[Occam]

Eric,

Like PeAK says, a transformer is quite complex, and as he has already adressed this at the requested technical level, I can only suggest an easy experiment for you and your cousin. Take an ART DI/O, in whatever state of modification, or stock. (or find someone to lend you one) Power it with the 'ultimate, next best, cheaper power supply", a $3.95 Atari 3.4amp 9vAC table wart. (or the Stancor @$29.95). 1st, determine whether this is an improvement of the stock wall-wart.  2nd, tell us if you and your cousin note any difference in reversing the non-polarized 2 prong plug on the upgraded powersupply. Use whatever Xblind protocol you deem appropriate. 3rd, let us know if either, or both of you have a consistent preference.
And lastly, tell us if (within the context of the system in which you did this experiment), if the terms "darker background", "more air" or whatever claptrap, are in any way applicable. I'm not saying that these are the most appropriate descriptors (I'd imagine that a 'darker background' might be the result of less noise on the signal ground, and the resulting lowering of the noise floor. That 'more air' could be due to the same, or might be the result of a change in the frequency/phase response.) Certainly, as Dan pointed out, there are few reviewers with the requisite technical expertise able to link their subjective impressions with appropriate objective measures, let alone coherently communicate qualities.
Personally, I do use a cheater plug to determine chassis potential for an isolated component, I'd never use a cheater plug in situ, because Providence has supplied mankind with 20amp dpdt, center-off, toggle switches

Or simply read this thread -
http://db.audioasylum.com/scripts/t.pl?f=amp&m=18748

When Mr. Crump speaks, I listen. Why? In addition to his reputation among DIYers and in the Audio industy itself, on those few occasions that I could compare my observations to his on the same components, I 'got' where he was coming from. This is same process we would use in evaluating a movie critic. If I disagree with the critic consistently, I don't care what they have to say, and I don't waste my time reading their opinions. If I largely agree with them, I value their opinion. Why should the credence we give a reviewer of audio components be any different?

If one is unable, unwilling, whatever,.... to calibrate your own obsersevations against those of a 'reviewer', whether of books, audio components, cars or movies, indeed, their words are meaningless. There are about 15 people in the Industry who's opinion I value higher that spit. There are 6 out of 70 members of my local audionerd club, who generally are in agreement with my own opinions. Its not that I generally like these six more that the other 64, its simply that I'm far more likely to seriously consider something based on their opinion.

[eric the red]

The  question "Why mod a perfectly good amp" is a good one. The easy answer is because manufacturers have price points that they are shooting for and it's a lot easier to sell a thick faceplate and nice casework that looks cool in somebody's rack then it is to sell something with better internal parts with plain-jane looks. Answer me this then audio smart guy: Do you think that the majority of two-channel audio gear that is being marketed to audio-hobbyists by mainstream audio manufacturers like Krell, Classe etc etc is audio junk that has basically the same internals packaged in different boxes that are sooner or later bound for the Audiogon classified/Ebay merry-go-round?  I can't imagine that in the current two-channel audio market that manufacturers of audio gear are doing much to improve the sonics of their products And yes, all hobbies have their own specific language and ours is no exception.

[audioengr]

As for modding a perfectly good amp, this is exactly what you want to do.  Why start with something that has little potential?

Here is what I would do:
1) replace the bridge rectifier with Harris FREDS
2) replace the two big filter caps with Nichicon golds
3) replace all of the op-amps with either AD825 or OPA 627.  If you need duals, use OPA 2107.  Check the voltages first for compatibility.
4) replace all of the local decoupling caps with smaller FK-series Black Gates
5) replace the signal-path capacitors with Black Gates
6) replace any feedback caps with polystyrenes

Then, I would look at topology and wiring changes, but this is more technical and difficult to describe.

[Ears]

Changed the pos speaker output terminals to some nice goldplated ones today and it was cake....no need for soldering as this is an easy bolt in swap.
The posts that came with the sla-1 left me using audioquest type 4 unterminated cables.
With the new posts I am able to use my ap oval 9 or any other terminated speaker cables.

