Headphone amplifier recommendations

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
And the basic schemo .....  (the pcb has a few additional caps, pot etc)

and basic specs are :

-> psu +15Vdc for 180mA quiescent current per channel, prefereably filtered and/or regulated
-> gain x8  (+18dB), dual 50K log pot at rca inputs, 1% matching Left-Right performance(!)
-> single ended class A, unbalanced rca inputs and stereo 1/4"jack  output
-> Zout 10ohms
-> output power
        80 mWrms into 300 ohms
      200 mWrms into 120 ohms
      650 mWrms into    32 ohms


and for testing, I'm using Audio Technica ATH-M40fs closed headphones with 60ohm impedance
 
It is the jlh 1996 enhancements, plus what I think is parallel'd output devices for moah power - no details of the pcbs avail unfortunately, so it's prolly a pcb traceout jobby  :)  (ebay pcbs are often poorly documented!)

The enhancements relate to more flexible quiesscent current set, as well as usage of bipolar psu, allowing direct coupling of speaker  (eliminating the single-ended psu 1969 circuit's need for an output coupling electro cap).

ANYway, a great little circuit in the compact headphone amp format, AND I hope, a fine 15Wrms class A speakeramp format.

Looking for a groovy psu+protection+mute-relay pcb how  :)

NOW - back to the fun part - measuring with my Motu/REW test setup    :)
 
Great work!

Currently am making four O2 amps. I need only to find new 3.5 jack sockets.
These which i bought are too short...
 
Thanks for posting this. I had been curious about those headphone amps and its nice to see a quality review of the build. I  built a couple of volume controllers that i got off ebay and I have to admit they seem pretty decent but no lcduino which seems top notch.  I used jlm dingos for the in outs but ive been eyeing Some of the other ebay pcbs like the attenuators and balancing amps as they look like they are actually good quality.
 
I've been using the jlh headphone amp for a couple of days now - it runs quiet and cool with +9Vdc supply, masses of power for driving phones. Sound remains impressive and very low distortion.

Noise performance for the class A circuit is reasonable with LM317 regulation, however the ripple could do with some additional knocking down.

So, for my next iteration of Class A  headphone amp, I'm going with this psu unit ...  a Class A psu implementation (no LM317 type regs)    .... good  (supposedly) for up to 250mA at for +/- rails up to 20V

Should knock the ripple right down, I think, and an interesting step away from 3 terminal regulator devices.

http://www.ebay.com/itm/Class-A-Parallel-Regulated-Power-Supply-Regulator-output-Adjustable-5V-20V/171999388137?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D2%26asc%3D38530%26meid%3D6ffc0f8d397841c98ae84783743e417b%26pid%3D100005%26rk%3D2%26rkt%3D6%26sd%3D172272494032

Combined with this 'improved'  JLH implementation - the 'Zen Amp' version. Still super simple and quite cheap. Should drive at least 2 lowZ sets of phones without difficulty. They can go down to 8ohms even !

http://www.ebay.com/itm/ZEN-Class-A-Headphone-Amplifier-Kit-Power-AMP-Kit-DIY-/140801221796?hash=item20c86824a4:m:mbBa8SM7SZrqHFBsS2Eg2gQ

Finally, I'll be adding a couple of other 'utility modules' into the same box and using the same supply

Firstly, this guy (but with the ne5532 substituted)  with the amazing TI LM49860 opamp

http://www.ebay.com/itm/DC-Version-NE5532-OP-AMP-Preamlifier-Pre-amp-with-ALPS-27-Pot-Shield-Cover-/182196144801?hash=item2a6bbc8aa1:g:kjYAAOSweWVXe4yN

and also this guy  (also with LM49860 sub)

http://www.ebay.com/itm/Class-A-NE5532-2-3300UF-Tone-Board-Finished-Board-Volume-Control-Board-/321872846500?hash=item4af11ddea4:g:5iQAAOSwFnFWAPtH

