What is the reason of having a cathode follower in Marshall type preamp circuit?

[Alan0354]

What is the advantage of having the cathode follower? I experiment with and without the CF and I notice the sound is softer without it. Which is not necessary a bad thing. I speculate that CF present a low output impedance which make the sound brighter ( not affected by any capacitance loading). What else?

[potatofarmer]

For make good distortion.

http://www.valvewizard.co.uk/dccf.html

[loudthud]

The tone stack has a fairly low input impedance at high frequencies. Especially when the slope resistor is lowered to 33K. Without the follower, the tone stack sounds kind of dark. If you can correct the response at other places in the signal chain, you don't need the follower. The follower also adds it own distortion to large signals. It clips softly on positive peaks and fast negative slopes are limited by the 100K pull down.

[Alan0354]

For make good distortion.

The Valve Wizard

I hate to challenge the article again. I read through the article and I have to partially disagree. This time is not by theory, but by a few days of observations on the scope:

The article only talked about the grid current of the follower prevent the anode of the gain stage from going more positive and thereby create a flat top. That I agree. But the next part talk about the lower portion can go down unclipped and follower the anode of the gain stage. That I do not agree in real observation with the loading of the tone stack and the lower limit of the gain triode.

If you look at the example of +B=300V and the anode of the gain stage at 200 and Rk=100K. The anode can never goes below 120V. So it clips at the bottom also. I tried changing Rk higher and lower. If going to 68K, the wave form just being pull down, like it clipped lower on the high side but extend the clipping on the low side lower. The low side also limited by the current the Rk can sink before it runs out of steam like Loud thud said.

Bottom line, the clipping almost look symmetrical no matter how I played with the Rk. You hardly get a lot of even harmonics. And base on sound test, it is very harsh if the +B is about 350V!!! Nothing warm about it. . I lower the +B to 300V and it sounded a little softer. But then the swing become so limited.

[Chuck H]

Are you sure that you're not clipping the stage prior to the CF and reading THAT wave form as processed by the CF??? This is paramount info. You can't analyze how a CF clips if you put a clipped wave into it.

[pdf64]

That's a tricky point, because a dc coupled cathode follower is a dual triode arrangement.
It wouldn't make sense to ignore or eliminate the preceding gain stage from the analysis, as without it the cathode follower probably wouldn't operate in the same way.
Pete

[Chuck H]

Well I don't mean to eliminate it. But scope the signal at that preceding stage to see if whatever clipping is being attributed to CF may actually be what's going into it rather than what the CF itself is doing.

[Alan0354]

Yes, I scoped and reduce the drive so the preceding stage are all lineal......well in distort a little, but no where close to clipping. It basically looks like a sine wave. Of cause, if the preceding stage is clipping, all bets are off.

I am not saying the output is symmetrical, but just not very asymmetrical that produce a lot of even harmonics. No matter how I change the Rk or changing the +B, it basically moving the DC offset and slightly changing the amplitude. It does change the symmetry a little, but really nothing to write home. You can change the symmetry more by playing with the normal common cathode stage by adjusting the Rk and plate resistor to clipper fast on the top or bottom of the sine wave.

[jazbo8]

You meant you are not able to duplicate the clipping characteristics as shown on Merlin's page with the TS connected at the cathode? If so, could you please post the scope shots.

[Alan0354]

You meant you are not able to duplicate the clipping characteristics as shown on Merlin's page with the TS connected at the cathode? If so, could you please post the scope shots.

I was referring to Valve Wizard, not Merlin. It depends on the Rk used, you can make the lower portion clip first if you increase Rk and the pull down runs out of steam. You lower the Rk, you make the top clip first. But very soon, the other side clips no matter what setting. You only have a very small range that you get one side clipped and the other side still like sine wave. So if people read the article thinking this is the way to create a lot of even harmonics, that is only true in a very small range.

[g1]

I was referring to Valve Wizard, not Merlin.

Surprise! Same person .

[Chuck H]

Sweet!

[Chuck H]

It doesn't take a lot of asymmetry to notice a bump in the 2nd harmonic. IIRC another thing the CF does is cut the corners a bit on sharp clipped edges. So that would lessen some very high order harmonics? This might make the 2nd harmonic more dominant just by shifting the balance. And of course there's the low impedance that diminishes the effect of impedance rise with frequency. So the tone might seem to be brighter, but less hashy. It's definitely a good thing for some designs. If you don't like it for your amp, don't use it.

[Alan0354]

Surprise! Same person .

Then I should not talk much anymore, I don't want to offend him. I truly did not know it's him, I just read the article.

[nickb]

Yes, I scoped and reduce the drive so the preceding stage are all lineal......well in distort a little, but no where close to clipping. It basically looks like a sine wave. Of cause, if the preceding stage is clipping, all bets are off.

I am not saying the output is symmetrical, but just not very asymmetrical that produce a lot of even harmonics. No matter how I change the Rk or changing the +B, it basically moving the DC offset and slightly changing the amplitude. It does change the symmetry a little, but really nothing to write home. You can change the symmetry more by playing with the normal common cathode stage by adjusting the Rk and plate resistor to clipper fast on the top or bottom of the sine wave.

Alan - is it possible that what you are seeing is the negative going slew rate limited waveform caused by the capacitive load of the TS? Lift the connection to the TS and see what happens.