3-Transistor Headphone Amplifier - diyAudio
Go Back   Home > Forums > Amplifiers > Headphone Systems
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Headphone Systems Everything to do with Headphones

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 26th April 2016, 01:20 AM   #1
jstott is offline jstott  United States
diyAudio Member
 
Join Date: Feb 2016
Location: Connecticut
Default 3-Transistor Headphone Amplifier

So, I was playing around trying to make the simplest amplifier I could. I ended up aiming for a low-power headphone amplifier, just to keep it simple, designed to drive ordinary cheap low-impedance headphones (iPod ear buds and the like, 20-30 ohm load at DC). The design works far better than it has any right to, so I though I would share it.

The amplifier design is fairly conventional (schematic attached). It's a single-ended class-A amplifier with no feedback, all solid-state [no opamps]. A small wall-wart and a voltage regulator are all it needs for power. TO-220 output transistors were happy without heat sinks, but I didn't want to push it into full clipping so I think I'll add some small ones when I do a permanent build. Because of its low gain and output design, stability and short-circuit protection are excellent.

Frequency response is flat to out beyond 1 MHz. Spice says distortion is is in the 0.5% range and, as expected, increases as the volume increases. I'm no golden ear, but I couldn't hear any audible distortion. Using the headphone jack on my laptop as a source, the tops and bottoms were definitely stronger with the amplifier than driving the headphones directly (but that could just mean the computer is happier driving the higher input impedance of the amplifier).

Nothing in the part selection is critical - I just used what I had on hand. If you want to substitute transistors, feel free. The output transistors, Q2 and Q3, need to dissipate as much as 0.5W [worst-case] and quiescent current is 70-100 mA, so I don't recommend using small-signal TO-92 transistors there, but any tabbed power transistors should be fine. I've also tried using a MOSFET instead of a BJT for Q2 and, other than needing a higher input biasing, it worked just fine. Q1 can be any small-signal transistor but the linearity is better with higher-current models like 2N2222/2N4401 rather than the smaller 2N3904's. R2 and R3 bias the input stage, R7 and R8 bias the output current source - if you swap transistors you may have to tweak these biasings a little. R10 is there to decrease the output power and swamp any possible parasitic cable reactance (hence no output Zobels). If you want to leave it out, the amplifier should still work just fine.

My only complaint is that there's more hiss than I expected. It's there even with grounded inputs, so it's definitely coming from the amplifier. Spice says this should be a nice low-noise design (sub-uV RMS noise), so that one's got me scratching my head. For trials I was using a bench power supply and some 10u decoupling caps, so it's not on-board regulator noise.

-Jonathan
Attached Images
File Type: png trioHA.png (331.0 KB, 994 views)
  Reply With Quote
Old 26th April 2016, 01:26 AM   #2
kevinkr is offline kevinkr  United States
diyAudio Moderator
 
kevinkr's Avatar
 
Join Date: Sep 2004
Location: Boston, Massachusetts
Blog Entries: 6
Interesting and elegant design, have you considered paralleling R8 with a large electrolytic or replacing it with a red led or green led - match Vf to what you need to bias Q3 to your desired Q point. This should help with the hiss as the johnson and excess noise generated by R7/R8 is currently amplified by Q3.

Potentially C2 is a bit small, have you calculated the -3dB point of the bootstrap?
__________________
"To argue with a person who has renounced the use of reason is like administering medicine to the dead." - Thomas Paine
  Reply With Quote
Old 26th April 2016, 02:57 AM   #3
jstott is offline jstott  United States
diyAudio Member
 
Join Date: Feb 2016
Location: Connecticut
Quote:
Originally Posted by kevinkr View Post
Interesting and elegant design, have you considered paralleling R8 with a large electrolytic or replacing it with a red led or green led - match Vf to what you need to bias Q3 to your desired Q point.
I tried the LED (in simulation). Diode biasing works with the BJT, but with a MOSFET I just didn't have enough control of the bias level (the extra 3.5V drop doesn't leave much headroom on the output).

Quote:
This should help with the hiss as the johnson and excess noise generated by R7/R8 is currently amplified by Q3.
Good point. Simulation said an extra cap wouldn't help so I left it off, but of course simulation also said there wouldn't be any hiss. I'll try it for real this weekend when I have some more time.

Quote:
Potentially C2 is a bit small, have you calculated the -3dB point of the bootstrap?
The -3dB point I get is 12 Hz. I'm content with that, but it's easy enough to swap it out with a 22uF capacitor.

Thanks for the feedback,
Jonathan
  Reply With Quote
Old 26th April 2016, 10:21 AM   #4
diyAudio Member
 
Join Date: Nov 2004
Location: Germany
I can see some potential for improvement there.

1. The output is referred almost directly to the +9V rail. PSRR is near zero. Might explain why the circuit is noisy. Try using +9V as output ground (while reversing C3).

2. The current through R7/R8 would seem to be too small for Q3 to serve as a good current source - try reducing them by about a factor of 10. R9 also strikes me as a bit small. Also consider adding a diode in series with R8 (1N4001-ish?), thermally coupled to Q3 (hint: transistor legs have pretty good thermal conductivity).

3. Consider splitting and bootstrapping R5 from the output and adding current feedback from the output to R6. Reverse (1.) then.

4. Also consider going for a PNP Q1 (turning associated circuitry upside down) and -9V supply, referencing everything to +Vunreg (which becomes your new ground). That would give you a clean ground for both stages.

5. TIP41C (6 A) are rather big for an application like this, try TIP31C (3 A) or BD139 (1-1.5 A) instead.
  Reply With Quote
Old 30th April 2016, 07:52 PM   #5
jstott is offline jstott  United States
diyAudio Member
 
Join Date: Feb 2016
Location: Connecticut
Default Schematic version 2.0

The core design is the same, but I've updated the schematic to include some of the suggestions by kevinkr and sgrossklass.
  1. The bias currents are now higher, which fixed the background hiss. Replacing R8 with an LED is another variation on the same theme.
  2. The input stage is limited in how much current it can source without distorting, so I increased the output emitter resistor (R9) to keep the quiescent current at about the same level as before.
  3. Smaller resistors also require bigger capacitors to preserve the bass.
  4. The output capacitor may be a bit marginal at 470 uF, depending on your headphone impedance. 1000 uF should be happy even with 20-ohm headphones.
  5. I decide the volume-limiting resistor, R10, wasn't worth the bother so I left it out.
  6. Finally, I noticed that if you over-drive the input, the bootstrap causes the input transistor to back-bias. If taken to extremes (ca 5-10V input), it could even destroy Q1 by exceeding its maximum base-emitter reverse voltage. To protect the transistor, I've added D1 which limits the worst-case base-emitter voltages to about -0.8V. During normal operation, the diode is reverse biased and it doesn't pass current.

As I said in the original post, component selection is based on what I had in the drawer; the output transistors are not optimal so feel free to substitute. Also, as sgrossklass pointed out, the amplifier has basically no PS rejection. A voltage regulator (LM317, LM7809, or similar) on the supply rails is pretty much mandatory to hold down the supply hum. Personally, I plan on using a 12V wall-wart and regulating down to 9V using a separate regulator for each channel when I get around to boxing the whole thing up.

-Jonathan
Attached Images
File Type: png trioHA_v2.png (361.9 KB, 855 views)
  Reply With Quote
Old 30th April 2016, 08:17 PM   #6
diyAudio Member
 
Join Date: Jan 2002
Location: Tyrone Ga. U.S.A.
Isn't C2 reverse biased ?
  Reply With Quote
Old 1st May 2016, 10:17 PM   #7
jstott is offline jstott  United States
diyAudio Member
 
Join Date: Feb 2016
Location: Connecticut
Quote:
Originally Posted by woody View Post
Isn't C2 reverse biased ?
The bias voltage (approximately equal to base voltage) is more positive than the emitter, so I think it's drawn right. honestly, I just used a voltmeter when I built it.
  Reply With Quote
Old 2nd May 2016, 02:46 AM   #8
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Maybe consider back to back 10uF electros there to form a non-polar. Then you are covered no matter how large the input swing.
Attached Images
File Type: jpg DIY nonpolar.jpg (41.0 KB, 741 views)
  Reply With Quote
Old 2nd May 2016, 03:40 AM   #9
kevinkr is offline kevinkr  United States
diyAudio Moderator
 
kevinkr's Avatar
 
Join Date: Sep 2004
Location: Boston, Massachusetts
Blog Entries: 6
The AC voltage across the cap will be extremely small in normal operation, and the cap will always be properly polarized. Once the driver stage clips behavior will get nasty, but since this is a headphone amplifier it is unlikely to be tolerated for any length of time.

The connection shown is not correct for a bipolar capacitor in my experience - if any significant level of DC is present one of the capacitors will become reverse biased which may lead at best to significant leakage and at worse a shorted cap. (Steering diodes would normally be used with the shown connection and this is not suitable for use in audio applications, but is not uncommon in DC blocking applications on AC mains)
__________________
"To argue with a person who has renounced the use of reason is like administering medicine to the dead." - Thomas Paine
  Reply With Quote
Old 2nd May 2016, 06:42 AM   #10
wwenze is offline wwenze  Singapore
diyAudio Member
 
Join Date: Mar 2008
Quote:
Originally Posted by agdr View Post
Maybe consider back to back 10uF electros there to form a non-polar. Then you are covered no matter how large the input swing.
Pretty sure that's not how to form a non-polar... and the picture is showing parallel anyway

Back to back is correct tho.
  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
Two Transistor Amplifier Stinger Pass Labs 15 30th January 2016 05:07 PM
Headphone Impedance and headphone amplifier TheGimp Tubes / Valves 89 3rd May 2014 02:46 PM
THAT transistor headphone amp (250ma discrete opamp) design sanity check. Russ White Headphone Systems 19 13th December 2007 01:52 PM
OTH -> One Transistor HeadPhone mod_evil Headphone Systems 7 13th September 2006 06:00 PM
Wanted 2paar beta matched transistor for my headphone amp! cartman Swap Meet 0 17th July 2005 09:51 AM


New To Site? Need Help?

All times are GMT. The time now is 09:22 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Copyright 1999-2016 diyAudio

Content Relevant URLs by vBSEO 3.3.2
Wiki