All that should run me around 100usd in a suitable box and give me some more education regarding class A discrete circuits as well as the latest and greatest 2.7nV/sqrtHz  +/-22V supply opamps  :)

 
Well I use the Grado / CMoy Headphone Amplifier and I like very much,
easy and cheap to build:

http://diyaudioprojects.blogspot.pt/2007/08/grado-ra1-headphone-amplifier.html

http://diyaudioprojects.com/Chip/CMoy-Grado-RA1-Headphone-Amp/
 
alexc said:
Sure - I'm just buying another myself  :)

http://www.ebay.com/itm/Newly-Pure-Class-A-Headphone-Amplifier-AMP-1969-Circuit-DIY-Kits-for-HD600-K701-/262199545918?hash=item3d0c4f903e:g:G1IAAOSw3ydV2YxK

or search for 'class a headphone amp' for some alternatives.

I went for this one becuse of the compact pcb 10cmx6cm - if more space is available, I think there are a few similar but better packaged pcb kits out there.

Also, I just went for these for 11usd  :)
Been following this thread( and your a-class madness) with great intrest
And i did actually order this cheap amp kit..
Pcb seems ok quality and ok design but are these included parts decent quality.
(ie. Do you use those parts or replace them with something.)
 
I found the parts to be quite decent quality on the tiny jlh 1969 headphone amp kit. The audio performance is pretty good, and can be very good - I'll be doing some more measurements soon.

......

One problem I had was mounting the pcb - the pot doesn't mount flush with the edge of the pcb at the front - meaning it sits back from a front panel  ..  just enough so that the small amount of pot thread isn't able to secure it strongly.  I use a froont panel thickness of 1.5mm.

I had to grind off some of the pcb front edge to make it work. Also, I had to use a couple of brackets to securely mount the pcb to the fron panel. A couple of nut-heads protruding on the front panel face.

Best to do before soldering up the jack and pot  :)


Apart from that - took me a while to figure the output devices mounting on the underside of the pcb  - no documentation of course :)  Get it the wrong way and you'll know it quick enough!

And of course, those itsy bitsy tiny smt resistors are a royal pia for me - I can barely see them let alone solder them. I used some tweezers and  so on.  What normally takes 15seconds takes a few minutes  :)  Progress, right?

The upside (apart from allowing the small form factor)  is they are supposedly tight tolerance which makes for excellent matching -- I verified very close performance, left and right.

That is pretty important to me, because I find the pcb really excels at driving a load 'balanced mono bridged mode' where symmetrical performance across the left and right is a plus.

ie. feed the pcb module unbalanced inputs which are the left and right phases of a balanced signal,  say from an XLR connector.  Then take the 'left' and 'right' outputs and connect the load across them.

You now have a 'mono' amp with balanced XLR input and balanced XLR output, capable of (ideally) 4x the power to drive a load (somewhat less in practice) and with super low THD, resulting from cancellation effects in the load.

The load can be anything from 8ohms to 10K. THD+N is super super low :)  and the resulting sound is pretty damn fine  ).

Can be used as a powerful line amp/distribution amp, summing amp, high performance headphone amp capable of driving several phones  etc.

At present, I'm doing a couple of 'recycled' builds, in the shells of what once were my first couple of pieces  ..... 

I'm just setting up to really do some detailed measurements etc  comparing with regular opamps and such  :)

Since then I have also bought a few more JLH class a based modules

- a headphone amp but in a larger form factor, onboard heatsinks, bipolar supplies (no output cap) and some of other  '1996 updates'

- low power amp (5W) tiny pcb version of the '1969 version' for use as a line /distro amp

- high power amp version  using T-O3  output devices and the '1996 updates'  (30Wrms of audio power, 100W of heat, per channel!)

- small capacity and large capacity regulated psu modules giving superior ripple performance over standard lm series 3 terminal regulators

.....

