Grid stopper for phase inverter

[TransLucid]

There's been a lot written about the importance of using relatively high-value grid stoppers on gain stages to prevent grid conduction (and the resulting nasty sounding distortion) when the input voltage is high. I've noticed that the signal input of the phase inverter often does not have a stopper, and, since it is often overloaded, I'm wondering why not.

When fed from a tone stack, the output impedance of the stack is fairly high, which may reduce the problem. However, in many instance the PI is fed directly from a gain stage.

Anyone have thoughts?

Bob

[Wilder Amplification]

Are you speaking of the long tailed pair phase inverter?

[MWJB]

It's obviously not an issue because amps without grid stoppers on the PI are capable of making maximum potential RMS power output. Plus the PI follows the tone stack & volume controls, so you decide how hard you drive it after W RMS has been achieved. There is also usually some signal attenuation prior to the PI coupling cap where the PI is fed directly from a gain stage.

[TransLucid]

Yes, but also from a more general perspective. If avoiding grid conduction is important when overdriving a gain stage, isn't it also important when overdriving the PI?

[MWJB]

LTPIs especially tend to run higher voltages than preceeding stages & half the voltage gain, or less, than regular gain stages...so they are less prone to distortion. As you get later down the signal chain, subsequent stages typically have less to do with the voicing/character & more to do with efficient transfer of the signal, preamp "overdrive" is typically culivated in the earlier stages...so you're not normally looking to generate overdrive at the PI. No reason why you can't, but it just seems counter intuative to have a PI that overdrives too easily, if whatever you do earlier is then unable to provide a clean signal. Some amps have controls that allow you to unbalance the PI...when this is achieved, clean headroom goes out the window.

[jpfamps]

There's been a lot written about the importance of using relatively high-value grid stoppers on gain stages to prevent grid conduction (and the resulting nasty sounding distortion) when the input voltage is high. I've noticed that the signal input of the phase inverter often does not have a stopper, and, since it is often overloaded, I'm wondering why not.

When fed from a tone stack, the output impedance of the stack is fairly high, which may reduce the problem. However, in many instance the PI is fed directly from a gain stage.

Anyone have thoughts?

Bob

Grid conduction into the PS is usually not good for too reasons.

1) Can cause blocking distortion as the coupling cap into the PS is charged up. This can be a real problem in amps that employ the long-tailed pair and global negative feedback, as once the power stage is driven into clipping (and it will in a guitar amp), then the the feedback loop breaks down and there is a large signal increase feeding the PS.

2) Both the common phase splitters, the long-tail pair and concertina, can generate some fairly nasty artefacts when driven into grid conduction.

Of these I would consider blocking distortion to be the worse problem.

Many older hi-fi amps use DC coupling between the final gain stage and the PS to eliminate this, and in my experienec it works well, although grid stoppers would be another solution.

[pdf64]

Theoretically it might be a problem, but in reality it doesn't seem to be. For instance, with a post phase splitter master vol on a long tail pair type phase splitter, blocking distortion on the phase splitter should be obvious and identifiable, but I've never noticed it. Whereas I do notice it on the coupling caps to the output tubes, eg, the vintage 0.1uF into 220k bias feed - I prefer 0.047uF into 220k. Pete.

[tubeswell]

I recently tried 470k on the grid of the cathodyne stage in my PR cloney thing, and really like the way it goes more smoothly into break-up now.

[TransLucid]

tnx for all your replies. I'm gonna take tubeswell's advice and try it, although 470k seems a bit high - no HF rolloff?

[pdf64]

No audio HF roll off, see towards the bottom of the page-
The Valve Wizard
Is your PI an LTP or cathodyne?
Pete.

[TransLucid]

Tnx for the reply. LTP, but I'm also interested in the general case.

There are two diagrams towards the bottom of the page you mentioned. In the first, the 470k resistor is just to bias the stage, and in the second it's part of a DC-coupling scheme, so there is no coupling cap to charge up and clamp the grid, so neither apply to my situation (AC-coupled with or without a grid stopper).

Bob

[jpfamps]

Theoretically it might be a problem, but in reality it doesn't seem to be. For instance, with a post phase splitter master vol on a long tail pair type phase splitter, blocking distortion on the phase splitter should be obvious and identifiable, but I've never noticed it. Whereas I do notice it on the coupling caps to the output tubes, eg, the vintage 0.1uF into 220k bias feed - I prefer 0.047uF into 220k. Pete.

With a post phase splitter master volume as you turn the volume down you are also reducing the open loop gain, and hence the feedback factor and are thus "correcting" the problem before it happens.

Grid conduction and blocking is definitely more of a problem in the power stage.

[Tubis]

In my experience you only need the grid blocker in front of high gain stages. In preamp for example a 100k-1k/22uF, yes. A 100k-10k no. The PI no.

[TransLucid]

My understanding is that grid conduction (and subsequent blocking distortion) is only due to conditions at the grid, not to the gain of the stage. Grid conduction will occur when the grid is positive wrt to the cathode. Of course, using a higher value cathode resistor can raise the cathode voltage, making the grid more negative wrt to cathode and thus avoiding grid conduction until the grid voltage gets large. This is also true for LTP PIs, where the cathode is at a fairly high DC voltage. Also, the coupling cap is often smaller with the PI, so that will help ameliorate blocking distortion.

However, the signal levels driving the grid of the LTP can be very high, high enough to cause grid conduction. So, I guess there are arguments both ways...

[Tubis]

My understanding is that grid conduction (and subsequent blocking distortion) is only due to conditions at the grid, not to the gain of the stage. Grid conduction will occur when the grid is positive wrt to the cathode. Of course, using a higher value cathode resistor can raise the cathode voltage, making the grid more negative wrt to cathode and thus avoiding grid conduction until the grid voltage gets large. This is also true for LTP PIs, where the cathode is at a fairly high DC voltage. Also, the coupling cap is often smaller with the PI, so that will help ameliorate blocking distortion.

However, the signal levels driving the grid of the LTP can be very high, high enough to cause grid conduction. So, I guess there are arguments both ways...

Yea that seems right. And thats why the PI doest need any blocker. But the coupling cap is usualy 22nF so i dont think thats the cure......