Now i would like to swap the  1/4 input jacks for rca jacks.
Has anyone done this yet and what did you do about the plastic covering over the board hardwiring connections?

The only other mod i have done,if you can call it a mod,is unhook the fan.
Before any further internal mods[caps wiring ect] the nex't step would be rca jacks so i can use a decent ic without the cheesy ratshack adapters.

Any info on changing the 1/4's to rca would be appreciated.

[Ears]

Just got done finishing swapping out the 1/4's to rca gold.

This entailed carefully breaking the plastic 1/4 jack housing and pulling the 1/4 jack carefully through the inside of the amp.
There is no de-soldering involved.

The surface mount connections are self explanatory as to which is - or + and i was able to solder direct to the connections.

Now i can hook my az matrix directly to the gold plated rca inputs, instead of using those cheesy 1/4 non gold plated terminals.

Not sure if there is any sound quality difference as i am letting it warm up a bit.

So far this has cost about 20.00 for both gold binding posts and rca terminals and less than an hour of time for both.

Anyone with novice soldering skills[ie me] can do this mod easily themselves.

[Ears]

After doing a little listening, i wood say the sonics have improved, if only because of the upgrade from aq type 4 to ap oval 9 and the direct connection of the  AZ matrix rca's without the use of the cheesey 1/4 inch adapters.

The changes are more bass and more air + a wider deeper sounstage.

Nex't for me will be to upgrade the stock pc.....any reccomendations?

[peakrchau]

I did some poking around on the weekend (without a schematic) and found the following:




Input Stage:
The quad op-amp AN6554 uses two of the units for left and right channel in differential to single ended gain configuration providing 3.52 dB gain (1.5x) for single ended inputs.  

The 14 pin package uses pins 8,9 and 10 for the left channel and pins 5,6 and 7 for the right channel. A 22K and 33k feedback resistor set the gain for this stage. A second pair of these is needed for handling the balanced inputs. Two input DC blocking caps and and output cap complete the circuit.

The phono plugs TRS (tip ring sleeve) provide IN+, IN- and ground connections for differential inputs. Single ended Phono plugs will short IN- to the sleeve/ground and provide single input via the TIP portion of the plug.

Both the XLR and Phone plug inputs provide for balanced inputs.  It should be possible to flip the phase of the amplifier by feeding a input to the "RING" and grounding the "TIP".

The whole section section is powered off +/-12V derived  from zener diodes off the +/-50V main DC supply.


Volume Control:
The outputs from the input stage go to a 9 pin header and connect to the front panel PCB. A third op-amp (of the four in the AN6554) provides a Unity Gain buffer from the volume control before going to amplifier section.   The fourth op-amp is provide for phase inversion in bridge mode.


Amplifer Section:
The output stage is a complimentary Darlington configuration arranged in common emitter configuration with about 10% local feedback.  This is driven by 1/2 of the 4558 opamp. Feedback components set the gain to about 25 dB (18.3x). Signals coming in are both high and low pass filtered. The output stage contains a classic zobel network to stabilize the amplifier. There is no DC servo control so large caps are needed to provide for high pass filter. Two small signal transistors (Q19 and Q20) make up an overcurrent detection circuit.  

Speaker Connectors:
The existing connectors cannot take spades due to the external plastic shrouding allowing only for bare wire.  The spacing is standard so I was able to replace it with a simpler flat shrouding. I can now take spades.

Summary:
For an inexpensive amplifier, the ART is feature laden. It is possible to simplify the circuit further by disabling some of these features. Parts swap are the order of the day for electorlytic caps and the signal gain resistors.
Care needs to be excercised if replacing the opamp in the amp stage as a higher gain/wider bandwidth opamp could destabilize it.

A phase inversion feature could be implmented easily enough but the details need to be worked out.

P.S. An up to date version of this text can be found http://www.angelfire.com/ca/rchau/audio/sla-1.html#1.3.1">here