Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Headphone Systems Everything to do with Headphones

Balanced Drive 2 Transistor SE Class A Headphone Amp
Balanced Drive 2 Transistor SE Class A Headphone Amp
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 2nd June 2019, 06:57 AM   #1
xrk971 is online now xrk971  United States
Got Foam?
diyAudio Member
 
xrk971's Avatar
 
Join Date: Oct 2012
Location: Metro DC area
Balanced Drive 2 Transistor SE Class A Headphone Amp
Default Balanced Drive 2 Transistor SE Class A Headphone Amp

Hugh Dean (AKSA) showed me a pretty cool circuit a few years ago, but it was only recently made public in this thread. It is beautifully elegant with only 2 actives, but worked like a charm. Very useful for making balanced (two identical outputs with 180deg phase) from a single ended input. Orginally, Hugh designed it as a live driver, but I played with the circuit and added a bigger TO220 output transistor and boosted the bias current. What I ended up with was a beautifully simple, elegant and great sounding amp for balanced drive headphones as low as 40ohms impedance. What is nice about this balanced drive circuit is that it doesn't cancel out all the nice second harmonic like many balanced drive circuits - it still has that nice sweet SE Class A profile.

Here is the schematic and some LTSpice performance predictions - predicted THD for 2Vpp into 250ohm cans is 0.006% - a very respectable figure:
Making balanced output from single end-cfp-se-bal-hpa-schematic-png

Here is the artwork for a home etched PCB:
Balanced Drive 2 Transistor SE Class A Headphone Amp-copper-etch-png

Here is the resulting hand-made etched PCB:
Balanced Drive 2 Transistor SE Class A Headphone Amp-copper-etch-result-jpg

This was the amp that was built - runs off a DC-DC step up from a 12v wall wart, then a simple cap Mx to get 48v rail:
Making balanced output from single end-cfp-bal-amp-build-01-jpg

Here is a closeup of the amp:
Making balanced output from single end-cfp-bal-amp-build-02-jpg

Here is the detail showing the underhung 2SA1837's:
Making balanced output from single end-cfp-bal-amp-build-03-jpg

And here is the measured FFT for 2.0Vpp into 49ohms:
Making balanced output from single end-se-bal-hpa-83ma-bias-2vpp-49ohms-png

How does it sound? Very quiet, very powerful - the dynamics are earth shattering on my OB-1's (rewired for balanced drive). Very nice sounding amp and super easy to make.
Attached Images
File Type: png copper-etch.png (49.5 KB, 470 views)
File Type: jpg copper-etch-result.jpg (564.0 KB, 448 views)
__________________
XRKaudio https://www.etsy.com/shop/XRKAudio

Last edited by xrk971; 2nd June 2019 at 07:06 AM.
  Reply With Quote
Old 6th June 2019, 09:26 AM   #2
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Nice work, X. This was quite amazing, the sound quality was remarkable as I recall. Ideal for driving very long leads, particularly in an active system with a large room and heavy lighting. And the distortion profile was terrific, very natural.

Thanks for putting up the circuit. Nico is building one right now to move audio around his home through wires in his roof, which is pretty noisy he tells me.

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 6th June 2019, 01:00 PM   #3
douggoldberg is offline douggoldberg  United States
diyAudio Member
 
Join Date: Sep 2005
What Vcc are you recommending for this use?
  Reply With Quote
Old 6th June 2019, 02:09 PM   #4
sSound is offline sSound
diyAudio Member
 
Join Date: Jan 2018
0.03% THD is not bad for 2 transistors
  Reply With Quote
Old 6th June 2019, 04:34 PM   #5
xrk971 is online now xrk971  United States
Got Foam?
diyAudio Member
 
xrk971's Avatar
 
Join Date: Oct 2012
Location: Metro DC area
Balanced Drive 2 Transistor SE Class A Headphone Amp
Per the schematic, 48v. This is achievable using a step up DC-DC converter to 52v then a CLC followed by cap multiplier. This can easily be done using one of these:

GB for Simple Cap-Mx Regulated Low-Noise PSU

Just skip the last voltage regulator with a jumper from pin 1 to 3.

