kevin gilmore

Here is another one. I know its different because the caps on top are different. And its 1oz copper.

changed size of cap cfp3rssfixedss - CADCAM.ZIP

latest board file with all fixes cfp3rssfixedss - CADCAM.ZIP

any ac voltage on the filament? likely the regulators are on the verge of dropping out of regulation. is your 230v about 3% low?

the unbal to bal part of the board had a problem with the current source, this was discussed over at diyaudio on the negative current source the 10k resistor went to the emitter. it needs to go to the base. first picture is the fix, second picture was the original with the error.

should be 12.6 vac rms measured blue to blue should be 8.5 vdc on that cap. should be 6.3 vdc after the regulators measured at the tube filaments. everything referenced to -350V

so you did not get the information you wanted over there, and now you come here. nice. from pictures elsewhere, the transformer winding is blue-black-blue and is 12.6v center tap, 2 diodes to capacitor full wave bridge. This is the filament supply and has high current ripple. The cap you have chosen is completely wrong for the job. All the original caps were chosen for a reason, if they are not dead then you should not be replacing them. And you should be careful touching that cap with the power on, as its sitting at -350v.

smallest 550v rated electrolytic is 47uf and still is 1 inch diameter.

capacitance multipliers don't work for low voltages and high currents. unless you do silly like denafrips. Which takes up huge amounts of space. And then one bad cap and you will wish you were dead trying to find it.

a pile of metrology gear owned by the university. not sure what the bandwidth was, but probably only a few khz at best. front end gear was peltier cooled to get the noise down. I really want the 8.5 digit agilent unit, but $5k for a toy with known issues seems a bit much especially since the screen room part is not possible at home.

5k is correct value

one other person over there had the same issue, turned out to be a miswired 4 pin jack

paid

actually all I have is the original above. this thing is a real fucked up mess designed by an idiot on lots of really bad drugs. using the tube filament as a load for the output transistor is all sorts of fucked up.

if you are using a .5 inch heatsink bracket, you have to use the other hole because the part is too big.

picture attached

so the 700s power supply voltages are exactly the same as the 717 and 727 from the transformer yellow,green,yellow go to full wave bridge rectified and regulated to +/-15 blue black blue (the filament supplies on the 700t) rectified and added to Vh- to make the - boost supply white red and brown orange are the 2 high voltage windings full wave rectified into CRC making +/-350 output boards identical to the t8000 each board has 3 x 2sc6127 and 1 resistor vas boards have 4 x 2sc6127 (2 x Darlington pair) and 4 x 150k/2W current source (on main board) vas is driven from dual low voltage n-channel jfet. (red heat shrink) (previously dual npn transistors) front end identical to 717/727 with dual n-channel jfet so about 99% identical to srm717 with non-obsolete parts and change of vas stage driver from bipolar to jfet on the D50 the front end jfet is known to be sourced by linear systems. So likely the k170 are also sourced by linear systems. so the 700T absolutely identical in every way to the srm-007t (and srm-t1 etc) with 2 x 6sn7 replacing 4 x 6cg7 plenty of room for replacement constant current sources to replace the mills load resistors both are priced the same, between $2895 and $2995

just because its a great amplifier does not mean that it likes to drive pure transformers. decca ribbon tweeters, quad esl56 and 63, klh 9 etc are all considered very hard to drive loads. transformers for electrostatics are much harder to wind than you might think, not a standard thing

Not the diodes

With those diodes is there any unregulated voltage on the first big cap?

Yes the opto just provides the on/off switch you need to use a timer circuit, or a seperate switch and a rolex

Ok, the stax mafia has spent quite a bit of time with the kse1500 first the headphones are 200v bias. The kingsound must have been modified to lower bias voltage otherwise these headphones will eventually pop. yes the kse1500/kse1200 is an extremely low power classAB amplifier with significant distortion. But great battery life. The kingsound is a chip amp driving transformers. Might actually sound better, but definitely not stax mafia approved. A really good battery powered portable electrostatic amp is harder than you might think. The stax d10 is another example that has issues.

its not a soft start. its a delayed high voltage turn on. For tubes that don't like voltage before the filament. no reason to use it except for tube or tube hybrid

ok, this scares the shit out of me. hennyo level of quality here. Second pic, the power supply in the upper right is an ebay high voltage switcher that's $7 I know its $7 because I bought some for the up and coming stax mafia portable, and its actually a pretty decent supply but can handle only about 5 watts of output power. You can see two more under the fan. Fan is required because those things get hot especially when you try to push them over 8 watts. And according to people at diyaudio that use them for tube preamps, the main diodes blow up at 10 watts. No way is this a kgsshv-carbon in terms of power. Barely 10% if that. Also limited to about 400v A pair of switchers on either side of the pot drive the HV supplies. So Switchers driving switchers. Not the most quiet of things. And the rest of it, yikes power supply pics here

Nope, you need the common ground.