Single ended class-A headphone amp using two transistors: T2

Status
Not open for further replies.
I think the Mainline is one of the most expensive diy kits I've seen. I had to google and make sure it wasn't a pre-built unit of some kind. If the Mainline is $1300 as a kit, imagine what it would cost retail... If the T2 comes anywhere close to that I'm a happy camper.

The Mainline is a nice amp which has fantastic synergy with Sennheiser HD800 headphones together they make for quiet an addictive pairing. Its also an enjoyable and rewarding to build I would suspect performance wise the Mainline has probably got a bit more to give and at the moment its limited by the capability of my front end (Oppo 205).

Building your own amp is a lot of fun if they sound superb and don't cost an arm and a leg that just makes the whole experience even better.
 
I had to pop into the Mainline today after reading the last posts....even though I am totally enamored with the T2, the Mainline just takes it away to a higher level...its indeed more open and clear and a spacious wide transparent sound.

That said, the T2 proves you can have a great musical experience for less than $200 plus the cans etc..

But if your a real music lover there is room for a Mainline amp as well, cost be darned!
Its not cheap and cheerful for sure!

:>)

I am fortunate to have both here and they both make me smile and tap my toes! and thats what its all about here..

Alex
 
Almost time... I splurged in some areas and cut back in others, using parts I already had on hand.
 

Attachments

  • T2.JPG
    T2.JPG
    338.9 KB · Views: 289
Member
Joined 2011
Paid Member
You spent big money upgrading the input capacitor (C22,C72), and then decided the 3300uF electolytic output capacitor can operate successfully with NO film capacitor bypass at all (C27,C77). Congratulations! Those represent a unique set of build choices.

Have a listen to the Fleetwood Mac track called "You Can Go Your Own Way" (link) Written for trailblazers who steadfastly refuse to follow the beaten path.
 
You spent big money upgrading the input capacitor (C22,C72), and then decided the 3300uF electolytic output capacitor can operate successfully with NO film capacitor bypass at all (C27,C77). Congratulations! Those represent a unique set of build choices.

Have a listen to the Fleetwood Mac track called "You Can Go Your Own Way" (link) Written for trailblazers who steadfastly refuse to follow the beaten path.

Not quite - you missed where I spent the big money! The input caps were already in my stash, as well as many of the smaller resistors, RCAs, switch, and wire. I spent a little on the output caps (Nichicon LKS, not that expensive). And I got a $5 alpha pot instead of the Alps out of curiosity.

The Mills 5Ws however weren't cheap.

The output bypass caps I'll add later. I'd like to hear the LKS without them for a little.

All in all I'm out a whopping ~$160.

Thanks for the link - I do love me some Fleetwood Mac...
 
Last edited:

6L6

Moderator
Joined 2010
Paid Member
Looking good! Very well done!!

Huge congratulations for using parts you had in your collection, that’s always a great feeling when you can do that. :yes:

I would be curious if your impressions of the change in sound if you ever replace the pot with something like a stepped attenuator or a fancy pot. I think you’d be pleasantly surprised.
 

6L6

Moderator
Joined 2010
Paid Member
I very much like the Alps RK27 (blue velvet) as is suggested in the kit. In my opinion I dont think there’s much better, although there are plenty of options that are more expensive. I’ve also tried some of the TubeCad attenuators, and Goldpoint with great success. One of these days I plan on using an Eizz attenuator.
 
Member
Joined 2011
Paid Member
I tried one of those DACT stepped attenuators from eBay, you can see it at the top of the photo attached to post #159 and copied here:

767906d1562969653-single-class-headphone-amp-using-transistors-t2-topless_with_black_board-jpg


In response to an inquiry in #189, I replied

I am glad that I tried out the eBay DACT attenuator module, and in listening tests it seems not to change the sonic presentation at all. I.e. same as the ALPS Blue Velvet.

However, I discovered that I like the ALPS more, for other reasons. The DACT has "only" 21 steps, which is too few for my taste. I prefer the continuous adjustment on the ALPS. Worse, the DACT has less friction; it's far easier to turn the volume knob than on the ALPS. I prefer the stiffness / solidity of the ALPS.

For me: good experiment, worthwhile results, non ambiguous conclusion.

edit- whoops, forgot to say, the DACT I bought was 20Kohms. I bought from the seller who offers them two to a package.
 