So, I'll be measuring up and posting the spectra and findings in my other thread .. soon  :)

I think it's fair to say I've been pleasantly surprised by the audio performance from such a simple circuit. 

But the 'mono balanced bridge mode' performance has been the real eye opener, particularly when driving a good output transformer - where the 'balanced mode' beats down the 'para-feed' operation by a good margin in the low THD stakes.

The net result is that I am now looking to integrate some JLH circuitry with some single-ended tube circuitry  :)

Had to happen :) 

Basically looking at a cheap and simple way of getting the wonderful spectra of say a 6922 tube's plate , to a speaker - with some power,  and without losing the h2 and adding as little h3+ as possible :)

Nothing revolutionary there - as usual plenty, plenty others have trod this path afore me!

Push-pull tube amps have a fair  bit of THD and a good amount of it h3 and h3+.

Class B or AB transistor amps have very low THD but still a fair amount of h3 and h3+

Class A tube, without nfb, has a lot of THD but most all h2.  (da good stuff!)

Preserving the simple, most all h2 spectra all the way to decent power ie. 30Wrms or so, is not so easy.

So far, from my own builds,  only my 'bigSE amp' has gotten me there.

I think I can better it AND in a slightly more convenient package!

My 'bigSE' amp weighs in at 25 kg for mono 22Wrms  :)      and has a 250VA plus psu traffo
 
alexc said:
I found the parts to be quite decent quality on the tiny jlh 1969 headphone amp kit. The audio performance is pretty good, and can be very good - I'll be doing some more measurements soon.

......

One problem I had was mounting the pcb - the pot doesn't mount flush with the edge of the pcb at the front - meaning it sits back from a front panel  ..  just enough so that the small amount of pot thread isn't able to secure it strongly.  I use a froont panel thickness of 1.5mm.

I had to grind off some of the pcb front edge to make it work. Also, I had to use a couple of brackets to securely mount the pcb to the fron panel. A couple of nut-heads protruding on the front panel face.

Best to do before soldering up the jack and pot  :)


Apart from that - took me a while to figure the output devices mounting on the underside of the pcb  - no documentation of course :)  Get it the wrong way and you'll know it quick enough!

And of course, those itsy bitsy tiny smt resistors are a royal pia for me - I can barely see them let alone solder them. I used some tweezers and  so on.  What normally takes 15seconds takes a few minutes  :)  Progress, right?

The upside (apart from allowing the small form factor)  is they are supposedly tight tolerance which makes for excellent matching -- I verified very close performance, left and right.

Just started assembling my kit and noticed same thing;
no documentation and no info on mounting the MJE340
+there is parallel cap in top of 2K2 resistor(?)
Pcb looks little different from your pic(few resistors are different value)

My parts seem to be copies; "benteng" bourns trimmers,
dunno if these panasonic M-caps are real.
But pcb looks really nice and if this works ok,
then for the price it's all good:)

 
I can take a  photo of the underside to show the bjt orientation ... will do tomorrow.

The little documentation, such as it is,  is a little inaccurate - I built to the pcb overlay and for the components which are of different values, I  left those to last then did a 'best guesstimate' approach. There were quite a few differences, but nothing serious.

I'm pretty sure the parts are mostly the cheap and generic end of town, but seem to be fine, excepting the pot mounting arrangement - best to check before final solder of jack and pot.

I wouldn't be basing my neve desk build on them, tho'  :)

Apart from that, still going strong - it's integrated into my latest summing bus box - a Lawo dv975 module  mashup :)

Looking forward to resuming the class a journey - next week I start soldering and measuring the new modules from the ebuy sino sweat shops  :)


 
alexc said:
I can take a  photo of the underside to show the bjt orientation ... will do tomorrow.

The little documentation, such as it is,  is a little inaccurate - I built to the pcb overlay and for the components which are of different values, I  left those to last then did a 'best guesstimate' approach. There were quite a few differences, but nothing serious.

