kevin gilmore

so I did something like that with an opa1632, everyone over at amb's picked on me, and in fact it sounded like shit. Then Ti released exactly the same thing. Still sounds like shit. Lots and lots of feedback is why. This discrete thing is pure class A and current sources, and no feedback. except for dc servo sounds so much better congo5 is testing the surface mount stuff, currently the output buffer blows up due to power over 15v, so will trim the current a bit, the surface mount transistors can only take 150mw. size vs power. It will run just fine on 15v anyway

yes that is how it works

sure as long as you don't mind the gain, or have a knob elsewhere (or a dac with a volume control etc)

for the input switch you want to ground the - input for SE inputs don't need a switch on the output because you can do balanced and unbalanced at the same time

all transformer leads need to be twisted and soldered. and then cranked down good

you need to lower the value of the resistor in series with the offset pot.

kimchi ---yum my local Korean grocery store has like 7 different kinds some in buckets equivalent to a gallon or more. But I would rather have the steak

all the equi=tech are torroids torroids will have less magnetic flux leakage, but will have a turn on thump probably easier to find the EI versions surplus on ebay for far less than you can have one made for.

10 to 15 ma for the driver board, and 25ma for the output board. But i have tested to 50ma and the heatsinks definitely get toasty that way. No difference in sound after about 25ma

the rest of the power in the circlotron is pure AC for the filaments, so not much of a turn on whoosh either

the way the board is layed out currently its really input -> attenuator -> ubal/bal -> diamond buffer because there is only one set of servo's and the attenuator is best at something like 10K

unbalanced to unbalanced, gain of 1 unbalanced to balanced, gain of 2 etc you can adjust the gain by changing the 1300 ohm resistors, or the emitter resistors up to 10x is possible with no increase in distortion these days, I don't think there is any reason to add much gain

remember that there is both a dynamic and electrostatic current feedback amp. using the bh power supply on the dynamic amp is going to result in much magic smoke.

deleted

final dns change monday morning machine will end up on 165.124.198.148 and dns will change a few hours after that 2/1/17 edit, final networking address, should stay this way till i retire

highest hfe, make sure to match hfe and vbe

i have not touched that one in a while. don't see any reason to, i like it the way it is

transistors are 2sc3324 and 2sa1312

posted final smt version of the ubal to bal board. the smt boards should be cheap because they are tiny, ubal board is 2.42 x 2.12 and the diamond buffer is 1.46 x .8 I would like 4 of the ubal boards, and 8 of the diamond buffer boards please

that was not a buffer. that was a 60 watt + power amp done with semisouth as output devices.. It was pretty nice for a small chip amp.

probably close to final 2.45 x 2.17 inch

lm49830 is a chip amp do you mean opa1632 and its clones, all of which require huge amounts of feedback and have all sorts of common mode dc issues

diamond buffers will stand vertical on that board. there are thru hole versions of both pieces already

moar better 2.65 x 2.27 inch