• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

Aspen Headphone Amp

Founder of XSA-Labs
Joined 2012
Paid Member
Yes, Hugh gave me the BF862/2SA1837 headphone amp design and it remains one of my best. It was in fact, the amp that beat all the other amps in a blind virtual audition.

Blind Virtual Audition of Several Headamps

596524d1486098267-mosfet-source-follower-headamp-bf862-2sa1837-headamp-pcb-closeup.jpg
 
Last edited:
I tried changing some stuff in the schematic posted by Nico Ras. This is what I finally built on a breadboard.

An externally hosted image should be here but it was not working when we last tested it.


When I plugged in my headphones and played music through this amp, my jaw just dropped in amazement. It sounds totally different from my CMOY amp. With hip hop, the bass hits so hard and is tight. On some songs I did not even know there was that much low bass. I just love the slam and impact of the bass with this amp. On some songs I thought that the music was recorded with bass distortion. After using this amp, the bass distortion is gone. There is also a sense of space in some of the music I tested. The treble also feels sweeter and not fatiguing anymore. I am gonna continue listening to more music. Thank you for sharing this amp schematic.
 
Buzzy,

I'm glad you built the CFP HPA and then listened it!
It really is an astonishingly good HPA and one of the simplest I have ever seen.
When you realise how good it is you wonder why so many complex amps are in the market, particularly knowing they are studded with ICs and massive global feedback.

My approach is always to find what works, simplify, simplify and simplify. Less components are cheaper, more reliable, and easier to assemble.

HD
 
Well, I found this thread today.

It was started by Hugh's on 2009, has 1145 replies and 110.042 views. Uau ... that's a lot of interest ! But, correct me if I'm wrong, it stopped on May 2011. The member KnightofAwesome was making a great job designing the pcb, but than ... nothing.

And again, Hugh is around and remembering that it is simple but astonishingly good. Also xrk971 made a prototype and is reporting as one of his best headphone amp.

Somebody has the latest pcb design files and BOM made by KnightofAwesome ? Can that be a re-starting point ?

So, by luck, somebody searched this old thread, and it may came to life again if the interest is around.
 
It is amusing, all this history with a dead end; yet the interest persists. Perhaps this is the reason XRK and I decided to collaborate some months back to build a very high quality headphone amp which could drive a rusty nail from DC to light - already we have a schematic, DAC, balanced and unbalanced input, high quality power supplies.

But this will be a commercial product, though it will be arguably the best headphone amp most people will ever hear........

Cheers,

Hugh
 
Here is another of Hugh's concoctions at about the same time 2007/8 that sounds superb. I have been using it for about 10 years now since Hugh and I searched for the most simplistic class A headphone amps.

My philosophy is the simpler a class A amp the better it sounds, my best take was three transistors, Hugh used only two. I don't even think he recalls this.

Hello Nico,
I have a few following queries, would be glad if you could answer them
1. What is the significance of R9 (32ohm to ground)?
2. Your build seems to use a few more caps than required for a stereo (6 as per schematic )
3. supply appears to be a 7812 regulator?
4. how to adjust gain? by reducing R8?

I am thinking to build one for me.

regards
Prasi
 
Last edited:
Hi Prasi,

I have reworked it for best performance, but there are limitations with only 12V and a stiff 32R load. Both limitations increase the THD but the harmonic profile is still very good, with typical THD of 0.03% at 1Vpp into 32R load at 1KHz. H2 is dominant.

At 32R load you are limited to an output of about 2Vpp. At higher impedance loads, such as a 280R Beyerdynamic, your output voltage rises to 9.2Vpp, which is very loud indeed.

Your resistor divider for the input voltage reference is a little off, and you should use a bias resistor of 22k, rather than 10k, to minimise changes to the divider reference voltage.

The adjusted schem is attached. Nice circuit, sounds wonderful!

Hugh
 

Attachments

  • AKSAcfpHPA1.gif
    AKSAcfpHPA1.gif
    80.9 KB · Views: 509
Hi Prasi,

I have reworked it for best performance, but there are limitations with only 12V and a stiff 32R load. Both limitations increase the THD but the harmonic profile is still very good, with typical THD of 0.03% at 1Vpp into 32R load at 1KHz. H2 is dominant.