I have the ebay DACT and it is excellent, but I agree there are too few steps. In a loudspeaker setup it's not as big a deal, but for headphones it becomes problematic. No complaints on the alpha so far. Channel balance seems to be very good.

I am still formulating my thoughts on the amp. I am enjoying what I hear so far.

I did try the diode bypass mentioned a few pages back and found it worthwhile.
 
I have the ebay DACT and it is excellent, but I agree there are too few steps. In a loudspeaker setup it's not as big a deal, but for headphones it becomes problematic. No complaints on the alpha so far. Channel balance seems to be very good.

I am still formulating my thoughts on the amp. I am enjoying what I hear so far.

I did try the diode bypass mentioned a few pages back and found it worthwhile.

I also tried an eBay DACT and did not like it for the same reasons as Mark. I have also tried this inexpensive RK16 pot from Alps and did not like it as much as the RK27. Maybe the Alpha is better.

RK16312A0BKH ALPS | Mouser

Not only did the RK16 not sound as good but I did not like the 3B taper as much as the 15A taper on the RK27. I have also tried the lowest cost, smallest Alps RK09 with a 15A taper and I like it better than the RK16 but not as much as the RK27.

So my conclusion is, taper really counts and the only company doing it right so far is Alps with their 15A curve. I may sometime take some measurements of a 15A RK27 and try to duplicate it with 1% resistors on a rotary switch. I have also tried Broskie's Glassware VC (two switches) and don't really care for the steps it offers either.

All of the above were nominally 50K devices except the Glassware which was 28K.
 
Impressions so far: The T2 is clean, clear, balanced in delivery and has what I perceive to be very low noise. The easy-to-build design and helpful documentation, coupled with relatively low cost makes it a no-brainer option for a diy build.

I'm using it with a heavily modified Soekris 1021 and Sennheiser HD6XX/650.
 
Last edited:
I also tried an eBay DACT and did not like it for the same reasons as Mark. I have also tried this inexpensive RK16 pot from Alps and did not like it as much as the RK27. Maybe the Alpha is better.

RK16312A0BKH ALPS | Mouser

I have also tried Broskie's Glassware VC (two switches) and don't really care for the steps it offers either.

All of the above were nominally 50K devices except the Glassware which was 28K.

I've one of the 36 step course and fine Glassware pots stuffed with Vishay dale RN60 resistors (The A5 version not sure if that's the one you are referring to) its been sat in my parts box for about four years now.

fwiw I've used a couple of the A5 attenuators in SET and OTL tube amp builds and found them sonically to work very well. My only gripe would be that it would be nice if the switches where of higher quality as they are a bit clunky in operation but for just $22 I guess that is to be expected.

The 36 steps can be configured as 1dB & 6dB steps (which I use) or 2db & 12db steps so in my case its possible to achieve -35dB of attenuation via 1dB steps using the two 6 position switches.

A5 Stereo Attenuator

Thinking about it I might use it on a T2 build (pcb is on its way, thanks Mark :) ) along with a steel enclosure I picked up for £5 off of ebay which has been sat in the cupboard for about the same amount of time. The enclosure is a little bit bigger than required and will need some fettling with to make it work.
 
Last edited:
Member
Joined 2011
Paid Member
If anybody is thinking about replacing the E-153 current regulating diodes, with a discrete circuit including (gasp!) more transistors, as for example in John Luckins's post #215 above,

here is a bit of data.

I built a couple of current sources, neither of which exactly matches his design, and put them on the Current-Versus-Voltage curve tracer. Here are the results. This is measured data from real world physical devices, on a curve tracer; I only used LTSPICE's schematic editor as a handy way to create schematics quickly and easily. These are not simulation results. This is measured data.

The 2-BJT circuit behaves as an ideal current source, at significantly lower voltage than the 1-JFET circuit. But it has more parts. And I have left out the PSRR-enhancement capacitors because they're not needed during curve tracing.

_
 

Attachments

  • J112_82p5_current_source.PNG
    J112_82p5_current_source.PNG
    27.5 KB · Views: 295
  • MJE340_2N5210_37R.PNG
    MJE340_2N5210_37R.PNG
    30.3 KB · Views: 966
Status
Not open for further replies.