Laowei

20 bit minimum on the spec sheet. Probably a guaranteed performance spec. Actual performance in an optimized circuit may be better? <damn autocorrect>

Why? Beats me.

^ True that. I always have left my DACs on to get the best sonics out of them. The Yggy just takes an unusually long time to really shine.

Listening to Frank Zappa's Sheik Yerbouti ATM with the Yggy.. Must have played it thousands of times over the past 35+ years since it was first released. No shit, it sounds like a live and remastered version with so much natural detail, tonality, transparency, sound stage, yada yada yada that I have not heard before. I am impressed.

Try leaving it on at least one time for a full week and give it a listen along the way. My Yggy has been on for 8 days now, and it sounds substantially better than when it's cold. And the process is not linear, the Yggy's sonics come and go but really bloom after a long power up time.

Good reading JimL. Shouldn't the current losses of the global NFB be included somewhere?

Really great design and execution, Kerry!

Great! I am looking forward to one day getting a pair made. Please take your time, I did not intend to pressure you. I appreciate all that you have already given so much to us DIYers as a gift.. And thank you too for getting me involved with Stax.

Will the gerber files be made public?

Amazing work as always Kevin.

I got mine yesterday. Too soon to tell its potential when fully warmed up, but straight out of the box it sounds really really good with my KGST and SR-009.

Sorry I was not clear. I was thinking of SMT for the active parts.

Smaller board with more SMT?

I have a pair of unstuffed v0.9 Carbon PCBs. Tracing the circuit, it appears the outputs are grounded gate, kind of like a SS Blue Hawaii. Is this correct?

http://www.apexjr.com/wire.html#TefloN For shorter lengths of Teflon silver coated copper wire,I have bought from APEX.

Can't confirm, looks like conventional bottom view numbering on the Japanese data sheet. http://ayumi.cava.jp/datasheets/6R-A6.pdf

I believe the 9RA6 has a 9 volt heater too.

I like the Japanese made 6S4As better than the GE, RCA. and Philips EGC that I have sampled. All have different plate structures. I believe them to be repinned 6RA6s, a Japanese tube with 40% lower Rp and higher dissipation. Heater current spec for the 6RA6 is higher too, and measurements on the JAPAN 6S4A bear that out. They are easy to spot with flags between the plates and the getter like on an EL-34. That said, I am currently using Tungsol 3 mica, square getter 6S4As. The internal plate structure is indentical to the Japanese mfg tubes without the flags. Probably the tooling was sold to JAPAN and tweaked for the 6RA6. These Tungsol sound great in the KSGT.

The KGST is not a production amp. Kevin originally designed it for 400VDC some time ago. That's where the 1K5 value resistor came from. Birgir helped in designing the PCBs, built and debugged the prototype. From that build he and Kevin settled on 350VDC supplies, so the value of that resistor may not be optimum for some builds. Depends on the actual supply voltages, actual CCS settings, actual tubes or any combination. Since that time Kevin and Birgir have gone on to design more than several new amps for Stax HPs. Let's keep them free and creative out on the cutting edge not having to get into the little BS running changes. Hell, the circuit is 99.5% there. You are coming to the party late, so don't expect anyone to be tidying up the designs. They are a constant work in progress, given to us to work out for ourselves. Besides you will learn more figuring out the circuits by debugging them. And you can ask away as you have on this forum for help. That is why this is called DIY.

Because this is DIY?

You don't want to use the isolators that I first ordered. They were wrong, I modified the heatsinks and smoked a CCS. Back in this thread there is lots of talk about isolators. No tme for me to search ATM, got to run for the factory shuttle.

The PEEK screws work great in my KGST running at a low measured 9mA plate load per tube. I would be cautious in using them in a higher stressed area. For the M4 and 6-32 size the rated assembly torque ~2.5 inch lbs. which is half the yield torque of ~5 inch lbs. For best thermal coupling, you should use stainless steel screws with isolators (and thermal grease) torqued to 10-12 inch lbs.

http://www.solidspot.com/peek_screws.html Best prices that I have found online. I have bought from them, good quality made in Japan product. Good people here in the US to deal with. Note: Per 100 unit quanities, Imperial is considerably more costly than Metric. Metric are offered in lesser quantities too, saving even more.

With the jumper pulled deactivating the servo, power up the amp. First adjusting the Balance trimmer for close to 0VDC between + and - outputs. Then adjusting the Offset trimmer for close to 0VDC between either output to ground. Repeat in order several times. Let the amp run for a half hour or more. Remeasure and readjust as needed. Turn the amp off and replace the jumper after a few minutes. The circuit is so stable once set, I don't bother with the servo. YMMV.

Best to go by the PCB, as that is further developed from the schematic. And yes, it is either A1486 or STN9360. Not both.