12AT7 PI in place of 12AX7

[daz]

If you replace the X7 with a T7 in the PI of a mashsall style amp, what do you see as the advantages and the disadvantages?

[Alan0354]

You definitely have a bigger input range. T7 has a grid bias of a few volts so you can swing more at the input. That's the reason I never change the T7 to X7 in my Fenders and my build.

I don't think the little bit lower amplification factor matters, with NFB, you don't see much difference.

Could it be that's the reason Marshall is better for distortion as the PI input is limited by X7 and you get better distortion. Fender use T7 so it maintain clean longer.

[daz]

You definitely have a bigger input range. T7 has a grid bias of a few volts so you can swing more at the input. That's the reason I never change the T7 to X7 in my Fenders and my build.

I don't think the little bit lower amplification factor matters, with NFB, you don't see much difference.

Could it be that's the reason Marshall is better for distortion as the PI input is limited by X7 and you get better distortion. Fender use T7 so it maintain clean longer.

Thats what i was hoping to hear. But i wonder why cascaded amps like soldonos and modern marshal have X7's still when amps like that mainly need a cleaner PA.

[Alan0354]

I am just guessing, maybe they want distortion all the way to the PI stage.

[daz]

I am just guessing, maybe they want distortion all the way to the PI stage.

Maybe, and people do often cite PI OD as a good thing. I've put a master after the PI several times and the OD that happens at the PI is worse than any pedal or any form of distortion i have ever hard. Loose, ratty, unusable. Maybe somehow they are able to shape the signal before it gets to the pi to sound good when it drive the pi. I dunno. But more likely I think pi OD with a OD style preamp is not a good thing.

[Chuck H]

The thing you must remember is that it's all in the interaction. The X7 PI is typically biased a little hot. You're never supposed to hear the cutoff end of either duty cycle. The vintage amps weren't designed with post PI master volumes in mind. The combination of PI clipping and power tube clipping (where the cutoff end of both is not heard) is a classic tone for many amps. The post PI master is just a way to get more distortion. Not necessarily better distortion. It's possible that a different bias point could yield better results.

On the whole I agree with Alan. The circuits assigned to different models do what they do. Changing them is asking for trouble unless you want something different. Which is fine too.

I use the X7 for an inverter most of the time because it saves me the trouble of keeping different tubes around for prototyping. You can always adjust the circuits after all. No rules that state an X7 LTP PI needs to have a 470 ohm bias resistor and a 10k tail. Another reason I use the X7 is that I still have a small supply of salvaged tubes. The T7's tend toward a more "function" type construction and are less resistant to microphonics than my X7 stash. I'd have no issue designing with a modern, audio oriented T7 though. The gain difference is really insignificant since WRT modern design it's likely that either tube will clip the power tubes into frappe. Actually, the T7 may have an advantage here since it should tolerate grid demands a little better in the average guitar amp circuit. I still contend that it's in the design more than the tube though.

[Alan0354]

I think he want to straight replace, not design into it. My guess he is going to get less distortion with the T7. Just a wild guess. Honestly, I never try to put a X7 into PI yet, maybe I should try that!!!!

[daz]

Well, In my case where i want my PI clean with plenty of headroom i think a T7 may be the way to go. I'm going to experiment with it and with different bias resistors at louder volumes. I think my PI is not nearly as clean as i thought, and much better with the 2k bias R, but still i think it could be better. The T7 might do just what i'm trying to get from it.

The thing you must remember is that it's all in the interaction. The X7 PI is typically biased a little hot. You're never supposed to hear the cutoff end of either duty cycle. The vintage amps weren't designed with post PI master volumes in mind. The combination of PI clipping and power tube clipping (where the cutoff end of both is not heard) is a classic tone for many amps. The post PI master is just a way to get more distortion. Not necessarily better distortion. It's possible that a different bias point could yield better results.

On the whole I agree with Alan. The circuits assigned to different models do what they do. Changing them is asking for trouble unless you want something different. Which is fine too.

I use the X7 for an inverter most of the time because it saves me the trouble of keeping different tubes around for prototyping. You can always adjust the circuits after all. No rules that state an X7 LTP PI needs to have a 470 ohm bias resistor and a 10k tail. Another reason I use the X7 is that I still have a small supply of salvaged tubes. The T7's tend toward a more "function" type construction and are less resistant to microphonics than my X7 stash. I'd have no issue designing with a modern, audio oriented T7 though. The gain difference is really insignificant since WRT modern design it's likely that either tube will clip the power tubes into frappe. Actually, the T7 may have an advantage here since it should tolerate grid demands a little better in the average guitar amp circuit. I still contend that it's in the design more than the tube though.

[Chuck H]

It's all about context. If you don't have the PI or power tubes clipping it makes good sense to choose tubes that give more clean headroom in a given circuit.

[teemuk]

As far as I know, the PI tube will not really effect overall headroom. Your headroom stops where power tubes clip and thus practically your headroom stops when power tube grid signals exceed certain levels (saturation or cutoff).

If you get certain output signal magnitude with 12AX7 then the lower gain tube will output a bit less. In other words, in comparison to higher gain tube it tolerates a bit higher input signal before output tube grid signals exceed those clipping thresholds. So I don't really see where any headroom comes in. If your amp was 100 watts before you changed the PI tube its probably going to output that 100 watts afterwards too. Now you just need to dial the master volume a bit higher to get the same sound and performance.

Lower gain will also affect feedback loops because lower gain in PI means lower open loop gain in the entire amp overall. So basically anything tampering with open loop gains will also tamper with the ratio of applied negative feedback. This may or may not, depending on design, pose discernible effects.

[daz]

Im already have variable NFB all the way from none to typical marshall.

[Gingertube]

What is different about a 12AT7 compared to a 12AX7?
- mu is slightly less, around 70 instead of 100 , that tends to give a liitle less gain
- rp is a lot lower, 11K instead of 65K, that tends to give a little more gain and a fair bit less ditortion
- gm is a lot higher, 6.5 mA/V instead of 1.6 mA/V, that means it is more susceptible to parasitic oscillation and grid stops are more important.
Grid Current is higher - watch the impedance in the grid circuit.

As far a substituting a 12AT7 for a 12AX7 in a diferentail phase splitter, It is likely to give you much the same gain (mu lower but rp less too), a cleaner sound, and a "s..t load" of trouble unless you put grid stops on the triodes, the old" 8/gm" Rule of Thumb suggests 2K2 or more for the grid stops, I'd use 4K7.

Cheers,
Ian