Is this what blocking distortion sounds like?

[motzu]

Hello,

my new 5F1 build is in the thorough testing phase, and after fixing the hum and most of the rattling at high volume (as much as the rectifier tube du jour will allow it), I've found this issue: when I turn the volume all the way up (12/12) and up the volume on my guitar (P90 pickup/bridge) to about 70-100%, I eventually get this fizzing sound accompanying the notes.

I assume that what's happening here is what's usually known as "blocking distortion", a power tube problem. My build is pretty straightfoward, if you take the Weber layout ( http://taweber.powweb.com/store/5f1_layout.jpg ) the only difference in my build is that I'm using a 470R/3W metal oxide screen grid resistor, and a 1K5/0.5W carbon composition grid stopper resistor on the 6V6GT output tube.

I'm assuming I need to up the value of the grid stopper to eliminate this problem?

I'm attaching an MP3 I recorded of the problem. Is this what blocking distortion is supposed to sound like in a 5F1: fizz.mp3 ?

[MWJB]

"I'm assuming I need to up the value of the grid stopper to eliminate this problem?" You shouldn't need to increase the grid stopper by much (try it, but if you need more than 5-10K stop because you're just doing a band-aid fix, you would be better off refining the layout), but other courses of action might be try 12AY7/5751/12AT7 in V1, try removing the 25uf cap bypassing V1, pin 3 cathode resistor, or turn the amp down from 12/12.

P90s can put out a lot more voltage than the single coils that the amp was designed for.

[motzu]

You're right, I was thinking about going with a 3K9 or 4K7 resistor there. It never crossed my mind to go over 5K even.

I'm not going to remove the V1 25uF bypass cap, I just added that and I'm pretty happy with the results (apart from this thing, obviously). Thank you for bringing that up, I should have mentioned that I don't remember this occuring before installing the V1 bypass cap (but then again it was only a few days after finishing the build), and I'm not sure it ever happened on input 2 of the amp (input 1 being the high-gain one).

Turning the amp down from 12/12 is probably the wisest recourse, but I'm a perfectionist and thus not a wise man.

[motzu]

I guess what I didn't make explicit in my initial post was that I'm asking if it sounds plausible to everyone that this is an actual case of blocking distortion and not some other problem. I'd suspect a cracked solder joint or something of that nature, but I've tested all connections for continuity, and they all passed a visual inspection. And it does only start to happen when the volume pot is above 9.5-10 on the high gain input, which is probably not something that has much to do with the quality of the solder joints.

The speaker is fine, I'm pretty sure. And once it starts sounding like that (with the volume up), sometimes the fizzing remains even after turning the volume down to where it would never have started. Couldn't be an OT issue, could it?

I don't have a 3.9K or 4.7K resistor to swap the 1K5 one with, so until I get one I still have time to try to understand exactly what's going on before going in again.

[Wakculloch]

That doesn't sound like blocking distortion to me. I, like you, would suspect connections, valve bases, valves then components.

[pdf64]

No, that's not blocking distortion - that pretty much only happens when a stage is overdriving.
You may have parasitic oscillation - have you got a oscilloscope?
A 10k or 22k grid stopper fitted to the second 12AX7 stage, one end direct to the tube socket terminal, would help if it was parasitics.

[motzu]

No, that's not blocking distortion - that pretty much only happens when a stage is overdriving.
You may have parasitic oscillation - have you got a oscilloscope?
A 10k or 22k grid stopper fitted to the second 12AX7 stage, one end direct to the tube socket terminal, would help if it was parasitics.

No, I do not own an oscilloscope, sorry.

If I'm on 12/12 on the high gain input with a hot P90 full on doesn't that mean that I'm overdriving the amp? Am I using the term wrong?

I also have a Peavey Classic 30 amp in the same room as the small 5F1 amp, and it doesn't do that (but I didn't try to bring it all the way up because it's too loud for where I'm living), if that helps.

If it's parasitic oscillation, what could cause it?

