Snubber theory, confused

[Bob-I]

I've always assumed that the "snubbers" on the OD stages of the Hybrid filter some of the highs to ground, smoothing the overall tone.

Today Ted Weber posted this schematic, a faux SLO.
https://weberspeakerscom.secure.powweb. ... _schem.jpg

where he has a plate to cathode 120pF cap on CL1. He claims that it's a treble boost, because the cathode is a non-inverting input, unlike the grid with is an inverting input.

So that leaves me confused. In all my amps the snubbers have the effect of mellowing the tone, but Ted's explaination makes it sound like it should boost the highs.

Can anyone shed some light on this?

[llemtt]

No comment about what Ted says...

the cathode of the first stage is bypassed by a huge cap so for any freq higher than a few Hz it's an AC ground!!! there's is no feedback and the 120pf cap loads the output of the stage resulting in a low-pass filter and other more subtle effects.

Connecting the snubber at the bypassed cathode or the positive supply is the same, the cathode connection is just a pratical matter because it let's you mount the snubber directly on the tube socket without additional wiring.

teo

[Darkbluemurder]

If I can trust my ears then a snubber cap reduces highs - the bigger the cap the more highs are reduced. A logical technical explanation for me would be that the highs are bled to ground through the snubber cap and the cathode resistor (please correct me if I am wrong).

[llemtt]

yes your logical model is oversimplified but it works well, just a little correction

..highs are bled to ground through the snubber cap and the cathode resistor..

should be

"highs are bled to ground through the snubber cap and the bypass cap"

because at high freqs the reactance of the bypass cap is a lot less that the cathode resistor

cheers
teo

[ODwan]

Effectively the snubber cap is in parallel with the inner plate to cathode capitance ( which is mostly not mentioned in datasheets but it's there!)of the tube, thus adding to it, thus reducing high freq response. This is mostly used to prevent ultrasonic oscillations.
Timo

[Luthierwnc]

Building on ODwan's statement, the snubbers are called a negative feedback loop. What they do is put ultra-high frequencies out-of-phase (the plate and cathode are 180 degrees opposite) so they don't oscillate (feedback) uncontrollably. Sometimes the amp just squeals or creates parasitic oscillations that quietly rob the signal. On the Hybrid A there is also a NFB on CL2 from the plate to the grid. With three cascading gain stages in a row, you need a little behavior modification Some of the larger late tweed Fenders do the same thing.

Usually when an amp uses little caps to take highs straight to ground, they are called suppression caps like the infamous Fender silver-faced models.

I hope this helps,

Skip

[Bob-I]

So I'm gonna ignore Weber's explaination sice he seems to be leaving out many factors.

Overall, I'm gonna trust my ear. Snubbers reduce highs.

[Luthierwnc]

In this case, there is a little more to it. They reduce highs in the audible range but they also keep the amp from yowling. The signal coming off the grid is magnified and sent out the plate 180 degrees out-of-phase. By taking the highest frequences off the plate and putting them back in the the source signal exactly the opposite of the way they went in, you cancel them out before they get to the tube. Sort of like the plot of the Terminator. An example of positive feedback is walking in front of your monitor with an archtop. Not that I have ever done that personally.

Without whipping this too much more, the snubbers do cut the highs but they also keep them from multiplying uncontrollably and destabilizing the amp. A handy reference for this is Dave Funk's book. It is a compliation of his articles for Vintage Guitar with a chapter on negative feedback. Tangentally related, there is another chapter about gain. He was making an amp that had four preamp gain stages much like our beloved "D". It shows the waveform and how it clips after each stage. Good reading.

Skip

[Bob-I]

In this case, there is a little more to it. They reduce highs in the audible range but they also keep the amp from yowling. The signal coming off the grid is magnified and sent out the plate 180 degrees out-of-phase. By taking the highest frequences off the plate and putting them back in the the source signal exactly the opposite of the way they went in, you cancel them out before they get to the tube.

