One poster has stated that this l'l amp sounds like a 700watt amplifier. To that, I can only respond - Exactly who's ass was he blowing what kind of smoke, up?
If that little pancake torroid transformer is 300va, I'd be truly amazed. Not unreasonable for the price, and overall, this amp is an impressive material value. But if you compare it other value products, Norh for example, you'll find substantially more core/watt output. My own (not a Norh) 100watt monoblocks each contain a 330va torroid transformer. My 50watt Llanno monoblocs (cool running low bias), have 350va torroids, each. This is not a criticism of the SLA, simply an acknowledgement of economic realities. Made in the PRC or no, the economics of capitalism is a constraint. A powersupply, shared beween both channels is to be expected. And the marketing driven 1U rack size, and its shallow depth, constrains both the transformer and capacitors. From the power output, I assume the rails are approaching 50v, too close for 50v electrolytics, so you're constrained to 63v, or higher, caps on the rails. Its not that bad, in that 63v electros are generally at the sweet spot in dissapation factor.
One 4,700mf per ps rail is all they can fit in these cramped quarters. Where would you put the minimal 30-40,000mf one might expect in a 'world beater' 100w/channel stereo amp? Well, you could replace them with Panasonic TS-HAs, leaving the total capacitance alone, they'd probably fit and sound better. But if you disable the fan, due to its reported negative sonic impact, and leave those unperforated covers on, those caps are going to degrade quite rapidly, regardless of the fact that they're longer life and higher temp rated than the generic 85*C provided. Because if you run the amp hard, that little transformer is gonna heat up.... and without the fan, the heat has got nowhere to go.
Then again, maybe the compliant nature of the ps configuration is ideally tuned to this specific amp. I doubt it. Rather, I think is the obvious result of size and economic constraints. For lack of specifics, I can't comment on the state of its local ps decoupling, the IC power supplies, etc...
So whats a fella to do? We could go upwards, placing the caps vertically, as opposed to the present lying down orientation, possibly getting double the capacitance(or more), at the cost of having to obtain a new top. But unless both the bottom and the new top covers are perforated, providing convection airflow, you really haven't accomplished much, other than temporarily provide an improvement, that will rapidly dissapate.
The only thing I can think of is either -
Remove the transformer from the amp, putting it in a seperate chassis (a BudBox for economy?) which would give room for additional and/or better quality (low impedence at relevent frequencies), and possibly local decoupling and bypassing.... You would remove a major? source of heat within the amp, while providing additional room for different/larger capacitors. I'd still perforate top and bottom if I found the fan detremental to the sound.
or
Buy 2 of the amps and run them as monoblocs, disabling one channel on each. This would give each channel it own powersupply, transformer, rectifier bridge, and capacitance. If anyone gets 2 stock units, I'd be interested in a comparison between a stereo unit and the 2 amps used as monoblocs. If you're willing to give up the ability to return to stereo/unit operation, you could gain more space by trashing one of the channel (being careful as the inputs are all on the right channel, and the outputs on the left channel, i.e. each channel of the stereo amps are not stand alone)
And the rectifier bridge, many expouse the benefits of Schottkey or fast/soft recovery diodes, snubbers optional. If a dropin replacement of such types wouldn't fit, one could obviously put the bridge (along with some capacitance, in the same that chassis that you're putting the transformer in. Or if you're running as monoblocs, put the new diodes on the heat sink you're not using.
And those numerous other little electrolytics not used in the signal chain, obviously, there would be various levels of incremental benefit by upgrading them. I'm not talking high priced exotics, nor am I advocating mass replacement, but where appropriate, things like Nichicon KGs, or the numerous brands of low esr (and low impedence at relevant frequencies) brought to you courtesy of the switching ps demand.
Any part upgrade, or architectural change, whether in the signal chain or power supply, must be evaluated not only for its absolute benefit, but in comparison to its next best and cheaper competitor. If our metrics were straightforward, and individual changes were orthogonal, optimizing performance for cost would simply be solving a system as a lp problem. But they aren't , they aren't, and it isn't.....
As to powercords, I/O connectors, circut breakers.... Knock yourself out if thats where you think that will maximally float your boat / $. Certainly, the Volex 17604 @$5ea., are the ultimate, next-best, cheaper powercord. Spend <$5 on a resettable circutbreaker that fits exactly in the existing fuse holder hole? As long as you can compentently work with AC, it won't hurt, and these breakers offer less resistance under high current draw, and you're not screwed when the fuse blows, and inevitably you can't find a replacement. As far as beautifully machined input sockets and binding posts, if it give you pleasure, either sonically or from an 'audio jewelry' perspective, and you can justify the cost, enjoy! Given the innards, that would not be the first place I would put my money, as you can buy a whole passle of OPA2604s for the price of one machined phono socket.
