New LM3886 Gainclone, pleasantly surprised

[Rodeodave]

Allright, this is my first Gainclone amplifier, and I must say I really like it. It's the third amplifier I've built so far, the others being a ZenLite and a MiniAleph (respectively).

The circuit...On the primary side of the 330VA 2x30V Avel toroid there's an IEC-filter, a fuse, a switch (Marquardt) and an NTC (I believe it's an Epcos NTC 10) which connects the two primaries in series (230V country).
On the secondary side there's a regulated PSU based on a design by carlosfm, realized on a modded chipamp.com pcb with MUR860 diodes. The two LM338 regulators sit on a small grounded heatsink.
The amplifiers are DC-coupled non-inverting LM3886TF with snubber networks fitted to the caps and the output, similar to Fig. 1 and 2 in the datasheet. The pcb's and chips are from chipamp.com as well.
The electrolytic capacitors are all Panasonic FC or ECA and the resistors are metal film. The coupling caps are 2.2uF 250V Mundorf M-Caps. The wires are all standard heavy gauge copper litz wire and the input cable is RG316/U coax cable.
The chips are mounted on quite large heatsinks whose fins are placed in line with the enclosures' venting slits. The enclosure itself is a 2U (80mm) Galaxy Maggiorato 330x230mm with 10mm solid silver frontplate from hifi2000 (modushop.biz). These enclosures are very well made, with a tough anodizing on the aluminum parts and a plastic coating on the steel top and bottom plate, plus they're optically appealing, methinks...that's why I bought four of them
The RCA connectors are the standard Neutrik (NYS367) and the speaker connectors are the standard ones from Farnell (BP2616), solder is oldschool SnPb. The power LED is amber, btw...

Firing up the amp...was easy, no wiring mistakes or wrongly oriented parts, and after some quick measurements that confirmed that all the voltages were where they're supposed to be, I connected some cheap speakers. Not very surprisingly the left channel had some very slight hum due to the wires running very close to the toroid. The scope confirmed that there was a maybe 4mVpp spiky-looking waveform at the left output, more square than sinusoidal. I was expecting more hum, but maybe tightly twisting all the wires does do something after all
I then used some thin copper sheets which I tightly wrapped around the toroid to shield it's stray field a little. It took about three A5-sized sheets to nicely cover it with sufficient overlap between the sheets. The top mounting plate of the transformer is in direct contact with the copper, and thus the whole shield is grounded via the bolt. Care has been taken to not have the shield form a shorted turn around the doughnut by using a foam coaster, the plastic coating of the enclosure does the rest.
This greatly improved things, now I have to place my ear almost right on the speakercone to be able to tell whether the amp is switched on. But that's the least one can expect from a regulated PSU and a high PSRR.
Even after several hours of operation nothing is above 40°C I'd guess (T_amb 25°C), very unusual, being used to class A...

The sound...I have to say I'm beginning to understand all that gainclone hype. I can only really compare it to my MiniAleph (f.e. see The Mini-A), but I think I have an idea of the source material by listening a lot with headphones. My audio source at the moment is a Lenovo s10e netbook running from battery, with it's Realtek ALC269 audio chip, mediamonkey and flac audio files, plus a cheap diy interconnect. The headphones are AKG 501 and Sony MDR-EX71SL (Linkwitz seems to like them, see Reference earphones). The speakers are Fostex FE108EZ in the recommended BLH enclosure with some diy speaker cables.
The MiniA has the sweetest midrange and incredible, almost holographic soundstaging. As if Johnny Cash stood right there in the room. As I know now, the bass is a bit loose. I can't prove it, but I think it's due to lower wattage and lower damping factor (below 20 I'd guess). The GC in comparison sounds a bit colder (more resolved?) but has/produces extended bass. It seems to have more control over the cone, effortlessly. The mids and heights are all there, all very balanced, but a certain magic is missing. The amplifier now has maybe 30hrs of operation, and after my ears were burned in after two or three hours the sound didn't change anymore. The amp is in no way tiring, but a bit more confined than the MiniAleph I'd say. All in all I'm very pleasantly surprised. And I envy the friend I have built it for...

[Bone]

I hope those copper sheets around the toroid are insulated otherwise you could end uop with a shorted turn and a severe fire!!

[Rodeodave]

Yes, they're insulated from the chassis and the bolt.

Originally Posted by Rodeodave

...
Care has been taken to not have the shield form a shorted turn around the doughnut by using a foam coaster, the plastic coating of the enclosure does the rest.
...

[aptquark]

man thats nice...and the choice of colors for the wires was spot on! Nice layout. I WANT ONE!

[Dennis Hui]

Those are surprisingly large heat sinks for a chip amp. Very very
nice construction.

You friend must be really happy.

[Rodeodave]

Thanks, guys! That friend of mine is indeed quite happy, we recently gave the amp a listen together. Hope it still sounds so pleasing through his speakers.

Anyone having an idea about the output impedance or damping factor (into 8 Ohm) of an LM3886TF?