
Balanced OTL headphone amp joefish Headphone Systems 0 23rd May 2012 09:26 AM
No NFB Headphone/Pre Amp VladimirK Solid State 97 21st October 2011 05:08 AM

New To Site? Need Help?

All times are GMT. The time now is 06:22 PM.

vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2015 DragonByte Technologies Ltd.
Copyright 1999-2015 diyAudio

Content Relevant URLs by vBSEO 3.3.2