Blocking Distortion?

[hesamadman]

I have a 50 watt build that I am having some "issues" with. It is a 2204 with an added tube stage. I have it setup so the tube stage can be bypassed and operate as a standard 2204. I think I am getting some blocking distortion. It doesnt seem very noticeable in all honesty. Heres only when you notice it. If you play a note really fast staccato like, the "high gain background hiss" drops quiet and fades back in. Its almost like a crappy noise gate, but it only really affects the noise the amp makes from being high gain. If you play some low note chugs it doesnt necessarily even affect the note. Just as soon as you cut it off, theres a quick drop and fade back in of the amp noise. I dont know if this is common or what. I mean the amp sounds and plays fantastic. Ive been trying to find this issue for a while now. I have tried many many things in the preamp to reduce this.Increase cathode resistors. Remove bypass caps. Coupling caps are all .022. Removed NFB (but put it back). Larger grid resistors. Is this common in ANY other amps out there?

[sluckey]

That sounds more like power supply sag to me. Did you create a separate B+ node for your added tube stage? If not, do so. Also try increasing the size of the filter caps. Just tack a 22µF across every filter cap as a trial. Do all of them at the same time.

[PRR]

Could be blocking. Amps do that. Is it a real problem?

Series grid resistors are one approach.

[hesamadman]

Also try increasing the size of the filter caps. Just tack a 22µF across every filter cap as a trial. Do all of them at the same time.

I thought the same thing about the power supply sag. The stage is connected to its own node. It had a 22uf cap. I paralleled a 22uf across it and I see improvement. The rest of the stages are 50uf cap cans. Should I clip some 22's across those as well?


 

Could be blocking. Amps do that. Is it a real problem?

It's really no problem at all. I just wasn't sure how common it was. And honestly at the point. The only way to notice it so much is to stand by the open chassis and hit a note. The noise inducted into the amp from being so close to the amp is all that fades and comes back.

[Tony Bones]

Since you have a sense of curiosity, as PRR said, the usual type of blocking - charging of the coupling cap - can be mitigated with a resistor between the coupling cap and the grid (that's where the current comes from that charges the cap.) Try 10k just to see if it kills the perceived blocking. If it works you can experiment with smaller values to see how they affect the fell of the amp.

There's another form of blocking that not everyone thinks about. It's charging of the cathode bypass cap causing a shift in bias. a 25uF cap and 1500 ohm cathode resistor have a time constant of about 1/4 second (really, it's about half that if you consider the R looking up into the cathode.) That's not too bad, but 3.3k with 100uF can result in something that feels like blocking.

[tubeswell]

Another form of output stage distortion (reported in Merlin's 2nd edition pre-amp book) can result from LTP imbalance under heavily overdriven conditions, whereby the inverting stage of the LTP goes into grid current limiting, causing the input coupling cap to charge, shifting the grid voltage more negative biasing the triode closer to cutoff. The signal at the inverting plate gets clipped strongly on its positive side due to cutoff, and its duty cycle increases (so it spends more time per cycle in positive than negative), and the non-inverting output does the exact opposite. So one of the signals being driven to the output tube grid will be driven harder than the other, which can lead to unequal blocking distortion in each power tube, which in-turn leads to crossover distortion and may cause the heavily overdriven tube to red-plate.


The solution is to use more centre-biasing for the LTP (to minimise grid current limiting and cutoff). I.e. this could be a 1k bias resistor. Not saying this is the problem, but what-the-heck?