Sunday, June 11, 2017

More AC Power, Please

I was just feeling warm with the incredibly ware, rich, natural, grainless, presentation with all the latest changes, including polarity fixes.  Then, last Thursday, the Krell FPB 300 shut down with only 1 light.  The manual says that means shorted speakers or cables, or insufficient AC power.

I run my entire stereo on a 20A dedicated circuit with 10G wire, but not specifically the Krell.  I see now that Krell recommends running the 400, which mine in power already effectively is (when they fix everything, they set it to the 400W program, the 400 and 300 they've always been the same anyway, except for Cast input and other input board changes.

The ultimate peak power doesn't change, but the ultimate bias levels permit 400W class A at least for a very short while (it can't do that for long, the temp will rise above the cutoff and go back to Plateau 2).  This is a lot of power.

Anyway, I also was somewhat worried about the new Master 7 Singularity DAC.  But after much trouble, I finally figured out how to test the XLR offset DC voltage, and there was none (though both + and - were 0.5v above the ground pin, which should be OK).   There might be other issues, but before finding my wonderful new dac unusable I should investigate less expensive possibilities.

I also (finally!) wound a set of new PTFE speaker cables from solid core teflon coated silver plated copper milspec wire I bought on ebay.  I used that the replace the Canare 4S11 speaker cables I had custom made by BlueJeans.  The Canare is low inductance and high capacitance.  This maximizes any peaks, and can set unstable amplifiers oscillating.  The Krell should not be unstable at all, but the draw of the cable is almost like a short at super ultrasonic frequencies, perhaps that is setting the Krell to anticipate the maximum bias level, even though I was listening at -15dB.

Whatever, I decided I also didn't like the screw on terminations, just the slightest looseness could cause arcing.  So I decided I was moving on, but not to the old junk but to the new project I hadn't gotten around to for a year.

So I put on the wires, ran the same song at what I thought was the same level (I had briefly turned it down, I couln't remember exactly what it was originally).  And, halfway through the song, the Krell shut down again.

OK, so now I tried replacing the Nordost Baldur 0.6 meter input balanced wires.  Now, no shutdowns, no shutdowns at louder levels, I raised levels all the way up to an earsplitting -4dB on this song (which has very impulsive flutes and metallic sounds, drums, etc, an audiophile recording which originated on SACD which I digitized to 44.1 to play on my Sonos system.  It's "Taquita Militar" from a Crystal Cables sampler.  It's a very tough high frequency test on an extended bandwidth system.

Anyway, so dumping the $500 audiophile cables (which by conventional standards should be OK, just slightly higher than normal capacitance, and extremely low inductance, and supposedly vanishingly now propagation delay) have fixed the problem.

(BTW, Baldur was discontinued by Nordost several years ago.  Perhaps a lot of people had trouble with such a fast interconnect.)

But I still think, if I had more AC power, perhaps the Krell wouldn't have shut down but actually rose to the challenge of the highest bias level.  Even free of the Baldur cable (though, I think I'll be trying some ohter things.  I've ordered Belden 1800F cut to length by Blue Jeans, and I'm looking into some other PTFE cables cut to spec).

Mind you, I don't really know if the problem is inadequate AC power.  But a friend of mine thinks I have so much equipment it shouldn't all be on once circuit, even 20A, and that is correct.  I should have at least 2 and maybe 4 circuits to be fully audio grade (each sub on it's own dedicated circuit, as well as the krell and the conditioner).  And with balanced connections to the Krell it won't hurt it to be on a different circuit, and the Subs already have balanced connections to their DAC (an Emotiva).

No comments:

Post a Comment