using an inductor and capacitor as a low pass filter at the 12ax7 cathode

[anson]

Hi everybody,

I know how commonly used a bypass capacitor is used arcoss a cathode resistor in a 12ax7 amplifier stage to boost the highs, at a frequency depending on the capacitor/cathode relationship. What I want to do is find a way to boost the lows instead of the highs by allowing ac bass frequencies to bypass the cathode resistor instead of ac high frequencies. Does anybody know how I could do this? I THINK that what I would do would use a vary large bypass cap to allow a boost to all frequencies (including bass), and then put an inductor in series with that cap to limit the amount of highs that can be boosted. would this work? And I was curious what values of inductor would be effective in passing audio-band bass frequencies.

I am basically basing this idea on my assumption that an inductor can be used as a low-pass filter.

Thanks for your thoughts!
Anson

[Enzo]

I think you have it backwards. A cathode bypass cap increases the low end response. Adds gain too. Addign gain might allow more high freqs to ultimately come out the amp due to overdrives and what not, but that is separate issue from the stage itself having high end response.

To make a stage more bass heavy and treble light, consider this. Gain is cheap, and there is usually more than you need. SO just roll off some highs, and the lows will be what's left. Turn it up a hair to compensate. The tone stack in your typical Fender or Marshall are subtractive, there is no boost. Just like them.

In the typical 12AX7 stage, a plain old vanilla voltage amp stage, the cathode resistor sets the bias for the tube. Signal applied at the grid will modulate the current through the tube. That tube current also flows through the plate load resistor and the cathode resistor. Current through a resistor causes a voltage drop across the resistor - that's Ohm's Law. So the signal at the grid also appears at the cathode. That means the cathode voltage varies with the signal, and thus the tube bias varies with the signal.

By adding a cap in parallel with the cathode resistor, the cathode voltage remains more stable, the bias remains more stable. Instead of the bias rising as the signal does, the bias stays put allowing the signal to be amplified more in the tube.

Just like any filter cap, this cap charges and discharges with the signal waveform. The smaller the cap, the faster it can charge or discharge. Low frequency signals take longer to rise and fall. If the cap is small enough to keep up with the charge and discharge demanded by the signal, then it won't be more stable than the signal. SO the larger the cap, the lower the freq will be that it can affect.

It can also be explained in terms of the cathode providing negative feedback, and I can follow those explanations, but it doesn't intuit on me as well.

[Dave H]

I think you have it backwards. A cathode bypass cap increases the low end response.

Are you sure about that?

[pvsage]

A cathode bypass cap boosts gain above a certain frequency, depending on the cathode bias resistor and the tube. The smaller the cap, the higher this "shelf" frequency. Most traditional/vintage schematics call for a cathode bypass cap that sets this shelf at 10Hz or lower with a 12AX7; more modern designs limit bass response a bit to help control blocking distortion ("farting out") or to give a brighter lead sound.

For ideas on cutting treble response, check out this page.

[daz]

Maybe i missed something because Enzo is the last person who's judgment i'd ever question. But i always though smaller caps would give you more highs. it would seem they'd drain frequencies to ground since they are connected to ground, but it doesn't seem to work that way. For one thing they add gain which means that are ALLOWING whatever freqs their value allows to be added to a dgree decided on by the resistor. In practice this works for me, so i can't understand what you're saying Enzo. i must be missing something because since it's you that said this i trust that somehow i am off here. But as far as the actual result's i have to tell you that using a say 10uf cap gives me a much woofier sound than say a .68. Right now i have a .68 on the first cathode and a .47 on the second and the fuzzy muddy low end it had is non existant. But if i put a much larger cap on either one the amp becomes a mud fest with lows that are fuzzing and indistinct. The way i've been looking at it is if you use a small cap, it only allows higher freqs. And then the more you raise the R value the gain of the full range of freqs is redcuced while whatever freqs the cap allows stay the same. so the more you reduce the R the more highs there are in relation to the full spectrum. i tried a 10k on V1b once ala 2204, but the gain was reduced too much and the tone was dark and lifeless. i then tried that again later on, but this time i bypassed it with a .47. the result was glorious. The lows because tight and more full of life and all muddiness was 100% gone. the entire amp sounded worlds better. If i try adding a bigger cap it becomes dark and muddy again.

anyway enzo, explain this to me in layman's terms if you will, because i want to understand what i'm missing, and i mean absolutely no sarcasm or disrespect...i really want to know.

[jrfrond]

The addition of the cathode bypass cap also lowers the output impedance of the stage above the rolloff point, in addition to canceling the effects of degenerative feedback, i.e. lower gain. This is something to keep in mind if you want to drive a tone stack from the plate rather than using an extra stage as a buffer (cathode follower).

Audio frequency inductors are too expensive, too big, and too prone to hum to incorporate them for high-frequency rolloff. This is why you really do not see them much in guitar amps. For me, the best way to roll off highs is to parallel a cap across the plate resistor of the stage, anywhere between 100-.01 has been commonly used. It must be rated for the plate voltage. Doing it this way minimizes the loading when a shunt cap is used from the signal path to ground. It can also be effectively used to stabilize high-gain stage which may be prone to oscillation. In this case, a small value (47pF or thereabouts) is used, and it is referred to as a "snubber". These plate snubbers can also be trimmed with a series resistor.