At 32R load you are limited to an output of about 2Vpp. At higher impedance loads, such as a 280R Beyerdynamic, your output voltage rises to 9.2Vpp, which is very loud indeed.

Your resistor divider for the input voltage reference is a little off, and you should use a bias resistor of 22k, rather than 10k, to minimise changes to the divider reference voltage.

The adjusted schem is attached. Nice circuit, sounds wonderful!

Hugh

Hi Hugh,
Thank you so much!
You are a master at this craft!
I will surely build it.
Thank you again.
Regards
Prasi
 
Hi Prasi,

I have reworked it for best performance, but there are limitations with only 12V and a stiff 32R load. Both limitations increase the THD but the harmonic profile is still very good, with typical THD of 0.03% at 1Vpp into 32R load at 1KHz. H2 is dominant.

At 32R load you are limited to an output of about 2Vpp. At higher impedance loads, such as a 280R Beyerdynamic, your output voltage rises to 9.2Vpp, which is very loud indeed.

Your resistor divider for the input voltage reference is a little off, and you should use a bias resistor of 22k, rather than 10k, to minimise changes to the divider reference voltage.

The adjusted schem is attached. Nice circuit, sounds wonderful!

Hugh

Hello Hugh,
thanks a lot for the schematic and your contribution to forum..

i would like to build one for myself but may i ask please,
what is the power ratings for the resistors? is 2w sufficient for R4 and R7?
also what is the potentiemeter value since i prefer to change volume on amp instead of source..

also, i'm curious about the differences from BF862/2SA1837 version.. i'm gonna try to make a layout that accepts both versions..

thanks in advance..
 
Aspen HPA

Hi Endia,

Thank you for your post!
The dissipation in the 56R resistor is Esquared/56 = (4.8x4.8)/56 = 0.411W.
Generally it's best to use a resistor rated to three times the actual dissipation, so for nearest preferred type I'd select 2W. It should NOT be a wirewound resistor, because they have significant inductance. Best is a high power carbon resistor.

I have noticed that the output cap (1.5uF) is too small and should be around 22uF. This cap, like the input cap, is important to sound quality so it could be a polymer electrolytic, or even a black gate IF you can find on. Failing that, try using a high quality low voltage MKS, polyester filmcap to about 10uF. If you use a 32R headphone, this cap should be large, but if it is a 300R impedance (like a Beyerdynamic) it need only be 4.7uF.

Hugh
 
Last edited:
Hi Endia,

Thank you for your post!
The dissipation in the 56R resistor is Esquared/56 = (4.8x4.8)/56 = 0.411W.
Generally it's best to use a resistor rated to three times the actual dissipation, so for nearest preferred type I'd select 2W. It should NOT be a wirewound resistor, because they have significant inductance. Best is a high power carbon resistor.

I have noticed that the output cap (1.5uF) is too small and should be around 22uF. This cap, like the input cap, is important to sound quality so it could be a polymer electrolytic, or even a black gate IF you can find on. Failing that, try using a high quality low voltage MKS, polyester filmcap to about 10uF. If you use a 32R headphone, this cap should be large, but if it is a 300R impedance (like a Beyerdynamic) it need only be 4.7uF.

Hugh

Hi Hugh,
thank you so much,

regarding to output caps, i'm not sure if i can find polymer caps locally,
instead, is it ok to use 1000-2200 uF aluminum electrolytics?

also i have lots of bc547 but not 547c, may i use them?

made a quick layout on an existing szekeres pcb i designed some time ago;
it has in/out jacks and a 7812 voltage reg. on board (86x86 mm)..

- both 56R and 22R resistors have options for paralleling..
- Q2 has options for TO-126 and TO-220 packages..
- there are space for 5R-68R-100K resistors for compatibility to BF862-2SA1837 version. 5R and 68R should be jumpered for the latest BC547/BD140 version.

thanks again.
regards,
Naim
 

Attachments

  • 9.jpg
    9.jpg
    548.9 KB · Views: 235