I've been doing a lot of web searching, and there are things that are close, but I can't seem to find exactly what I'm looking for.

Headwize comes close, and Headroom comes close, but neither is quite "there".

I'm trying to find plans for a small, preferably Class A, headphone amp. Something that I can use to listen to music late-night, or from a portable, with as high a quality as possible, for a reasonable amount of money. I've worked my whole life in electronics, so a "kit" or plans would be preferable.

Something along the lines of this, or larger would be great.

Anyone have any links, or ideas about this?

Thanks; Artie

Have you had a look at Audiologica?
I had DIY plans myself a while ago, but never got started. But, I don't know that much about this, just thought it would be fun to build my own.

Btw: If I'm not mistaken, the site linked above is tangent's, maybe he can help you?

Yeah, check out the PIMETA

I built the META42 a couple years ago, and the PIMETA is a slightly updated version of it. It is an excellent amp. Tangent's site has a lot of good information. If you need help calculating gain values and stuff, I can help you, as well as the people on HeadWize. I built mine with an overall gain of 3 for Grados, and the local gain for the opamp to 250, for use with the AD8620 opamp. This amp will most likely be better and cheaper than the Total BitHead (depending on the parts you choose), and much more fun of course.

Personally I found that biasing the opamp class A had no noticable effect on the sound, but I left it as is anyway since I use a wall power supply and don't have to worry about current draw. Even so, the amp is still only drawing 15mA quiescent at 24V if I remember correctly. I believe removing the class A bias the amp only drew 5mA.

EDIT: And BTW, if you use Foobar special edition, it has an excellent builtin crossfeed plugin, so you don't have to worry about building this addon circuit to the PIMETA if you don't want to.

Another thing, don't underestimate a good clean source. And of course, a good pair of headphones!

Thanks guys, (gals?). Even though thats still not exactly what I'm looking for, it looks like a fun little project that might surprise me with the results.

Then again, maybe that is exactly what I'm looking for, and I just don't know it.

I'm hoping to find something comparable to the Grado-style headphone amp, or the Headroom. That is, something thats ultra-hi-fidelity. Using discrete components possibly.

But, I'll give the PIMETA a go.

Thanks again for the links.

Artie

The PIMETA is definitely much more advanced than the Grado one, and probably the HeadRoom, as I said before. YOU pick all your components, so you can be sure you're getting what you want.

To get a comparison, here's the schematic to the Grado RA-1 amp:


Now, take a look at the ones on the PIMETA site. And the opamp in the RA-1 costs a dollar. The PIMETA is pretty damn "ultra-hi-fidelity." To an extreme. Many people who build them stack the buffers, bias them in class A, and buy super expensive opamps, things that probably are above and beyond normal audio sanity. And The design is solid. I'm not quite sure why you think its somehow subpar.

If you absolutely want to build the ultimate discrete amp, build a Gilmore variant: http://headwize.com/projects/showproj.php?...ilmore3_prj.htm

You'll find plenty of information on this design in the HeadWize and Head-fi DIY forums.

Wow . . . I'm amazed that thats all thats in an RA-1! Doesn't that thing cost over $300?

I think I'm going to build both the RA-1 and the PIMETA, just to hear them, then, if I still need more, I'll try the Gilmore.

From what you've said though, I imagine the PIMETA will do just what I need.

Thanks again freakngoat. I'll let you know how this turns out when I get it done.

Artie

Ooops . . . missed this sentence the first time 'round.

I think, perhpas from my old "audio" days, I equated IC's with low-fidelity. Especially inexpensive IC's.
I'll be very happy if the PIMETA proves me wrong.

Yeah, crazy, huh? I'm sure it (RA-1) still sounds good though; opamps are very nice devices in themselves and don't need much external circuitry.

The PIMETA improves on the RA-1 design with multi-loop topology, current buffer, and optional class-A bias. The power supply uses a precision voltage rail-splitter with current buffer, and high-speed low value caps as well as large value caps for greater stability and transient response. Now add to all that the ability to choose whatever components you want, and hand match their values--that's pretty nice. Also you can optimize the gain for whatever headphones you use.

Yeah please let us know how it goes. As for me, I took all the components out of my META42 for use in a headphone amp for my MP3 player project, so I need to buy some more parts to reassemble that amp. Now I miss it.

The one thing I'm a little concerned about, is that the PIMETA uses some Burr-Brown parts. Burr-Brown is known for their exceptional quality, but they were bought out by Texas Instruments. I hope those parts are still available.

Artie

They sure are. In fact, those Burr-Brown parts (the buffers) were chosen based on the fact that the previous recommended Elantec buffers are no longer available.

Anyone know if there are any true digital heaphone amps out there? We've seen a few digital home theater amps, and it seems like a natural progression to move this technology to the headphone realm.

freakngoat: I noticed on Grado's website, that they talk of a "modded" RA-1 that had a bit higher output power to drive non-Grado headphones, (RA1-HG). Any chance that you might have that circuit also?

negritot: I think you might be confusing terminology here. There's no such thing as a "digital amp". All amplification takes place in the analog realm. I think that maybe this comes from an early marketing ploy that speaker manufacturers used to try to sell speakers when CD's first hit the market. They were called "digital speakers". A completely made-up term. They were the same old analog drivers and cross-overs that they always were.

If you pressed a sales rep for an explanation, they back-pedalled a bit and said they they meant that they were designed to handle the increased dynamic range of digital audio. (Which was, again, more BS.)

Unfortunately, I suspect that modern amp/reciever makers are trying this same sales tactic to sell their otherwise unexciting new wares. I can't say for sure, because I haven't actually shopped for anything new in years.

Those old analog amps still work great!

Artie

If you change the term "Digital" to "Class D", what would the answer be then? Class D is the so called "digital amps" right?

I don't have any first-hand experience with Class D amps, but the website that you linked to said this:



I don't want to be too quick to judge, but it doesn't sound synonymous with high-fidelity.

Its likely just configured for higher voltage gain. All you do is just change resistor values. The original RA-1 is configured for a gain of about 4.8. This is not high enough to drive higher impedance headphones such as the higher end Sennheisers and AKGs to sufficient volume levels, whereas it is plenty for low impedance Grados. Like I said earlier, I configured my META42 for a gain of 3 for use exclusively with Grados. It gets PLENTY loud.

To be honest, the schematic for the RA-1 can pretty much be taken out of any introductory electrical engineering book or opamp datasheet. It is a standard non-inverting opamp configuration with a DC blocking capacitor and volume control. The high gain version is likely not any more advanced; like I said they probably just changed resistor values.

check TPA6120A2
http://www5.head-fi.org/forums/showthread....25&page=1&pp=25

I bought an amp off the Head-Fi forums and it is supposidly an RA-1 clone with the same opamp and all. I could take it apart and take some pictures if anyones interested. It does sound spectacular with my sr-60's.

This guys stuff comes highly recommended:

http://www.z-audio.com/


C

It's not unheard of (though it is rather unwieldy) to use a Gilmore or other discrete class A as a portable, using a lead-acid battery. Such setups strain the definition of "portable" though.

In general discrete class A amps suck up lots and lots of juice and there's no way around it. Moreover, they aren't really guaranteed to have exceptionally good noise characteristics, due to the heat involved - I've occasionally been able to hear the hiss from my Gilmore on my Etys.

My advice is to choose an opamp based amp strictly on the objective ratings and price. A friend of mine bought a META42 clone for some obscene price ($50 or so) on eBay and loves it.