Posted by IslandPink (A) on September 07, 2004 at 09:04:55

Just wanted to pass on some observations about the LL1676 and perhaps some of the other amorphous-core transformers, regarding the burn-in time . This subject has been discussed before, but I recently had such a noticeable breakthrough , that it needs reporting, I feel . I have recently had some correspondance with Kevin, Lynn Olson, Gary Pimm and Gary Dahl on this subject.

Basically, I'm using the LL1676 as an input and phase-splitter , as in the recent 'Karna' schematic on Lynn Olson's website , see :
www.nutshellhifi.com/triode1.html
This is a reversed orientation to the suggested Lundahl sheet, and gives a 1 : 0.5 + 0.5 signal split .

I had been somewhat disappointed by the sound some time, it sounded a little 'forced' and 'hifi' , and was getting only moderate improvements after perhaps 250 hours of normal playing . Now, I reckoned the typical input voltage to the 1676 at this stage had been fractions of a volt up to occasionally 1-2 Volts , with 20K loads on each secondary .

What I tried at this stage , was connect-up the CD player output to the line stage ( I was only using smaller phono amp signals for a long time up till then ) and turn up the wick , with the volume on max , to put typically 5-10 volts RMS into the 1676 . I also disconnected the rest of the circuit (ECC99 grids ) and put 2K loads across the secondaries .

I played this arrangement with the CD on repeat for about three hours, just to see if I had progress.

Well, the results were excellent . I got lots more low-level detail, ambience and 'alive' feeling from the system . It lost much of the 'solid-state' sound I'd been struggling with .

I tried another 4 hours a few days later, and got another (smaller) improvement , which now takes the amps transparency into new territory as far as my listening experience goes .

So , I think the advice so far given ought to be supplemented with the proviso that the actual loading level ( it may be voltage that is important, rather than current ) needs to be a significant fraction of the device max handling, to get full run-in within a reasonable length of time . I would advise anyone not getting expected results out of these transformers to consider some burn-in procedure, with care of course.

Perhaps Kevin can comment further or advise on the best way of doing this. Gary Pimm has elsewhere suggested high frequency and voltage working wonders on transformers and wires, but one has to consider the winding current-handling capabilites of course.

Hope this is useful,

Mark J