Click the image to open in full size.
__________________
XRKaudio https://www.etsy.com/shop/XRKAudio
  Reply With Quote
Old 6th June 2019, 04:37 PM   #6
xrk971 is online now xrk971  United States
Got Foam?
diyAudio Member
 
xrk971's Avatar
 
Join Date: Oct 2012
Location: Metro DC area
Balanced Drive 2 Transistor SE Class A Headphone Amp
Quote:
Originally Posted by sSound View Post
0.03% THD is not bad for 2 transistors
Correct. And itís mostly 2nd harmonic and a little 3rd. Beautiful sounding. Very natural and yet very dynamic.
__________________
XRKaudio https://www.etsy.com/shop/XRKAudio
  Reply With Quote
Old 15th June 2019, 01:41 PM   #7
Mark Johnson is offline Mark Johnson  United States
diyAudio Member
 
Mark Johnson's Avatar
 
Join Date: May 2011
Location: Silicon Valley
Balanced Drive 2 Transistor SE Class A Headphone Amp
Hugh Dean's 3-Transistor circuit used an emitter follower to drive the base of the output PNP ("Q1" in the schematic of post#1 above). So, in his circuit, the impedance at Q1 base is low, and the impedance at Q1 emitter (=(Zbase/BetaQ1)) is very low. That's the output impedance which drives the top end of the load, node "Vout_p" in the post#1 schematic.

However the 2-Transistor variant in this thread eliminates that emitter follower, and now the base of the output PNP "Q1" is driven by a high impedance: the bootstrapped collector load of NPN transistor "Q3". Bootstrapping the collector load emulates a constant current source, resulting in an extremely high impedance. However, the output impedance which drives the top end of the load, "Vout_p", (Zbase/Beta), becomes a very much larger number. I will hazard a rough guess that the output impedance at "Vout_p" is about (R4 parallel (30*R5/BetaQ1)) = 87 ohms.

When discussing the 3-Transistor circuit, Hugh said

Quote:
Originally Posted by AKSA View Post
Zout of each balanced output is around 10R, so this can be used for a 600R balanced line.
But in the 2-Transistor adaptation of this thread, I am concerned that Zout of the upper output is much greater than 10R. I.e. the balanced output drive now has drastically unbalanced output impedances.

Do you happen to know the Zout of each balanced output in the 2-Transistor circuit?
  Reply With Quote
Old 16th June 2019, 12:47 AM   #8
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Mark,

I have not tested it with instruments but simulation gives good answers.
The upper output is indeed high impedance but sink and source issues make it unattractive for heavy loads. At 10k it seems just fine.
I believe Zout source is the defined by the collector resistor - 1k on my original design - but the impedance is not so high; it is 1K, the value of the upper resistor on the collector, as expected. Compared to the 26/10=2.6R Zout of the lower output, this is poor, and this is the reason I added an EF to equalise.

The ultimate of this circuit would be an EF on each balanced output, direct connected and cross linked to 10mA CCS on the opposite rail. This then offers identical Zout - 2.6R as suggested - for both outputs. But then we have six transistors, and it's grown up to be come a Cadillac...... with three transistors almost the same performance is available, and indeed Nico Ras tells me has built one with a ceiling twisted pair cable of 30m between two distant rooms of his home and tells me it is outstanding.

I had hoped that by using a cfp I could extract low Zout from both sides, but in truth, it seems only from the lower, with the master emitter end. The upper does have an emitter on the slave, leading me to believe it could also be low, but in truth it is dominated by the master, not the slave, something I should have realised from the operation. But, it might be possible to drop the cfp phase splitter, use only a single device, and place the removed slave into EF duties on both sides.

I doodled a second version with a single transistor in place of the cfp. Both phases are passed directly to BD139/140 EFs using cross linked 2k2 emitter loads from the 15V power supplies (bipolar). With 2k loads for output and 1Vpk input, I saw -113dB H2 and -121dB H3 with 0.00048% at 1KHz. It seems that because the phase splitter, with a high beta transistor (KSC1845), the single current through the stage ensures very good phase splitting, but the Zout imbalance must be redressed with two EFs. This is still a three transistor configuration, and extremely high performing for low parts count. Of course THAT has terrific chips, but it's a nice exercise to do it from first principles.

Thank you for your post.

