PreAmp Cathode bias Capacitor

[big_teee]

Trying to understand the effect of the Preamp Cathode Cap.
Marshall 1987 Circuit puts a .68uf Coupling cap in parallel with the 820 ohm resistor from the Cathode to ground on V2a.
What does the Cap do, and what effect do you get when you change the Cap Smaller or large in uf Value?
The Marshall 2204 puts the .68uf Cap on the Cathode of V1a, none on the 820 resistor of V2a.
Thanks for any explanation of the bias Cap on the cathode.
Thanks,
Terry

[Chuck H]

The cap isn't a bias cap. It's typically called a bypass cap. And that's what it does. It allows AC to bypass resistance in the cathode circuit. This eliminates local NFB and increases the tubes gain above the capacitors knee frequency. With the typical 1.5k preamp cathode resistor the increase is about 6db. So, the bigger the cap, the deeper the frequency that is bypassed (and boosted).

The effect will be much less, but still significant, with the 820R cathode resistor. And since the cathode resistor basically IS the cathode circuit impedance as seen by that cap it's value will affect the knee frequency of any bypass capacitor since that changes with impedance.

Of course there are also circuits that don't use a bypass cap. Basically, you don't always need the extra gain, you need a more linear amplifier stage or there is some sort of AC interaction implemented at the cathode like a NFB circuit.

[big_teee]

Thanks Chuck for the Reply.
So the Capacitor allows the Highs to bypass the 820 ohm resistor?
Using a .22uf, .47uf, or.68uf, etc. The bigger the cap the deeper the frequency.
So the .22uf would be brighter than the .68uf, is that correct?
Like a tone Cap in a guitar?
Thanks again,
Terry

[Chuck H]

So the .22uf would be brighter than the .68uf, is that correct?

Not exactly. It's a little more accurate to say that the .22 cap would be less deep. Since both caps will pass the same high frequencies but the .68 will dip further down into the mids. I used to have a chart that showed the typical cap value frequency knee with a 1.5k cathode resistor, but I haven't uploaded it to my new poot yet. The knee for a 820R resistor will be higher for the same cap value. It hardly matters what the actual frequency is though since listening tests are the best way to choose the value. It's still a good chart to have. Perhaps someone else who has it uploaded will read this post.?.

Like a tone Cap in a guitar?

Kind of. More like a "presence" control. In a typical presence control the highs are removed from the circuit by bleeding them to ground so that they don't become part of the NFB signal and therefor don't get cancelled.

[Zozobra]

Not exactly. It's a little more accurate to say that the .22 cap would be less deep. Since both caps will pass the same high frequencies but the .68 will dip further down into the mids. I used to have a chart that showed the typical cap value frequency knee with a 1.5k cathode resistor, but I haven't uploaded it to my new poot yet. The knee for a 820R resistor will be higher for the same cap value. It hardly matters what the actual frequency is though since listening tests are the best way to choose the value. It's still a good chart to have. Perhaps someone else who has it uploaded will read this post.?.



Kind of. More like a "presence" control. In a typical presence control the highs are removed from the circuit by bleeding them to ground so that they don't become part of the NFB signal and therefor don't get cancelled.

Mercilessly stolen from Merlins website!

[defaced]

The text that goes above that graph is equally important. If you change the conditions, the graph also changes.

his graph shows how the frequency reponse of a typical ECC83 with 100k anode load and 1.5k cathode resistor, for different values of cathode bypass capcitor.

The Valve Wizard -Cathode Follower

[Zozobra]

Yeah I probably should have included that. I do wonder how many people bother to do the maths once they've got the concept nailed.

[Chuck H]

I think it's also worth noting that in this particular application the -3db point doesn't SOUND like the caps knee. Maybe because there is only 6db total increase. For example, a 2.2uf bypass cap with a 1.5k cathode resistor doesn't seem to boost bass frequencies much. It sounds more like a mid bump. And there is a clear audible different between 2.2uf and 10uf which, looking at the graph, you wouldn't expect to hear in a guitar amp. The best way to "get it" is to just experiment with different values and hear it for yourself.