Now I'm really confused. The grid is 180 out of phase with the plate, but the cathode is a non-inverting input, so anything you put in comes out the same phase, but amplified. So this would support what Weber said that the plate/cathode connection would boost the highs.

It was always my belief that the snubbers sent some of the signal to ground through the cathode resistor.

Without whipping this too much more, the snubbers do cut the highs but they also keep them from multiplying uncontrollably and destabilizing the amp. A handy reference for this is Dave Funk's book. It is a compliation of his articles for Vintage Guitar with a chapter on negative feedback. Tangentally related, there is another chapter about gain. He was making an amp that had four preamp gain stages much like our beloved "D". It shows the waveform and how it clips after each stage. Good reading.

Skip

I'll have to pick up that book, do you know the name?

[ODwan]

Now I'm really confused. The grid is 180 out of phase with the plate, but the cathode is a non-inverting input, so anything you put in comes out the same phase, but amplified. So this would support what Weber said that the plate/cathode connection would boost the highs.

It was always my belief that the snubbers sent some of the signal to ground through the cathode resistor.



I'll have to pick up that book, do you know the name?

Actually the grid is 0°, the plate is 180° and the cathode is 0°. So plate and cathode are out, grid and plate are out and grid and cathode are in phase. Thats why a cathode follower is a non inverting stage.

I have Funk's book, too and agree that it's good reading. Nearly as good as Kevin O'Connors. I think it is called Tube Amp Workbook Vol.1: Fender Amps.
I wonder if there's ever gonna be a Vol.2 !?

[Luthierwnc]

I don't think Dave Funk is making amps anymore. I never tried to make money doing it but better men than me have gone toes-up. Don't think we'll see volume 2.

Bob, the quick rule of thumb is that the signal flips at each tube. The exception is pulling the signal off the cathode. Except for Bassmans and their decendants, you don't see it very much because it wasn't a cost effective use of a tube in the era when these things were mass marketed. That is why the gain part of the Funk book is so cool because it shows the wave being shaped in the first two triodes. In the next two, it lops the top off, flips it and "clips" the top off again. I need to re-read that.

Skip

[Bob-I]

[Actually the grid is 0°, the plate is 180° and the cathode is 0°. So plate and cathode are out, grid and plate are out and grid and cathode are in phase. Thats why a cathode follower is a non inverting stage.

Now that makes complete since. SO the snubber is a NFB loop using the cap as a filter..... cool.

[ayan]

[Actually the grid is 0°, the plate is 180° and the cathode is 0°. So plate and cathode are out, grid and plate are out and grid and cathode are in phase. Thats why a cathode follower is a non inverting stage.

Now that makes complete since. SO the snubber is a NFB loop using the cap as a filter..... cool.

Well, read Teo's first reply to your post. The cathode (in a bypassed stage, like all of Dumble's preamp stages) is at AC ground, so the snubber is just shunting the plate to ground via a small capacitance. Identical, i.e. the exact same, results can be achieved by bypassing the plate load resistor for the same stage.

Gil

[skyboltone]

Bob, the quick rule of thumb is that the signal flips at each tube. <snip> That is why the gain part of the Funk book is so cool because it shows the wave being shaped in the first two triodes. In the next two, it lops the top off, flips it and "clips" the top off again. I need to re-read that.

Skip

Thanks for that one Skip. Makes some stuff clearer for me.

Dan

[mat]

(newbie alert!)

Can I assume that the smaller (and same uF) snubbers, the easier it is to get the feedback from the amp ?

I can not get feedback from my clone in any pot settings (non HRM, hybrid A) with my tele (single coil) or Les Paul style (humbuckers). I also have mid cap (from .01 to .022 to .047) switch + mid boost and local feedback switch with off, 22M and 44M and of cource the PAB.

I have now 330p and 440p snubbers. I'll try next as both 220p..