[erix]

Have a look at the images of Headroom amp guts.

The Home uses one Talema transformer (toroids available from Digi-Key) and two regulators to make +/- voltages.

The Maxed Out Home uses two Talema transformers and four regulators (for + & -) for seperate left and right channel power.

The Max uses the same configuration but with Avel/Lindberg transformers.

All these power supplies are mounted to the same circuit board as the amplifier parts with no perceivable sonic degradation. It's tider that way.

I would embark on cloning the Maxed Out Home PS for the uber-amp simply because you can get all the parts at Digi-Key.

Barring that (and if you insist on a seperate board for the power supply) you could use a ready-made (or easy-to-assemble) one like Welborne Labs' PS-1.

I've built stand-alone power supplies using the that portion of the circuit of the SDS amplifier project on Headwize with good results.

ok,
erix

 

[eric343]

geom_tol: Sorry, it's 93F4189.
thomas: According to my oscilloscope, it's either a linear supply or a switching supply that's so well designed that the "megahertz" noise is insignificant. (< 20mV) Also, since the opamps we're using have their bandwidth trimmed down via multiloop, noise at those frequencies shouldn't matter.

Also, the +5V is only .86A, and I doubt it's not just "thrown in for free", it's just that I don't see an Elpac linear p/s that supplies only +/- 12V.

 

[Audio-Me]

Thanks for setting that straight, two transfomers four regulators on same board. The problem is, where to get a custom board? aos?

I'd like an even better power supply than the MOH, so which toroidal transformers and regulators to use??? What kind is the one in Walter Jung PS?

 

[puppyslugg]

Hi erix,

What IC regulators did the MAX use? Was it the LT1085/33? Caps?

Thnx.

 

[morsel]

Rail splitters have been discussed in gory detail in many back posts, but here is a summary.

Two resistors do not provide a stable virtual ground unless they are buffered. A TLE2426 contains a precision voltage divider and a 20mA buffer. They are so cheap and small, it makes more sense to use a TLE2426 and a 100mA buffer than two resistors and a buffer. It saves you the trouble matching resistors or adjusting a trim pot, and takes less space. Using a TLE2426 without a 100mA buffer is a cheap low power compromise option that saves even more space.

 

[Audio-Me]

Well wouldn't a dual output ps have a ground? Thus no need for a virtual ground since there is a real ground?

 

[Budgie]

Yes. Dual supply will not need a rail splitter.

(I know this hardly contributes to this thread, but I am bored to hell right now!)

 

[puppyslugg]

morsel:

Are you planning to power your meta with linear reg. psu's? How many? Will you be diying them? Which circuit?

Thanks.

 

[thomas]

Quote:

Also, the +5V is only .86A, and I doubt it's not just "thrown in for free", it's just that I don't see an Elpac linear p/s that supplies only +/- 12V.

The problem is that an additional 5V supply (on a linear ps) must use another transformer/winding and regulator circuit to get there. This would cost just as much as the main output, essentially doubling the cost.

Anyways, if that supply sounds good, use it, just IMO it does not look good for a headphone amp. I've never heard a switcher that didn't take away all the "air" from the sound...(btw, the easiest way to tell what it is is by opening it up, not measuring it

)

If you want to keep the size down and don't want to work with high voltages, you can use an AC wall transformer, then put the rectification/filter/regulator circuit inside the case...

Of course, the best solution is a torrodial transformer (centre tapped)-> walt jung regulator, but that would easily cost more than the amp itself... Then you can use the real ground rather than a ground driver...

 

[erix]

Quote:

Originally posted by puppyslugg Hi erix, What IC regulators did the MAX use? Was it the LT1085/33? Caps? Thnx.

Ya know, I wish I could go back in time to the WOH tour stop in Mnpls! I'd have taken some pictures of those naked amps.

I believe I've read that the regulators are LT - more than likely the 15V fixed variety.

As far as the caps - they were dark blue with gold (or silver) lettering which could make them Panasonics.

There are two good sources of information on the construction of the Headroom amps (three if you count Tyll but it costs too much in whiskey to get him loose enough to talk... Just kidding, Tyll! Love Ya! XOXO). The first is Headroom's own site and the second is the review of the Blockhead in Stereophile.

ok,
erix

 

[eric343]

thomas: Just out of curiosity, what will make the innards of a switcher look different than those of a linear? P/s design is *not* my specialty...

 

[aos]

I think MAX uses two-stage regulated supplies so LT* would apply only to the first stage. I believe I saw some opamps in pictures which would then be in the second, more sophisticated stage.

 

[tangent]

Quote:

what will make the innards of a switcher look different than those of a linear?

A linear-regulated power supply is generally a lot simpler than a switcher. It will have a large transformer, a few diodes or maybe a single monolithic bridge rectifier, a large capacitor or two, and a few linear regulators (usually TO-220 parts).

A switcher will be much more complicated. It will have a smaller transformer, and a whole boatload of capacitors, resistors, ICs and such. Overall, it will be rather small.

Quote:

I've always wondered about the rail splitter in the meta, i don't see the reason for using another piece of active circuitry when a pair of resistors does the exact same thing.

Morsel gave the right answer, but I'd like to reiterate: the TLE2426 is smaller than a pair of resistors, even if you tombstone them, and it gives you the option of leaving the 2001G out if you want. Since the 2001G is up at the front of the board, in small enclosures you need all the space at the front of the board you can get.

 

[erix]

Quote:

Originally posted by aos I think MAX uses two-stage regulated supplies so LT* would apply only to the first stage. I believe I saw some opamps in pictures which would then be in the second, more sophisticated stage.

Yeah, I saw those but forgot to mention them. They could be a variation of the Jung Super Regulator (as seen in Audio Amateur from 1995). You can get circuit boards of that regulator here.

I've not seen the actual article or the schematic but I think it might bear a family resemblance to the power supply in Kevin Gilmore's amp.

ok,
erix

 

[eric343]

OK, Mr. Doubting Thomas

, I opened up the MED-113TT, and lo and behold, a few diodes, three capacitors, two chips (both in TO-220 style cases, one larger, one smaller) and a huge transformer. Read: huge as half the size of the P/S is transformer.

Now, is this linear or switching?

Incidentally, about the whole 5V thing, the reason it has the +5V output is that there weren't any *without* the +5V that I noticed.