12AT7 VS 12AX7 in PI

[67polara]

What is the difference between the sound of these two tubes in PI section is one better than the other for this task?

Tony

[Tone Junkie]

Tony back a page or so or even maybe on a differant site sorry my brain is slow tonight, some one linked to a written piece saying that the 12at7 had a lot more head room for a PI tube than a 12ax7 I will try to find the article and link to it.
Bill

[67polara]

Thanks, I just had this NOS RCA 12at7 I have been dieing to use LOL.  I just plugged it in and Wow I like it. 

Tony

[darryl]

I just plugged it in and Wow I like it.

If it sounds good, just use it.   

[tubeswell]

Quite a few of the BF amps used 12AT7 LTPs with smallish coupling caps on the input (because that's all that's needed with the excessive bandwidth available). However I do believe that the BF circuit set ups are even not the most optimal for 12AT7 PI tubes, and in particular, they may go 'even better' with 47k-ish plate resistors and possibly even slightly-cooler biasing.

[Tone Junkie]

What would be your suggestion my friend on the biasing.Thanks Bill

[darkbluemurder]

However I do believe that the BF circuit set ups are even not the most optimal for 12AT7 PI tubes, and in particular, they may go 'even better' with 47k-ish plate resistors and possibly even slightly-cooler biasing.

Fender used 47k plate resistors with 12AT7 tubes in the PI of the silverfaced amps. However, with the biasing they went hotter and not cooler (270R instead of the BF 470R cathode resistor). They really cooked those 12AT7 tubes in these amps. Rivera went a little cooler with 680R in the II series (with a 12AX7) whereas Dumble mostly used 820R (also with a 12AX7). There is no harm in going with 47k plates and 820R cathode. 

[HotBluePlates]

Quite a few of the BF amps used 12AT7 LTPs with smallish coupling caps on the input (because that's all that's needed with the excessive bandwidth available).

I would disagree that the reason is available bandwidth.

What is the input impedance of a vaccum tube stage? As long as the grid is more-negative than the cathode, then there is almost no current flow at the grid, and the input impedance looks essentially infinite. So the input impedance seen by a preceeding stage is equal to any grid-to-ground resistor present.

When there is feedback in/around a stage, or if the grid reference resistor goes somewhere other than ground, the input impedance might not be what we expect.

Notice that in long-tail and split-load inverters, as used by Fender, don't have the 1M resistor go to ground but rather to some largish resistance which then goes to ground. Without deriving it, know the apparent input impedance is much greater than 1M. Since the input impedance is very large, then coupling cap can be much smaller than you'd expect and still maintain a low roll-off frequency.

Aiken, Merlin and a number of the old tube books do a good job of explaining why the input impedance is higher, and exactly what that impedance is.

Note also that feedback may raise or lower the input and/or output impedance of a stage, depending on how the feedback is derived and applied. Feedback is a tricky subject with no one-size-fits-all explanation.

[HotBluePlates]

What is the difference between the sound of these two tubes in PI section is one better than the other for this task?

Tony

Depends on the circuit and what you want to hear. It would be fastest to swap tubes to hear for yourself.

... some one linked to a written piece saying that the 12at7 had a lot more head room for a PI tube than a 12ax7

A properly-designed stage will probably have more input headroom when using a 12AT7 than a 12AX7, because the 12AT7 has lower gain. That said, a properly-designed output stage (phase inverter and power tubes) will have the output tubes be the limiting factor as far as clean headroom goes. That's because big output tubes (at least in the past) cost more than small output tubes, and you want to design your amp to enable the amp to make all the clean output power it can.

If you're trying to decide which tube to use, the one specified by the manufacturer is the "right one" or the one they designed to be right. But it is generally safe to swap any of the 12A_7 tubes to see if you like one more than the other.

[tubeswell]

Quite a few of the BF amps used 12AT7 LTPs with smallish coupling caps on the input (because that's all that's needed with the excessive bandwidth available).

I would disagree that the reason is available bandwidth.

Oops - when I said 'available bandwidth', I was assuming that the roll-off point that was produced by the interaction of the cap and the input impedance was implicit - sorry I guess I should've spelled it out. The obverse is that too-large-a-cap will result in blocking distortion, but the available bandwidth with a .001uF cap and a 12AT7 is such that you don't need a larger cap there to get an ample signal for your geetar.

[HotBluePlates]

I see where you were coming from.

I was trying to point out that the small cap was due to the effect of bootstrapping on input impedance, and not due solely to the use of a 12AX7 vs. a 12AT7. A 12AX7 used in one of these circuits also exhibits a higher-than-expected input impedance.