power supply filtering

[CraigB]

Hi, All!  The weekend's almost here.  Time to get working on another amp project  

See attached.  I don't think I've ever had the need to place e-lytic caps in series (to get the benefit of a higher voltage rating) and parallel (because I want to double the amount of filtering).  If what I'm after is a total of 100uF filtering with a voltage rating of 700v, I can't see any reason why this wouldn't be a fine way to do it.  Any disagreement with that or "why don't you do it this way instead?" is welcome.

I have the 100uF/350v radial caps on hand, they are relatively small/short, and will fit my board better than two 220uF/350v axials.
Thanks!  

Craig  

[kagliostro]

I think your "scheme" is correct and if you have those cap why don't use it ?

prefer other cap you already have, only if they are rated for a higher temperature

(700v rate must be read more as 550/600v)

---

I think you know that if the rectifier is a vacuum tube the first capacitor must be of a max value depending on the tube you use, no problems with SS rectify

K

[stratele52]

Good schematic  this is a total of 100 MFD . Could be too big for tube rectifier, ok for silicon rectifier

[HotBluePlates]

Also, 50uF is not much below 70uF. Old-school e-lytic tolerances were on the order of +100/-50%.

If you feel you need 100uF of main filter, your plan is perfectly acceptable.

If 50uF might work, you could save two caps. If that filtering is insufficient, a second series pair (two 100uF's for 50uF total) after a choke or small decoupling resistor might give you better filtering. Recall that the series impedance/resistance of the choke/resistor helps the cap better knock down ripple.

Then again, you might not want to spend for the choke, and may not wish to waste heat/power in the decoupling resistor.

[tubeswell]

And furthermore, you could replace that lower 220k bleeder with a ~1/5th voltage divider (say 180k and 39k in series), decoupled at the knee with a 10uF 100V cap, and use that point for heater elevation. (This should give you ~1/10th* of the B+ voltage at that point)

* actually slightly less that 1/10th - but its only geetar amp math :-)

[kagliostro]

Tubeswell

please can you explain better

decoupled at the knee with a 10uF 100V cap

or draw a little scheme

Thanks

K

[sluckey]

.

[kagliostro]

Oh, now I understand 

Thanks Sluckey

K

[tubeswell]

Sluckey beat me to it

[CraigB]

Thanks very much for all the responses!  Solid state rectifier -> 100uF -> VVR where only the power tubes and PI are regulated.  I've built this same amp before, only with a tube rectifier.  This one has a little different power transformer.

And furthermore, you could replace that lower 220k bleeder with a ~1/5th voltage divider (say 180k and 39k in series), decoupled at the knee with a 10uF 100V cap, and use that point for heater elevation. (This should give you ~1/10th* of the B+ voltage at that point)

* actually slightly less that 1/10th - but its only geetar amp math :-)

I like that idea.  Thanks!