[big_teee]

I think it's also worth noting that in this particular application the -3db point doesn't SOUND like the caps knee. Maybe because there is only 6db total increase. For example, a 2.2uf bypass cap with a 1.5k cathode resistor doesn't seem to boost bass frequencies much. It sounds more like a mid bump. And there is a clear audible different between 2.2uf and 10uf which, looking at the graph, you wouldn't expect to hear in a guitar amp. The best way to "get it" is to just experiment with different values and hear it for yourself.

I agree with testing.
The .1uf was too bright but did have some gain.
So I was trying to figure out what Caps to purchase to try.
Obviously I need the .68uf, Should I also try the 2.2, and larger.
That's what I love about my turret board Amp, Experimenting is so easy.
I'm hooked on the Hand Built Amps.
Terry

[defaced]

I do wonder how many people bother to do the maths once they've got the concept nailed.

Hey now! That's me you're describing

I can do the math, but when it comes down to it, I'm often in the same camp as Chuck: try it and see for yourself.

[Tom Phillips]

…So I was trying to figure out what Caps to purchase to try.
Obviously I need the .68uf, Should I also try the 2.2, and larger….

Since you are into experimenting you can test a whole range of cathode bypass cap values. Start with no cap and try anything you desire up to 25µF or even more. You will be able to hear the effect of the different values for yourself and experience how the sound changes even when the cap is smaller or larger than your final preferred value. Keep in mind that the final sound is also dependent on the circuits that are downline from the input stage. The point being that a 0.68µF cap in one amp won’t necessarily produce the same sound in another amp.

Remember that you can make cap values that you don’t have in stock by hooking up parallel and series combinations of the various caps that you have available. Use them for temporary sound tests and then order the final value later.

Regards,
Tom

[Chuck H]

The point being that a 0.68µF cap in one amp won’t necessarily produce the same sound in another amp.

And a change in bias resistor value will change the knee frequency. And you'll get more effect in the first gain stage than the one right behind the PI. And ...

Best to just experiment.

You'll hear the biggest difference between no cap and 4.7uf (with a 1.5k bias resistor). After that you start getting to frequencies below the average guitar speakers range. Bigger values will still affect harmonic reactions within the amp itself. This is significant and still changes the dynamics and feel, but not so much the audible frequencies from the speaker.

Values smaller than .1uf (with a 1.5k bias resistor) are really only boosting frequencies above the average guitar speakers range. This really only increases some hiss and glassy trebles. So...

.1uf, .15uf .22, .33, .47, .68, 1uf, 1.5uf, 2.2uf, 3.3uf, 4.7uf, 22uf, and huge, like 100uf or 220uf. I don't hear any significant difference between 4.7uf and 6.8uf or 10uf. Likewise between 15uf and 22uf (6.8uf, 10uf and 15uf) so I'd skip those. Same with values like 33uf, 47uf and 68uf. Again, this is as it applies to a 1.5k bias resistor. With an 820R you should double the stated values, or near as possible with standard values. All this is just IMHE. Milage may vary.

[jrfrond]

As a tangent to what was just said, it's pretty clear to see that, in most cases, the typical 22uF bypass cap is just way too high in value to be of any value in a guitar amp's first stage. On the contrary, it can actually cause issues with "rolling" bass frequencies in high-gain amps, where the low note start sounding loose and farty instead of tight. This is why Marshall uses the .68uF in the first stage of their amps, and there's still plenty of low end left. In conjunction with tight filtering and stack cabinets, you get that bottom-end "whoomp" with an almost lick drum-like attack that Marshalls are famous for.

Just FYI- the Fender Hot Rod Deluxe switches in a small-value cathode cap as a Bright switch, forgoing the usual scheme of a cap across the volume control. In this manner, the Bright switch works even with the volume dimed.

[loudthud]

The .68uF in the cathode creates a low frequency shelf, not a pure rolloff. I always liked the sound of a fully bypassed cathode as a first stage. If you need a shelf, you can add a parallel cap/resistor in series with the coupling cap to the next stage. YMMV and will probably be less...

[Zozobra]

In my next build I'm going to try out LED biasing for the input stage so I can have effectively full bypass without resorting to using a huge electrolytic.