Best Place To Cut Bass: Input Stage Or After PI Couplers?

[rp]

Title says it all. Want to try .022 instead of .047 in a 5C8. Any need to do all four?

[sluckey]

This kind of tweaking is very subjective. Let your ears be your guide.

[TUBEDUDE]

Reducing low end early on means more headroom in later stages.

[strelok]

For me I always start with the cathode bypass cap on the first stage. Granted, when I'm trying to cut bass response its generally because I'm trying to reduce tubby or farty sounding distortion at higher overdrive settings due to blocking distortion or the like. This is the most effective spot to try first in my experience, then if its still to much and I've gotten the value down to something small like .68 or so, I'll move to the coupling cap of that stage, then to the next cathode bypass etc etc.

[pdf64]

With realistic control settings, as signal level increases from 0, the power tube control grids will be the 1st point to clip and as it reduces, the last to stop clipping.
So depending on how you want to define headroom, it's probably the place where headroom is lowest.
That makes them prone to bias shift / blocking distortion under overdrive.
Reducing the coupling cap / grid referencing resistor time constant there is an effective strategy to mitigate the issue; as is adding grid stoppers to designs that don't have them, such as the 5C8 http://bmamps.com/Schematics/fender/twin_5c8_schem.pdf

[Firestorm]

Coupling cap into the the PI. Unless you need to fix something else, too.

Not strictly pertinent, but in my SEs, I wanted a Tweed Champ with a 12". Cap there is (stock) .02 so I put a .001 and a .02 with 1M pot between. Rolls off the lows when you want. In a P-P, I would guess a range of 200pFs to 1n.

[strelok]

That's a good point about output/PI coupling caps and blocking distortion. I generally don't think about the output coupling caps as I tend to stray away from anything larger than about .022uf instinctively for that very reason, but that's also dictated in part by the type of tone I'm generally after.

That being said in my own personal experience if you're trying to reduce the bass response in the of the pre-amp then its best to start with the first bypass cap followed by the coupling cap second bypass cap etc. My reasoning behind this is that if you cut bass response in a later part of the pre-amp when there is still an excess of bass let through in a previous stage, when that stage distorts the harmonics of that excess bass may still get let through and doesn't sound all that great IMO.

Of course this all assumes that the pre-amp has relatively reasonable RC values throughout. Recently I did some work on a friends Mesa Lonestar. One of the issues was a scratchy master volume pot. Cleaning it didn't help. It also only scratched for a few seconds after the standby was switched on. I put a voltmeter on it and sure enough there was some DC voltage on it when the switch was flipped. I thought the cap might be leaky until I looked at the schematic which reveled a .68uf coupling cap feeding the 1meg master volume. The rolloff frequency of that network comes out to something like .2hz! No wonder there was dc voltage on the pot when the standby was flipped. Replaced it with a .047uf that I had laying around and it went away. I can't imagine why they thought it was necessairy to use such a large value there. As you might guess changing it out to something more reasonable didn't change the tone of the amp much, but then again it is a Mesa.

[Littlewyan]

How does having a large coupling cap allow DC through? Surely it should block DC no matter how big it is.

Also one amp that goes against reducing bass early on is the Trainwreck Express. When I simulate that amp the bass isn't cut until it gets to the voltage divider between the 2nd and 3rd stage. Then again the first and second stages stay clean so I guess it works.

[pdf64]

Note that it was only scratchy for a short while after being energised.
That would pull the cap up to the Vp, and the long time constant to 0V means it would take a long time for the volume control side of the cap to get pulled down to the 0V reference.
So its just a momentary source of dc.
Yet another issue caused by switching the HT that is the nonsense of standby.

[Structo]

On my ODS amp, I started with the preamp but still had too much bass.

So I reduced the coupling caps on the PI output.

They were .1 uF and I reduced them to .05 uF. (ala #183)

That worked great!

[strelok]

Yup it sees the sudden spike in voltage as a "AC" voltage of sorts. All caps do this. Its just that this one happened to be big enough to make it noticeable. The voltage across the cap at switch on decays exponentially and is approximated pretty accurately by the formula a*e^(-t/k) where a is the voltage at switch on and k is related to the time constant of the filter network.

You guys are proving what I meant to add is my previous post. There's no one universal way of doing it. You have to take it on a case by case basis and do some experimenting. These are just a few ideas I'm spit balling that have worked for me on a semi-consistent basis. One thing I've found that seems to be relatively consistent is you *probably* don't want larger RC values proceeding smaller ones. For instance if you have a gain stage with a given frequency roll off given by the cathode resistor/cap and coupling cap/grid leak to the next stage. Then you probably don't want the stage prior to have a roll off given by those same components to have a lower roll-off. Probably.