kevin gilmore

well it only really makes a difference if you have a dac with significant amounts of dc offset. otherwise who cares.

that is a good question. i'm not really sure of the answer, why not build both and let us know if it sounds any different. however using the input board gives much more common mode rejection. At least 10db, probably closer to 15

depends on how much heatsink you have. 150mv is a good starting point

I would go with the new input board above, and then 2 x cfa2 output boards per channel.

With a sine wave generator driving the amp, would make a perfect tig welder. Change the frequency to control bead size.

uberamp will drive anything i can think of. no protection circuits of any kind, and just about unlimited current capability.

the carbon for dht definitely works as does the kgbh-bipolar-dht the carbon for 6ca7 evidently does not work, congo5 has some spare boards contact him for a pair, might be an assembly error (doubtful) or some bad parts the shrunk boards fit a 100mm heatsink vertically.

2 input boards, 4 driver boards, 4 output boards one grlv @ 30v or 2 if you really want to do dual mono one grlv @ 40v or 2 if you really want to do dual mono one or 2 bigass transformers, stack of unregulated caps and 25 amp diode bridges mounted to the bottom. current version looks a bit different. for the driver board, can stuff parts either top or bottom (other than opamp) and all the holes line up to stack all 3 boards on top of each other.

the opto version only works in one direction, so the output needs to be at least +10 before turning on the servo

yes you definitely need the opamp, otherwise the opto has no drive.

use 2 and make sure the pass transistors have enough heatsink, as in bolted to the main heatsinks

the professional... (picture 9) http://www.head-fi.org/t/798779/sennheiser-he-60-baby-orpheus-w-hev-70-amp-original-box-immaculate the amateur seriously, ray has turned into a real sleezeball...

don't let any of the idiots over there know about this, the first thing that will happen is that someone will plug the tubes into a bhse directly and destroy $2k in tubes and probably trash the rest of the thing. these dht's have virtually identical curves to 6ca7 in triode mode.

i have over 200 op27 in stock. where i got them, no clue. but better than paying for them. finally had to buy opa445 from mouser as i ran out. one chip dead out of the package. have not seen that in a very long time.

you know that its possible that the hifiman amp has output transformers in which case the 300b would actually work as push pull. If so, what is that massive heat sink in the middle for. Still completely stupidass.

SE to balanced is what you probably mean, and you can do that. you can also do SE output, but then why?

Sure. Lower noise

I use 10k pots for everything. makes it simpler

http://gilmore.chem.northwestern.edu/currentamps.jpg

the stax mafia are definitely planning on a multibit dac. Problem is finding switches good enough (and fast enough) to do 24 bits, 1/2 lsb

http://gilmore.chem.northwestern.edu/g9f21766.jpg

https://en.wikipedia.org/wiki/Joe_Alaskey

I got my hands on the old Tektronix all tube curve tracer, the high voltage version goes to 1700 volts. I tested about 50 6ca7 tubes, no 2 indentical. not even close. did the same thing with c2m1000, and to the resolution of the screen, fairly easy to get bunches of matched quads but there is still some variation.

Is that how it works. I've been doing it wrong all these years.

its not like a voltage amp, so a balanced version is still the same output current. other than the cast output circuit i'm unaware of any balanced transconductance amps.