PS Caps and Low Frequency Response

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

starfish

Well-known member
Joined
Aug 22, 2005
Messages
78
Location
Los Angeles
I've read that replacing old filter caps in tube guitar amps will improve the low frequency response. Is this also true in tube mic pre's and tube mics where the current draw is much lower, or are these usually recapped because of noise issues? Thanks
 
No, no and No, in that order.

The car audio community has progagated that myth out of a partial understanding and a whole lot of ignorance.

Filter caps will reduce hum from power supply ripple. That's what they do. Low value caps will also cause average power rail sag at high power, but that's a loss of headroom (before hum-modulated clipping onset) and nothing to do with LF extension.

Noise issues? -Not usually associated with caps. -Every time a musician brings an amplifier to me saying "it's noisy, it needs recapping" it's never been the caps. Signal caps can be responsible for Low frequency rolloff, but not -as you mention- filter caps.

Keith
 
With a guitar amp the output tube part is the section that you can sometimes hear the difference with caps. It seem to be the ESR part of the cap I use caps that have<= 1.5 ohms per 20uf at 450/500V for the power amp section.

If you add too much cap value or too low an ESR cap you can have problems with some amp power supplys. Sometime adding a low value resistor to limit the max current surges helps with very low ESR caps like polypros.
 
[quote author="Gus"]With a guitar amp the output tube part is the section that you can sometimes hear the difference with caps.[/quote]

No doubt.

...but extending low frequency response?

No.

Keith
 
but if you have power sag because of small cap value the first thing to suffer is low frequencies (bass) because (assuming full-bandwidth program material) they are usually loudest and thus require most power. Am I making any sense here?
 
Thanks for the answers. I guess I'm trying to find out if and when to replace old filter caps. Is increased hum the main issue? Does capacitance decrease as they get old? If there is no hum should I leave them be- even if they are 30-40 years old?
 
I have measured some old caps in guitar amps still measured good after >20 years for value,DA,ESR and leakage at the RATED voltage.

There are some crap Als out there.

I would not let someone work on an amp if they did not have access to a good cap tester.
 
[quote author="Family Hoof"]but if you have power sag because of small cap value the first thing to suffer is low frequencies (bass) because (assuming full-bandwidth program material) they are usually loudest and thus require most power. Am I making any sense here?[/quote]
Yes, but the argument is a fallacy.

The low frequencies are present in highest proportion (size) so they trigger the clipping as the power rails sag, but if you are sweeping a tone at a constant amplitude, there is absolutely NO difference in amplitude across the band, therefore ALL frequencies trigger it when they are present in the same amplitudes.

There is NO LF extension due to PS caps. Misunderstood by many.

Consider DC like in car audio. There's no 120Hz (or 100Hz in Europe) recharge pulse... it's all pure DC and resistive. Now... how does a cap assist beyond acting like a bit of a flywheel? At equal amplitude, whether DC, 100Hz or 120HZ recharging, ALL frequencies are equal.

Keith
 
[quote author="SSLtech"]

There is NO LF extension due to PS caps. Misunderstood by many.
[/quote]

I'm hesitating to add some theoretical crap about the equivalent AC-schematic for an amplifier stage. The supply-cap IS in the signal path then... well, put in another way, you may ground the top side of a collector-resistor in the equiv. schematic since the supply cap is assumed to be sufficiently large - equal to the power-supply having zero source impedance for all frequencies.

If that supply-cap would be of a smaller value in a common emitter-amplification stage then one could even say that the collectorload becomes higher, so more gain :roll:

As I wrote, I was hesitating... now I'm hesitating whether or not to erase the stuff above - since likely those decreased cap values won't enter this territory here too soon I assume (but too lazy too think a bit further a bit now).
 
> The supply-cap IS in the signal path then...

Right, and a woefully weak rail cap would (does) affect low frequency response.

But in almost all cases, the size of cap needed for good buzz-reduction is way more than is needed to meet your LF goal. Especially in a gitar amp where nothing below 82Hz matters, and speaker efficiency is high through the buzz range 120Hz-600Hz.

Also, negative feedback around the amp should make it fairly insensitive to rail impedance (though perhaps more prone to motorboating if the rail gets too loose).

> in tube mic pre's and tube mics

Rail cap value should be much more than the value of the coupling caps in that stage. For most low-level work, that can mean a few uFd. For power stages maybe a dozen uFd. But buzz-reduction and decoupling considerations usually force the rail caps to be bigger than this. So the first sign of trouble is usually buzz or motorboating, not loss of bass.
 
Rail cap value should be much more than the value of the coupling caps in that stage. For most low-level work, that can mean a few uFd. For power stages maybe a dozen uFd. But buzz-reduction and decoupling considerations usually force the rail caps to be bigger than this. So the first sign of trouble is usually buzz or motorboating, not loss of bass.
Thank you for clarifying this.
 
Back
Top