stax t8000 clone (well sorta)

[wink]

[kevin gilmore]

its birgir's birthday today, the big 35!

so...

special security screws on the top, 6 sided ultra miniature metric spline, apple tools work

power conversion relatively easy once you get rid of the goop

 

 

[Whitigir]

Woaaa! Thanks KG.  I can't wait for what is coming next

[joehpj]

Whoa! Looked much more complicated than previous Stax amps.

Looked like regulated heater. Zener string for front end tubes?

Edited July 5, 2017 by joehpj

[kevin gilmore]

2sc6127 for the outputs and the voltage gain stage, resistors, for pull up on the voltage gain stage, darlington output

 

need better picture of servo, so that one later, stacked ksa1156 for the tube drive, the rest of it is a modern 717 with available parts. 

 

wiring, 100v   J1,J3,J4

120v   J2,J3,J5

 

Edited July 5, 2017 by kevin gilmore

[JimL]

Cool, looking forward to the full schematic.  PS looks like 78/79 front end regulation (the two heatsinks) and for the extra slot, but R-C high voltage?

[kevin gilmore]

next to the transformer is 7815,7915  which makes +/-15 to the slot and also the servo board

filaments are DC

HV is unregulated Cap,100 ohm resistor,Cap

servo pretty much identical to the T2 servo

the 15k resistors are something else

 

 

Edited July 6, 2017 by kevin gilmore

[wink]

16 hours ago, kevin gilmore said:

its birgir's birthday today, the big 35!

I have heamhorriods older than that,,,,,,, .

 

btw, happy birthday, Birgir...  may the Kæstur hákarl be extra savoury....  

[kevin gilmore]

here you go, a whole bunch of resistors are the wrong values, when I have time

Power supplies are 350v

 

 

t8000schem.PDF

Edited July 8, 2017 by kevin gilmore

[Whitigir]

Thank you Kevin!! You are the best of the best

[JoaMat]

Thanks for schematic.

Guess I'll stick to the t8000DR. The servo is interesting though.

[kevin gilmore]

power jumper block cleaned of all the epoxy

j1,j3,j4 for 100v remove all others

j2 j3 j5 for 120v remove all others

j5,j6 for 240v remove all others

and servo board pictures, unnecessarily complex

tube rollers are going to have a lot of trouble with this, the servo has a pretty limited range, and the Russian similar's definitely will not work

 

Edited July 6, 2017 by kevin gilmore

[eggil]

Kevin: did you even listen to it before you took it apart?

 

[Whitigir]

I bet Kevin was more eager to strip her up before speaking or listening to her .  I appreciate these above findings and pure electrical porno

Edited July 7, 2017 by Whitigir

[wink]

 

2 hours ago, eggil said:

Kevin: did you even listen to it before you took it apart?

He shook the thing to see if it rattled.......    

[kevin gilmore]

17 hours ago, eggil said:

Kevin: did you even listen to it before you took it apart?

 

I was listening to it while taking pictures...   multitasking

14 hours ago, wink said:

 

He shook the thing to see if it rattled.......    

absolutely, don't want any metal nuts floating around in bad places

2 of the screws that hold the bottom plate on have very long screws and nuts, both were loose, and one is almost impossible to get to without taking off the front panel.

Edited July 7, 2017 by kevin gilmore

[fat joey]

i am sorry, suddenly lost all the interests about this when i heard it's only a slightly more powerful tube-fronty 727 without changing the pot

 

 

 

-

[JoaMat]

Kevin, function of switch in test position? Resistors in red circles below, have something to do with test?

[kevin gilmore]

simplified schematic   just updated, I got the current sources flipped.

when put in the test position, it puts fixed resistors to the plates of the tubes to -15

so then you adjust the main offset and differential. then move back to the on position and further adjust the servo pots.

when working right, keeps the outputs to less than 100mv. but range is very tight.

Edited July 8, 2017 by kevin gilmore

[Rossliew]

How's the sound compared to your other KG amps, specifically the Carbon?

[joehpj]

4 hours ago, kevin gilmore said:

simplified schematic

A few noob questions.

There are some more parts after voltage dropper to get +100V. Is that because the B+ in T8K is unregulated? It seems stax use regulated PS for the tubes.

What's the function of Q19, Q21, Q27, Q30? These seems to first appear in T8K.

Stax used some low-noise parts like KSA992 and KSC1845. And stacked KSA1156. Any chance to adopt them in Tube-carbon/Sandwich or even Carbon? 

Feedback is 200K in Carbon and 300K in T8K. So maybe there chance to lower the feedback in Carbon?

Most important is how do you think about the front half of T8K comparing to your design?

For the negative supply for the tubes, I like the way tube-carbon does which doesn't need -15V supply. But I don't know the noise level difference between GRHV and 7915/GRLV.

Thanks.

Edited July 8, 2017 by joehpj

[kevin gilmore]

q19 and q21 are in all stax amps. limits the voltage of that particular section, they never trigger, and I eliminated them a long time ago.

q27 and q30 are used as temperature matched diodes. same as Q8 which is mounted on the heatsink

the 300k resistors result in a voltage gain of 1k (actually a bit more) when combined with r38,r39. carbon does the same thing with lower resistances

they use stacked ksa1156, a pair of 400v parts. I use the stn9360 which is a single 600v part. there is no reason to do this in solid state front end amp because the 100v is 15v

main power supplys are not regulated. and the servo jumps all over the place when the air conditioning kicks in.

 

 

[JimL]

1 hour ago, joehpj said:

A few noob questions.

There are some more parts after voltage dropper to get +100V. Is that because the B+ in T8K is unregulated? It seems stax use regulated PS for the tubes.

What's the function of Q19, Q21, Q27, Q30? These seems to first appear in T8K.

Stax used some low-noise parts like KSA992 and KSC1845. And stacked KSA1156. Any chance to adopt them in Tube-carbon/Sandwich or even Carbon? 

Feedback is 200K in Carbon and 300K in T8K. So maybe there chance to lower the feedback in Carbon?

Most important is how do you think about the front half of T8K comparing to your design?

For the negative supply for the tubes, I like the way tube-carbon does which doesn't need -15V supply. But I don't know the noise level difference between GRHV and 7915/GRLV.

Thanks.

Q27 and Q30 have base and collector connected together and so behave like diodes and in series with the LEDs D2 and D3, provide a constant voltage for the constant current sources Q26, Q28 and Q29.  

 

I believe that  Q18-21 form a modified Vbe multiplier - this part of the circuit is not new - the exact same topology is seen in the T1, 600 and 727, among others.

 

The circuitry around Q31-33 is a voltage regulator to stabilize the voltage to the tubes and intermediate stage.

 

I'm sure KG will come along and correct me if necessary.

[spritzer]

On 6.7.2017 at 7:05 AM, wink said:

I have heamhorriods older than that,,,,,,, .

 

btw, happy birthday, Birgir...  may the Kæstur hákarl be extra savoury....  

Thanks.    I'm glad my own birthday present from me to me is bringing Kevin joy. 

[spritzer]

Kevin has sent me all the pics now and well...I'm just shocked how bad the T8000 is.  Now the circuit is blahhh to say the least and how exactly were Stax working on this for 3 years?  Somebody looked at the stuff they already had three years ago and then remembered it again 6 months ago?  I mean nothing is new here and that even extends to how this is built.  Most of the innovation is just because of necessity while the rest could just have been used in the T1 from 1987.  Seriously, links on a modern pcb with surface mount parts?