I'm pretty sure the parts are mostly the cheap and generic end of town, but seem to be fine, excepting the pot mounting arrangement - best to check before final solder of jack and pot.

I wouldn't be basing my neve desk build on them, tho'  :)

Apart from that, still going strong - it's integrated into my latest summing bus box - a Lawo dv975 module  mashup :)

Looking forward to resuming the class a journey - next week I start soldering and measuring the new modules from the ebuy sino sweat shops  :)
ok, thanks!
Looking at the datasheets,
looks like markings(on MJE340) should be facing towards back of the board(ie. where rca connectors are)
 
With the output devices mounted on the underside, the device writing faces should face the rear.

ie. looking back from the rca jacks towards the front of pcb, you can read the devices writing.

 
alexc said:
With the output devices mounted on the underside, the device writing faces should face the rear.

ie. looking back from the rca jacks towards the front of pcb, you can read the devices writing.
Yes!

now on quick (listening)test, this does sound really good.
Bias was set somewhere between vcc.
I have to get proper heatsinks as this does really generate heat! :)

just wondering would I hear any differenece,
if I would buy one kit and replace all Chinese parts
with "non fake" counterparts..
 
Exactly as you said :)

I can't really comment on parts selection and it's effect on audio performance, mostly because this is my first of the 'ebuy kits from china'.

I do think there are parts and then there are *parts*.

So - I am using this particular kit as is, learning whats what from it, and appreciating that at least the left-right side matching is pretty dern good. That's a  big plus, to me.

Now all that is interesting enough in a 4 transistor circuit! but sure enough, the pleasing simplicity can be elaborated on for some more performance.

The small form factor makes possible some interesting 'summing amp' implementations and eq amps and such.

Even more so, the small/simple thing makes feasible 'balanced bridge mode' driving loads, which is where it can get truly scary hifi.

Enough so for me to get more research on,  and buy some more modules ...  at different power levels.

As well as reworking my test setup with some serious opamps and such  :) for some proper thd profiling :)

For the big stuff, I am doing my own parts buys for the sino pcbs.

Generally I *do* buy american for key parts ....  that I want some confidence they are genuine  ...  as the good lord intended them to be :)

Certainly for power bjts and mosfets and any kind of decent opamps and so on.

SO, as usual - 'all of the above'. 

 
I'm still testing my amp, but now I have cracking noise that started on other channel..
got to look thru my solders and look for bad components :'(
If I make next one after this "proto" then I will replace all parts and use just the pcb!  :)
 
ln76d said:
Hey,

am looking for headphone amp, can be cheap but decent or at least easy to mod :D
Anyone found something decent direct from China (without european/us sticker on it :D )?
Anything which can work up to 800ohm headphones.
For listening during recording.

Hello there, In76d.

Recently I finished a 2 year recording project, and took the opportunity before beginning the next, to rebuild my studio setup.
This included updating my headphone cue amps.
I required balanced in and through.
So, I looked at the boards mentioned here on this thread, as well as commercially available examples.
I also checked out using a power amp headphone distribution system, but came to the conclusion that each type of headphone would have to be individually impedance matched to the resistor network for correct frequency and phase response.

In the end I chose the Symetrix 304 Headphone amp. This is now out of production, but has the simplicity and quality I was looking for. Four stereo channels with volume control, balanced in and through. This model is the surface mount version of the earlier SX204 through hole,  (which doesn't include a balanced through socket). Both the SX204, and 304 are often available second hand at low prices on eBay. I was lucky enough to find some NOS Symetrix 304, still in the original sealed packaging on offer.




 
I traced this chinese amp pcb;
there might be some errors,
but seems that there is some value changes and few added caps
compared to the online schematics.

Does anyone know what is this based on? (other than basic JLH)
are the mods chinese design?

Did a little googling and seems like those MJE's should be matched(in normal JLH)
and I also noticed that I shouldnt clean these parts with acetone -eats the graphics right out.
at least I now know that the caps weren't panasonic:)

 
Back
Top