[pvsage]

I've seen "stabilizer" caps in a couple of different places in preamp stages - from plate to ground, from plate to cathode...the last is used in the first gain stage in a lot of Marshall's JCM-series...

If I'm not mistaken, these interact with the triode's output impedance to form a simple low-pass filter.

[booj]

At lower frequency's the impedance of the inductor will increase and the gain will decrease. Not what you're looking for. It may work in a negative feedback loop, but that would take some experimenting.

[Enzo]

I was describing things as informally as possible. Guess I did a poor job.

A smaller cathode bypass cap will add gain to the higher freqs only, yes. Increasing the cap size doesn't add highs, it widens the bandwidth of that gain. SO smaller caps accentuate the highs by not accentuating the lows as much. The larger the cap the more lows it adds to that accentuating party.

daz, you explain that a larger cap adds bottom response to make things a mud fest. I thought that was the point I was making in post #2. And is it not exactly what anson wanted to do in the first post? Add bottom? perhaps my post was not stated very clearly. He wanted to use a large cap (good for more low end) then somehow defeat it with a choke. COunteracting the effect of the larger cap is the opposite of what he wants to do. I don't disagree with any of your findings, they are as I would expect. A 10uf will be bassier than a .68.

Smaller caps yield more highs by yielding less lows.


Listening to the responses, I think the issue is not that folks don;t know what larger and smaller caps do, I think the trouble is they don't understand why it happens. Or at the very least, they intuit it way different that I do.

[anson]

Thanks Pvasage, that link was real helpfull.

What i really wanted to accomplish with my amps was to kill the bass early on so that it overdrives nicely and doesn't become a swamp of mud, and then boost it in subsequent stages so that it still has balls in the end. But I wanted the bass boost to be done in a shelving way, so that the highs are still there but just reduced in volume. I didn't want to just use a low-pass filter because by the time the bass frequencies I want are effectively boosted, the sparkly highs are all cut off.

On older amps, like sunns, hiwatts, and marshalls, you see massive 330 u or 220 u (or something on that order) bypass caps to boost the frequency response of that stage all the way from super high frequencies to subsonic frequencies. But lets say i didn't want to boost the highs after all, then I'd have to engineer some way to only allow the bass frequencies to bypass the cathode resistor. And that's why I considered putting an inductor in series with the bypass cap to allow only the low frequencies to be boosted as a result of the removal of negative feedback for those frequencies. I can't just (i don't thank, anyway) put an inductor in parallel with the cathode resistor because that would allow DC current through and that would f**K up the bias of the 12ax7 stage under discussion. That's why i considered putting the inductor in series with a capacitor, to keep DC from getting to the cathode therby messing up the bias.

But from what jrfrond says, it might be impractical to use inductors for this purpose. I do remember, however, remember seeing inductors on the global NFB loops of early oranges and some sunns, so that's what prompted me to consider an inductor as a real tool.

A bit off subject, but would using a cathode follower be an effective way to keep the bass frequencies with out causing ugly overdrive of the next stage?
Enzo's basically got it right, I just want to have bottom, but I'd like it to be as non "foggy" or "swampy" or "crappy" as possible.
Anson

[Enzo]

OK, then the cathode is not the place to do this, use the idea someone posted earlier - bypass some highs around the plate load resistor - then they wont get amplifed and passed on.

ANd you can also use simple rolloff later, such as a cap across teh volume pot.

Here is a simple experiment to do. Pick a stage, the input stage most likely, of an amp, find the plate load resistor of the stage - most often 100k - now using a couple clip wires, tack various caps in prarllel with that resistor and see the effect... well hear the effect rather. use caps of high enough voltage, same type you'd use to pass signal on to the next stage. Try smaller caps like 100pf and 270pf, and try laerger caps like .01. Just to get an idea of what it does.

It is easy to pass bass through an amp, it is a lot harder to "add it back in" later in the circuit.

[Joey Voltage]

It depends on what you are really trying to do. If you are either A) trying to achieve a smoother lead tone, or B) trying to add some low end wallop for a heavier sound, there are better ways to do either, or both.

FWIW a larger CK won't necessarily make a mud fest, you have to take other things into consideration as well. The coupling cap of the stage, in conjunction with the load of the next stage will also dictate how much amplitude these frequencies will have going into the proceding stage. Plenty of modern amplifiers use larger CK values, and then attenute the unwanted frequency range with high pass filters. This may add some harmonic complexity.

[anson]

Yeah, It's a heavier sound that I'm looking for
.

[Joey Voltage]

As another poster suggested, you may want to try some sort of active boost for this. The most common place to do this is the global -fb loop in the PA. an inductor will work here, or you can use another gain stage for this,.......but if you have an FX loop, you might as well just stick in an external e.q. there if you have one.

[Joey Voltage]

Yeah, It's a heavier sound that I'm looking for
.

If you take a look at the topology of many heavy sounding amps, you will see that they DO NOT try to sculpt a bass heavy tone early on in the preamp, especially the first stage. The first stage is most important in this respect, as the tone shaping that occurs here gets amplified by every succeding stage. If you study enough of them, you start to see how similar they can be, and how formulaic the designs are.