Padded Paramount 300B, advice needed.

[richmi]

Hello,

In 2008, Paul Joppa gave me some instructions privately through the Audio Asylum on how to modify (pad) my Paramounts, as I had too much gain in my system.

We used leftover resistors from the Foreplay III kit.

Two 33K resistors were connected in parallel, which in turn were connected in parallel with the 249K resistor going from 20U to 18U, resulting in a 16.5K resistance value for the shunt arm of the attenuator. I have since replaced these paralleled resistors with a single 17K resistor.

The red wire coming from the input jack was disconnected from 20U. On end of a 80.6K was connected to that red wire, and the other end connected to 20U, as the series arm.

This modification would give 16dB attenuation, or about 5 clicks of the Sweetest Whispers (which is indeed the case).

I had to do something under the hood of the Paramounts, and broke one 80.6K resistor. So, I replaced both of these with 100K resistors, as these were the closest value I had.

Now, the input impedance is 116K (if I remember well), whereas it was 96K before.

What I would like to know is if I can leave these resistors (100K) in place, or are there any chances this change can damage the amplifiers? If this can be dangerous, I will order new 80.6K resistors.

Since the signal now goes through 100K and 17K to ground, I would think that there is less signal wasted to ground. I clearly don't understand electronics enough to know if there is something else happening. However, only today did I realize that the original modification also affects B2 which goes through 17K instead of 249K (connected to 20L through a 220 ohm carb comp resistor - this connection is somewhat not obvious, this is why I did not notice it the first time), then to ground. I suspect this change also impacts attenuation. A simple and short explanation of how this works here would be appreciated, if anybody cares to do it.

Thanks a lot for any comments.

Richard

[Paul Joppa]

The bottom line is that your current setup is just fine and will have no problems. The values are quite flexible; attenuation is set by their ratio.

There are two limits on the actual values. First, the sum (which is the input impedance of the pad) should be high enough that it does not demand too much current from the preamp. Foreplay, for example, works best into a load of 10K or greater. Second, the parallel combination of the resistances (which is the output impedance of the pad) must be low enough that the high-frequency response of the amplifier is not degraded. again for example, the Miller capacitance of a 12AT7 is about 90pF. To keep the -3dB frequency below 50kHz, the resistance of whatever drives the grid must be less than 35K. The parallel combination of 100K and 17K is 14.5K.

A minor matter is that the resistance from the grid (B2 in this case) to ground should also be less than the amount specified for the tube. I call this minor since that value is usually 1 megohm, that is 1000K, and in this application you will never get anywhere close to that! The reason for this limitation is that there is always some (extremely tiny) current going into or out of the grid, and if that current goes through a very large resistance is will develop enough voltage to upset the bias of the tube.

[richmi]

Thanks PJ for the explanation.

Richard