Choke input vs Capacitor input power filtering

I have been wondering about this for awhile now, looking at the curves of rectifier tubes it is apparent that choke input offer more stable operation, voltage stays fairly constant despite current changes, but more voltage drop then a capacitor input. Where capacitor input filters sag with increases in current draw but have less voltage drop. So it seems that Capacitor input filters are fine for single ended amps where current draw is fairly constant and choke input for the other classes, assuming sag is not desired. But everynow and then I see a single ended schematic that uses choke input. Does choke input have other advantages, does it clean things up better or is this done for other reasons, to get some voltage drop on purpose or plain ol not knowing better. I am just about to start building a small practice amp not sure if I should go for a choke input or throw a cap between the choke and rectifier and use those extra volts since this is an SE amp and sag is not an issue.

adam

[quote author="adamasd"][...] Does choke input have other advantages [...][/quote]

Hi Adam,

a choke input keeps the current draw through the rectifier pretty constant overall, opposed to the short big current peaks you get with a C input, which keeps stress away from the rectifier tube. (The PSUD program at duncanamps.com is great to visualize these things. Some tube rectifiers [EDIT] will [/EDIT] die very fast when used with a C input.
A disadvantage is that the filter caps need to be rated for the full sqrt(2)*transformer voltage although the final voltage will be way lower, as the choke input only works with a minimum current flowing (= not with tubes pulled for testing or before the filaments are hot).



HTH
;Matthias

Edited to change bad wording, sorry...

Commercially, the real question is whether iron or capacitors are cheaper. Up to 1930, iron was cheaper. After 1940, caps were generally cheaper, but for heavy current a choke might allow use of a low-cost rectifier where a cap would need a fat rectifier.

Choke-input rises 50% at no-load, which means you need high surge ratings until the tubes warm up, and a Class B stage may need a very heavy bleeder.

I'm not sure that cap-input has less sag, but it depends a lot on the rectifier, iron, and cap parameters. Historically, if the minimum current were not too low, choke-input was the way to go for low sag.

It gets trickier. SE amps have no buzz cancellation, so need clean power. Two reactances give more buzz reduction than one, even if the one is quite large.

I think I got it now, at least well enough to experiment abit and see what the resaults are. If I did this last night I may have blown some caps. Glad I decided to look into it abit more first.

What is the purpose of a rectifier that is supposed to die very fast with a capacitor input?

Thanks,
adam

> What is the purpose of a rectifier that is supposed to die very fast with a capacitor input?

Darn english language.

He does not mean they are designed to die. But the rectifier datasheet has a hot-switching current rating, and implies that if you exceed this, something bad will happen (and yes, it could kill the tube, though maybe not too quick). The datasheet often gives a input capacitor value and warns that if you use more capacitance you will have to add resistance to stay inside the ratings.

(Yes, I guess this means they are "designed to die" in the sense that they could sell you a bigger tube at a higher price that would survive bigger caps, but they designed instead for a good price that can cover many realistic needs.)

GE 5V4 datasheet page 2 shows choke and cap input on the same chart, and shows choke-input sagging less.

AHHHH!

I just could not figure out what he ment by that, but it seems quite obvious now that an english to english translator came about. Maybe it is time for the board to give up english and switch to esperanto.

thanks,
adam