[pdf64]

It didn't sound particularly overdriven to me - think of Hendix's fuzz face freak outs, you can hear the sound getting farty/almost cutting out sometimes. To get that blocking distortion bias shift, you need to hit a stage with a sustained massive overdrive, which that clip is nowhere near. Mild overdrive of the initial pick attack, but the main body of the note seemed clean.
Oscillation occurs when there's sufficient gain (>1) and phase shift of 360 degrees (which puts the output in phase with the input). That easily occurs with a wide open, cascaded pair of 12AX7 stages, as there are 2 signal inversions (giving the necessary 360 degress shift) and a lot of gain (say 60 x 60).
Think how close the 1st grid is to the 2nd plate.
And that's before the additional gain and phase shift of the 6V6 and OT, and the proximity of the speaker socket to the 12AX7.
Parasitic oscillation occurs when the conditions aren't quite right for free running oscillation, but when a regular signal is applied, at certain times in the waveform the correct conditions come together, and you get a fizz / blip type of sound (or might be ultrasonic) that rides on the main signal.
Which is what I think I can hear on that clip.
Poor lead dress is the main culprit (on previously good designs) for oscillation. Or the design may be borderline, and certain tubes/atmospheric conditions/lay lines, set it off.
You can get rid of oscillation, parasitic or otherwise, by mounting grid stoppers (including the 68k input mixers) directly on the tube socket terminals. Correctly sized, they work with the tube's Miller capacitence to roll off upper frequencies.
And by analysing/improving the lead dress, thinking of the grid wires as receiving antenae, running them close to the chassis, well away from signal wiring further along the signal chain.

[motzu]

Thank you for taking the time to explain. About the 68K resistors, I've heard it before that they ought to be installed directly on the tube socket, but for most 5F1s I've seen (in real life and in gutshots posted on the Internet), they're actually installed all the way on the other end: on the input jacks (see this picture of a Victoria 518: http://proguitar.de/ProGuitarPreisli.../518TInnen.jpg ).

As for the massive overdrive issue, there was massive overdrive, but once I started getting that effect I've backed down on the volume pot of the guitar so that you will be able to hear the note-riding distortion better (at full-on overdrive you could have mistaken it for normal stuff).

[MWJB]

"Thank you for taking the time to explain. About the 68K resistors, I've heard it before that they ought to be installed directly on the tube socket, but for most 5F1s I've seen (in real life and in gutshots posted on the Internet), they're actually installed all the way on the other end: on the input jacks (see this picture of a Victoria 518:" They're not specifically "grid stoppers, they are the voltage divider for the attenuated input, a 5F1 does not need grid stoppers aty any stage of the 12AX7 if built correctly, if you just had 1 input you could just use the 1Meg load resistor.

If the sound you don't like ONLY started when you fitted the 25uf bypass cap at V1, then that is the culprit (5F1 weren't designed around P90s, they may have added the cap to juice up the signal with Fender single coil pick ups).

[motzu]

They're not specifically "grid stoppers, they are the voltage divider for the attenuated input

No no, I didn't say they were grid stoppers, I was merely replying to this: "you can get rid of oscillation, parasitic or otherwise, by mounting grid stoppers (including the 68k input mixers) directly on the tube socket terminals".

[MWJB]

Sure, but a 5F1 should not have, or need, grid stoppers on the 12AX7...oscillations should be cured by revising layout & lead dress.

[motzu]

Sure, but a 5F1 should not have, or need, grid stoppers on the 12AX7...oscillations should be cured by revising layout & lead dress.

I was talking about increasing the value of the 6V6 grid stopper, and I had assumed pdf64 was too.

[pdf64]

My preference here is to fit a grid stopper to every grid, post #6 I advised an additional one to the 2nd 12AX7 stage. I see it as basic good practice, and so position the 68k input mixers on the socket terminal, rather than the circuit board or input sockets.
I acknowledge that Fender, and subsequent derivations, get away without it, in a lot of cases, even in some very sensitive parts of the circuits.
But it stops this kind of problem, and can often help to reduce circuit quirks, such as channel crosstalk (in combination with good lead dress).
It's useful to assess each grid circuit, and consider how to best arrange the wiring to miminise interference and suseptibility to oscillation.

[Krwkka]

Hey, Motzu. Did you ever solve this problem?