Decoupling Paper In Oil capacitor question for KT88 design.

[Layberinthius]

How, I've been searching the galaxy trying to find out this quite simple question.

http://www.plitron.com/pages/Product...o/vtvkt88.html

Here is the axis of my evil.^^

And my weapon is six 8uF 660v "60Hz" Paper In Oil capacitors.
(For a stereo setup)

Now the design calls for 10uF decoupling caps in the audio stages
but with constraints on leather-wallet skin I have been 'forced' to use 8uF caps as replacements, If the 8uF caps work at the plate voltage 515 volts and are "Sprague Eccol" brand will I see any
negative effects?

I'm thinking in the region of lower tolerance to distortion in the stages?.

And I am of course considering scraping up some axial 600v 2uF
caps for in parallel ops.

Help me any way you can :P

Cheers.

[Sch3mat1c]

Your link is broken so I have to guess at this... Simple answer, use lytics. But since that ain' gonna happen, I'm going to guess it's not going to care +/-20%.

Tim

[fdegrove]

Hi,

But since that ain' gonna happen, I'm going to guess it's not going to care +/-20%.

Or bigger iron...

Cheers,

[Layberinthius]

http://www.plitron.com/pages/Products/Audio/vtvkt88.htm

Try this one.

[Sch3mat1c]

I don't agree with their NFB method (no tuning whatsoever), nor the small size of the bypasses, but sure, 20% isn't going to make a difference.

Tim

[Layberinthius]

How would I go at merging two of the power supply schematics onto one chassis?

I was considering having one large merged power supply 'umbilical coord' to go to the amp stage and also merging two channels into one chassis/ground-plane but having neg DC bias and B+ seperate coming from the PSU chassis into the circuits INSTEAD of binding them in parallel??

The PSU ground plane would be the same for either side too.

I'm definatley not keen on doubling most component values and making a behemoth of a PSU even tho it may be better for sound quality...or so i've heared.

The PSU chassis would be seperate in a steel painted cube cabinet with forced-venting and a dust collector seeing it will be on the ground next to my setup on the carpet but on rubber-feet 2/3 an inch off.

http://www.plitron.com/images/Diagrams/vtvkt882a.gif

[Layberinthius]

Hmm I can't even understand myself right now

[Sch3mat1c]

Looks like you'll need a PT (and so on) with double current ratings... you'll want to double capacitances in the supply as well. You can then just build two amp circuits and run them off the same supply. No problem building them in the same chassis, in fact then you can connect common things together (like the 10uF oil cap chains which filter the low level stages... again you'll want double capacitance and half resistance).

Tim

[Layberinthius]

But but but

Well I could place a pair of the standard trannys in parallel and double the component ratings, like caps, double them accordingly in the same package, that would variably half the size needed for some post-transformer components.

Of course I could still use the 1200v diodes.

I was considering placing most of the not-heavy components on PCB and cutting a fan mount underneath or above the diodes to keep them from burning the pcb and extending their life (not that they blow too often but I don't like heat buildup) although this may cause dust to form I've considered that and am going to place a filter upon the intake which would invariably half the fan performance, anyway.

I'm worried to death at the chokes and wirewounds.

How do chokes go when placed close together? does shielding take care of most magnetic play in such a high-powered job?
and more importantly what could be the effect upon resistance when placed in parallel? Would I have any problems with voltage?

What kind of effects could I see after the chokes? when they're parallel?

On to the caps and I was considering to make size a matter with them and just use film.

Cheers.

[Sch3mat1c]

Actually it's not a good idea to parallel trannies! Small differences in output voltage cause currents equal to the difference flowing through the DCR of the windings. Can't you find one spec'd at double current on all windings? If you insist on paralleling, you might as well grab doubles of a few more parts and make a pair of monoblocks.

Inductors placed in parallel: DCR and inductance drop in half (assuming equal; if not, works like any other parallel resistance: R = R1*R2 / (R1+R2) ), parasitic shunt capacitance also doubles (not an issue with PS chokes). The choke shown is rated 400mA which should be fine for both channels' load.

You can calculate power dissipation in the resistors: take the 4.7k 5W for instance. Schematic shows 480V on one side, 525 on the other, a difference of 45V. P = V^2/R, or less than .5W. You could use a 1W resistor with absolutely no worries here, even in a cramped chassis. Speaking of resistors, carbon comp are notorious for drift and noise, you should use at least a modern carbon film type.

Hmm...WTF!?!? That schematic is utterly WRONG. You cannot have 50V across the 27k (which feeds the input stage), because that's 1.8mA. Low for a 6SN7 but that's not the beef. The 1.8mA also flows through the 47k plate resistor, developing 87V. This puts the plate at 233V, an utter impossibility for the following cathodyne splitter stage, and even if it were possible, it would be a bad operating point for the first stage anyway. (Better to have Vplate around 1/3 to 1/2 Vsupply. When direct driving a splitter like this, ideal is 1/3 to 1/4 Vsupply.) Be careful with this design, don't expect it to measure as notated.

Tim