Output tube snubber kits, any experience?

[Pedro Vecino]

Yes. I know it well but with its typical values (250K and 4700pF) it has a much more severe behavior. An AC30 on its top boost channel of course needs it but on Plexi types it interferes excessively with their overdrive texture. In AC30s with crossline master volume it also has a variable behavior depending on the master volume setting.
I think that the conjunctive filter effect as you have described generates more complex interactions. Ideal when only a pinch of salt is needed and if the amp accepts it. I tried previously with a single ended and it didn't work well. I had to look for other formulas.

[nickb]

It is difficult to believe that that area between 5K and 10KHz, seeing the small difference between the curves, acquires so much relevance but I want to think that this is where the harmonic content in overdrive sits and that is always delicate matter.

Thanks again!

Looking at frequency response curves in simulation for such a situation as this is going to give a completely meaningless answer. The reason is that the circuit is linearized to compute the response. Here, the amp is overdriven so in a very non-linear area of operation. Nor will the feedback be able to flatten the response any longer. The transient response tells the real story, but only at one frequency at a time and for the given level of overdrive. Nevertheless, taking the FFT of that should be informative.

The capacitor will resonate with the leakage inductance, typically 100Khz region or thereabouts. The current through the network will reach it's 3db point at roughly f= 1/RC, it's complicated by other circuit factors but that will get you into the ballpark,

[Helmholtz]

The capacitor will resonate with the leakage inductance, typically 100Khz region or thereabouts.

I measured a number (about 40) of PP OTs, mainly Marshall, Fender, Mesa types as well as replacements.

Leakage inductance varied between 3mH and 100mH. Reflected speaker inductance contributes maybe another 200mH (in series with a few kOhm of AC resistance).

(Primary self-capacitance ranged between 80pF and 800pF.)

[Chuck H]

It is difficult to believe that that area between 5K and 10KHz, seeing the small difference between the curves, acquires so much relevance but I want to think that this is where the harmonic content in overdrive sits and that is always delicate matter.

There's a school of thought that subscribes to the 3dB rule. I don't entirely buy it. I think there are subtleties that, taken in accumulation, make a significant difference. Where and how they're implemented as well. That's why I posted the sims. You can now look and draw your own conclusions. A 1dB difference in one place tempered against other differences is the nuance of what we do here. I only hope I have helped.

[Helmholtz]

FWIW, the RCA receiving tube manual (RC-30) recommends a "corrective filter" using a series resistor of 1.3 times OT primary impedance (Raa) and a cap value of around 50nF.
Its purpose is equalizing speaker impedance and response at higher frequencies. It is meant for non-NFB amps.

[Chuck H]

It is meant for non-NFB amps.

Excellent!

But I'll also mention that it wasn't proposed for guitar amps Where the purpose of this filter (for my own considerations) would be mitigating spikes under clipping conditions and perhaps some trimming of ice pick high end without overtly affecting other frequencies. I guess it's also unclear what the properties of the speakers were (compared to purpose built guitar speakers) for the above circuit recommendations.?.

[Helmholtz]

But I'll also mention that it wasn't proposed for guitar amps Where the purpose of this filter (for my own considerations) would be mitigating spikes under clipping conditions and perhaps some trimming of ice pick high end without overtly affecting other frequencies. I guess it's also unclear what the properties of the speakers were (compared to purpose built guitar speakers) for the above circuit recommendations.?.

The manual describes a method to find the best cap value with given speaker. It requires matching the amps gain at 400Hz and 1kHz by varying capacitance with speaker attached.

In practice this would mean feeding a signal to the PI input and varying the Zobel cap value until speaker voltages at 400Hz and 1kHz are the same.

[g1]

I'm just going to throw in these practical examples as I don't come across these filters (across OT primary) too often. Garnet seemed to like them. One is a guitar amp and one PA.

[Helmholtz]

I'm just going to throw in these practical examples as I don't come across these filters (across OT primary) too often. Garnet seemed to like them. One is a guitar amp and one PA.

Considering that these amps have global NFB and the cap values are relatively low, the Zobel seems to be used here as a snubber to protect the OT from high voltage spikes and resonant HF voltages - rather than "correcting" the speaker's audio response.
Also note the high voltage rating of the caps.

That's what I meant with different purposes in my earlier post.

I think the OP uses the Zobel to shape the tone of a clipping NFB amp, which would be a third purpose.

Dr Z seems to like and use a Zobel (aka Boucherot network, conjunctive filter, corrective filter etc.) for a similar reason:
http://4.bp.blogspot.com/-HPHDF2LjWb...9cc07a21_b.jpg

[Chuck H]

Also note the high voltage rating of the caps.

This is something you might learn the hard way if you're like me and don't have all the technical chops. I used a 600V cap the first time I implemented this circuit in a 2xel84 amp with 370Vp. I thought I'd be ok since it exceeded the plate voltage. I was unaware of the voltage circumstances happening in that circuit at the time. It should be noted that the design was intended for clipping without a master volume. It didn't take very long for the capacitor to basically short. It was a self healing type and in the process of repeated self healing became enough of a short to start cooking the 10k resistor. I knew the amp sounded "off" and I didn't pin down why before the resistor started to show browning on it's color bands.

I'm currently using a very high voltage "snubber" type cap and a "high energy" resistor rated for the highest voltage I could find in the Mouser catalog. Though I can't remember specifics on just what I used (without cracking the amp open) that did the trick. No failures and it's been many years.