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com

Last edited by AKSA; 16th June 2019 at 01:11 AM.
  Reply With Quote
Old 16th June 2019, 01:20 AM   #9
Mark Johnson is offline Mark Johnson  United States
diyAudio Member
 
Mark Johnson's Avatar
 
Join Date: May 2011
Location: Silicon Valley
Balanced Drive 2 Transistor SE Class A Headphone Amp
Hugh, thanks for your message. I'm rather pleased that circuit reasoning plus ordinary algebra still gives useful insight, even in an era where "SPICE seems to be all I need!" is the default assumption.

It sounds like you now believe this circuit (the 2T version of this specific discussion thread) ought not be used to drive low impedance loads, say, lower than 1Kohms. Certainly not 600 ohms and absolutely & certainly not "... great sounding amp for balanced drive headphones as low as 40ohms impedance" as mentioned in post #1 above.

It's promising that preliminary investigations show alternative approaches with better performance into low impedance loads. I wonder if the phase splitter were a common-base amplifier (driven by a "Rush cascode"), maybe you could get Really Damn Good performance out of 4 transistors. Two for the phase splitter / Rush cascode, and two more for emitter follower output drivers. Using emitter resistors instead of emitter CCS loads.

MJ
  Reply With Quote
Old 16th June 2019, 02:25 AM   #10
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Hi Mark,

Well, I must be a little diffident here. I do not believe that low distortion means good sound, and I don't believe that high distortion means bad sound. Let me explain.

The issue is the nature of the distortion and the coarse nature of the THD figure. My rudimentary topology ensures that if there are imbalances between the two antiphase outputs they will show up as mostly H2, a little H3 and maybe a little H4 too. But we should then examine the human hearing issues and realise that most - not all - but most, like H2/H4 distortion.

So if we produce an amp for headphones which creates musical distortions many people generally like it. This is anaethema to engineers who strive for zero distortion, but this is a belief; the reality is not quite this way, the same way as people like to buy cars entirely on their paint color. XRK and I have done a lot of work in this area of psychoacoustics, and I have done a lot of music in my life (keyboard) that gave me an appreciation of harmonics in the musical scale. We have found that HPA with high levels of H2 and all subsequent harmonics dropping linearly (monotonic decrease a la Hiraga) the sound seems 'natural'. Absurdly subjective, and therefore easy to ignore, and lots of good engineers do just that. But if you create H2 at -75dB, H3 -90dB H4 -105dB and H5 -120dB the sound quality seems to be very natural, slightly sweet, and engaging. These issues are not easily measurable, but recent digital developments (focusRite) gives us the way to do it cheaply now. Tube aficionados know this effect well. I built an amp with tubes 25 years ago with H2 at -45dB. It did not sound wooly at all; it was entrancing, and people really liked it, even with orchestras. That was a 6SL7 operating at B+ 340V and 1.6mA and 100k plate load. The remainder of the circuit was zero distortion, all SS with feedback. The amp had high distortion, yet it sounded just wonderful.

This set me on the path of psychoacoustics. I take note of THD, but I try to examine the individual harmonics and their relationships. XRK has helped me hugely to measure and produce better amplifiers which people like to listen to for hours on end.

Nothing of this is new. It's been there for sixty years. Linsley Hood and Jean Hiraga in the sixties knew about it and Nelson Pass is well acquainted with it over the last forty years. But opposing ideologies about the engineering and marketing strategies have obfuscated the issues. I might be completely wrong, but I don't think so. But to prove this with academic rigor some careful testing needs to be done, with real money, and then someone has to take on the market machine which dominates the audiophile world. Remember the Japanese wars of THD in the eighties? WOW, some dreadful amps came out of that.......

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote

Reply


Balanced Drive 2 Transistor SE Class A Headphone AmpHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Clone Audio Technica Class A AT-HA5000 FET Transistor Headphone Amplifier NTSBuilding Headphone Systems 19 14th September 2018 04:04 AM
LM4702 or LME49810 work in class A mode to drive the output stages in class A/AB rhythmsandy Chip Amps 29 29th December 2012 11:54 AM
"Balanced" vs single ended headphone drive chris719 Headphone Systems 11 14th September 2012 06:45 PM
Balanced Drive sounds better? How?! b1o2r3i4s5 Headphone Systems 11 3rd March 2012 02:10 PM
How to drive balanced? vossie Solid State 2 20th January 2004 12:08 PM


New To Site? Need Help?

All times are GMT. The time now is 09:15 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki