Headphone amp

jaycee · 21st April 2007, 12:43 PM

Hi,

I got bored and started noodling around in LTSpice. At first I was simulating the usual buffered opamp, using 2 diodes for bias, something like ESP's Project 113. Then I decided to try out diamond buffering. I came up with this:

Pretty basic I think and seems to sim well, but I'm not a great expert on designing amplifiers. Any experts out there care to tell me what they think?

I was even thinking about trying it out using BC850/860 instead of BC546/556, but i'd have to get some of those...

Juergen Knoop · 21st April 2007, 02:35 PM

I would reconsider the gain setting.
A gain of 22 seems far too much.

regards

jaycee · 21st April 2007, 02:44 PM

True, ideally I want full swing with a 1V RMS input signal. It's easy enough to change R6 though.

edit: I set R6 to 3K and that does the trick. It still seems stable but I dont think the model of the OPA(2)134 is particularly good in that regard.

peranders · 21st April 2007, 02:57 PM

The output stage will run in almost class B.

Forget about the cascodes in my design. Check which values the emitter resistors have.

jaycee · 21st April 2007, 03:43 PM

Hi Per-Anders... I must say I've seen your boards and they are works of art

Sadly I'm just a home hack, doing my boards in a tray of sodium persulfate being warmed in a bath of heated water.. I haven't tried anything double sided yet as I havent worked out a good (cheap) way of doing vias.

I am guessing the cascode stages are T3/T7, and T14/T4/T8. What purpose do T1 and T2 serve? I am guessing they act as references in the current mirror.

I had thought about emitter degeneration and will probably add it as it makes the PCB easier to construct into the bargain.

The output is class AB but only just, there is about 18mA of bias according to the sim. Class A would be better though, especially at this low power.

Conrad Hoffman · 21st April 2007, 08:00 PM

The design looks nice, and I'd be curious if it performs as well as the sim. No reason it shouldn't. Long ago I did a class A headphone amp I really liked, though the design was a bit unorthodox. It was basically an op-amp driving a power FET, but with a separate op-amp/FET current source underneath. The interesting thing about that is the output of the op-amp runs at the gate voltage of the FET. Thus, until you reach fairly high levels, no zero crossing occurs in the op-amp- it runs pseudo class A as well. The thing sounded great, and I can probably dredge up a schematic if any body's interested, though you'll likely be a bit skeptical at first

peranders · 21st April 2007, 08:21 PM

Quote:

Originally posted by jaycee
The output is class AB but only just, there is about 18mA of bias according to the sim. Class A would be better though, especially at this low power.

Remember that you must consider temperature stability also. The simulations give you nothing about this.

jaycee · 21st April 2007, 09:26 PM

Yeah, I was concerned about thermal stability. Not sure how to go about that one yet. I might try and put it together with some parts I have laying about, on a piece of veroboard or so.

As you said, the sim doesnt tell the whole story.

21st April 2007, 12:43 PM	#1
jaycee diyAudio Member Join Date: Apr 2004 Location: Norwich, UK	Headphone amp Hi, I got bored and started noodling around in LTSpice. At first I was simulating the usual buffered opamp, using 2 diodes for bias, something like ESP's Project 113. Then I decided to try out diamond buffering. I came up with this: Pretty basic I think and seems to sim well, but I'm not a great expert on designing amplifiers. Any experts out there care to tell me what they think? I was even thinking about trying it out using BC850/860 instead of BC546/556, but i'd have to get some of those...

21st April 2007, 02:57 PM	#4
peranders Electrons are yellow and more is better! diyAudio Member Join Date: Apr 2002 Location: Göteborg, Sweden Blog Entries: 4	The output stage will run in almost class B. Forget about the cascodes in my design. Check which values the emitter resistors have. __________________ /Per-Anders (my first name) or P-A as my friends call me Tube Buffered Gainclone in work \|Thread

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Another DAC + Headphone amp	ghetto	Digital Source	14	10th July 2008 04:02 PM
Using a headphone amp as a pre-amp?	Basalt	Solid State	4	19th November 2007 09:55 AM
ESL Headphone X-former	bjackson	Planars & Exotics	5	19th September 2005 03:45 AM
Headphone amp	audioPT	Headphone Systems	3	20th September 2004 10:17 PM
Zen v4 headphone amp	Rob M	Headphone Systems	11	21st May 2003 12:58 AM

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	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

21st April 2007, 02:35 PM	#2
Juergen Knoop diyAudio Member Join Date: Jan 2007 Blog Entries: 3	I would reconsider the gain setting. A gain of 22 seems far too much. regards

21st April 2007, 02:44 PM	#3
jaycee diyAudio Member Join Date: Apr 2004 Location: Norwich, UK	True, ideally I want full swing with a 1V RMS input signal. It's easy enough to change R6 though. edit: I set R6 to 3K and that does the trick. It still seems stable but I dont think the model of the OPA(2)134 is particularly good in that regard.

21st April 2007, 03:43 PM	#5
jaycee diyAudio Member Join Date: Apr 2004 Location: Norwich, UK	Hi Per-Anders... I must say I've seen your boards and they are works of art Sadly I'm just a home hack, doing my boards in a tray of sodium persulfate being warmed in a bath of heated water.. I haven't tried anything double sided yet as I havent worked out a good (cheap) way of doing vias. I am guessing the cascode stages are T3/T7, and T14/T4/T8. What purpose do T1 and T2 serve? I am guessing they act as references in the current mirror. I had thought about emitter degeneration and will probably add it as it makes the PCB easier to construct into the bargain. The output is class AB but only just, there is about 18mA of bias according to the sim. Class A would be better though, especially at this low power.

21st April 2007, 08:00 PM	#6
Conrad Hoffman diyAudio Member Join Date: Mar 2007 Location: Canandaigua, NY USA	The design looks nice, and I'd be curious if it performs as well as the sim. No reason it shouldn't. Long ago I did a class A headphone amp I really liked, though the design was a bit unorthodox. It was basically an op-amp driving a power FET, but with a separate op-amp/FET current source underneath. The interesting thing about that is the output of the op-amp runs at the gate voltage of the FET. Thus, until you reach fairly high levels, no zero crossing occurs in the op-amp- it runs pseudo class A as well. The thing sounded great, and I can probably dredge up a schematic if any body's interested, though you'll likely be a bit skeptical at first

21st April 2007, 09:26 PM	#8
jaycee diyAudio Member Join Date: Apr 2004 Location: Norwich, UK	Yeah, I was concerned about thermal stability. Not sure how to go about that one yet. I might try and put it together with some parts I have laying about, on a piece of veroboard or so. As you said, the sim doesnt tell the whole story.