Sunday, June 25, 2017

Things I do differently (as of June 27)

I'm back to the Dac 19 until I have time to retest the Master 7, or set up a 4th Behringer DEQ 2494 to allow extra delay in the bass line--and also, I would have 4 screens running, one with a spectrum of each frequency range, and one with the level monitor--see, I love these things, except they could be better, such as polarity for one and 0.1db adjustment and delay to 1000msec.

But it's not the same as before I returned from vacation (and soon acquired and hassled with the Master 7, now in hiatus) because:

1) I have subs on a different circuit.

2) I play FM tuners via Lavry AD10 digitization, rather than Sonos digitization.  I could not do this before, failed years ago and over and over, because of the subs competing for power to deal with low frequency noise with the Krell, causing Krell to shut down.  Old tuners have lots of LF noise.

3) I've gotten all Oppo connections, even 88kHz samping (which I had though was simply broken in the Tact for years, except the Tact and Behringer both show the digital looking normal) mainly by tightening the RG6 connections.  This permits me to play discs of all kinds and use the digital outputs.  The Oppo is a better universal disc player for many reasons, including that it handles DVD-Audio better than anything, and does Blu Ray Audio, and to have the digital output from it online is a top level breakthrough.

4) I have corrected the absolute polarity, which had been reversed since using the Dac 19 mostly--because it inverts, though I remember in the very beginning I had figured out that "in" on the setting button meant "in" correct polarity, which is the reverse of my intuitive expectation, then I promptly forgot that for the next couple years and was playing out of polarity.  I now have a simple polarity test.

5) I corrected the relative polarity on the subs.  As of yesterday I am using a polarity reverser connected to short XLR cable to the sub.  That way it doesn't require a tricky preamp adjustment in one channel followed by physically reversing all the other channels, a setup bound to lead to future mistakes.  This, of course, has easily profound consequences.  I have probably down adjusted the bass a bit because it was too much.

6) On re-examination, Sonos outputs do not need to be at -8dB to be 0dB "bit perfect" digital outputs. I had either misdetermined that some time ago, or some update changed or fix it.  Now I have all Sonos digital outputs set to Fixed and there is no clipping as I once noticed.  So now I am getting the full dynamic range through Sonos again, and it's up there with how CD's sounded before.

7) The Acoustats survived a tough ordeal (and contrary to my expectations for awhile) and are now fused at 3A, which should make them even safer.  I have determined the transformers needed for my broken spare interface and re-order.  When fixed, this interface will be modded to match it's pair.

8) Solid core 14ga silver plated teflon coated wire, essentially equal lengths in both channels (a first for me), now secured to Acoustat posts with Furutech bananas.  The wire softens the edge but doesn't lose resolution, compared to Canare 4S11 I was using, very slight difference if any.  The locking bananas are essential and better than previous ones I was using.

9) I remember certain things about using the Krell.  After using breaker to shut off (always turn off at front panel first) wait for complete cool down.  Or, if you fail to do that, and it starts shutting down, do that and remove from power for awhile.  Otherwise, it seems (don't trust me) to be safe to turn off from the front panel when warm, make some change, and then turn on again, without any waiting beyond what it takes to make the change.

10) Because it sounds so much better now for all of the above reasons, I am enjoying actual digitial inputs now.  I haven't used re-sampling from my digital players for awhile.  For playing CD's I can use the Integra Research RDV-1 and take it's digital output (it has the Apogee clock which is probably better than Denon DVD-9000).

11) I now have a fully working DVD-9000 online, it plays DVD-Audio and everything.  My old DVD-9000 has become a spare DAC for now.

12) The super tweeters are off for now.  They need to have their time alignment fixed, and that depends on the disposition of the Master 7.  I now suspect they were 300 ms delayed, which is terrible.

13) I have all room remote control of the L-1000T thanks to a new repeater, and I've reprogrammed the kitchen remote for that.

14) Batteries were replaced in living room remotes, two needed serious cleaning including the DVP-9000ES remote.

(And all this is not counting the Master 7, which seemed to work pretty well until the day it was taken offline, being by far my best sounding DAC, and was part of lots of listening and testing as well as frustration.)

That's the scorecard of actual and mostly huge and objective improvements, just in the past 6 weeks.

What has driven a lot of this in the past 3 weeks is random shutdowns of my FPB 300.  Each time I figure out some fix which seems to help get it past something, and it's back.  While this has in some sense helped lead to some improvements, it's been very frustrating and not helpful to the rest of my life.

Sunday I ran the Dac 19, which had never caused problems before in 2 years, and it was performing flawlessly, until midway through the movie Steve Jobs the amplifier died again.  And wouldn't restart.

I changed dacs to the Denon DVD-9000, and set up a dual DEQ system for the subs to delay them about 330msec to account for that much delay in the DVD-9000.  It played for awhile, then shut down again (playing bass heavy Santana Supernatural, but crossed over at 100 Hz).

So now it appears shutdowns occur for all 3 dacs, so the best explanation remains that there is an intermittant problem in the amplifier or the speakers.

Now, shut down a third time Sunday 10 minutes to midnight playing Supernatural again through the Oppo.  This time (not always before though) it shut down while apparently the DVD-9000 was muting--I heard the relays click and a moment later the Krell relays clicked and it shut down.

Tests I finally did last week showed that either speaker connected to the right channel caused shutdown, but neither speaker connected to the left speaker does that.  All that time I was using the apparently correctly functioning at the time Audio GD Master 7 Singularity.  I should have done those tests the first day.  By the time I had done them, I had already desoldered the HF transformer connection for testing, then I had to solder it back (because, overlooking the amplifier, I immediately and wrongly jumped to the conclusion that the speaker had been gone bad.

On Sunday night I duplicated the shutdowns using the Dac 19 Anniversary that I bought two years ago and has never been a problem, AND on the Denon DVD-9000 that I had been using for a year and concluded was both my best CD and HDCD player (but being semi-broken, it won't play DVD).  You can't say the output of the DVD-9000 wasn't correctly engineered--it was Denon's statement piece.

Since then extensive speaker play of the left channel only on very demanding music has caused no shutdowns at all, many hours so far, whereas on both channels (and right only) shutdowns were occurring in about 5 minutes except at low levels they might go longer.  (My 'rebooting' fix did seem to give me a longer playing time for awhile, but days later the problem came back.)

This problem has got to be a failure in the right channel of the amplifier.  I will be using something else until it gets another factory repair.

One thing: if no failure occurs in a brief test (maybe few hours) you can't tell if you have really fixed something, even if last time it happened in 5 minutes.  That's what intermittent is like

So The Thing which means something is failures.  When failures occur, you can be sure something bad was in the chain then, which you can then narrow down by elimination, replacing with other things.  But once again, if failure doesn't occur in limited testing you can't be sure.  But if you find something consistently in the failure path, with before and after changed, it is most likely the culprit.

No comments:

Post a Comment