negative feedback question

[navdave]

If the 33R resistor is the shunt and the 3.3k setting the bias for the tube does the the 3.3k resistor also block some feedback from reaching the tube?
http://www.el34world.com/charts/Schematics/files/ampeg/Ampegv4poweramp.pdf

[PRR]

Kinda.

[navdave]

Kinda.

Hmmm so not enough to make a big difference?
So the 33r and 3.3k are forming voltage divider for the feedback?
And at the same time setting the bias for the tube with a total resistance
of 3.303K ?

[sluckey]

The 4.7K and 33Ω form the voltage divider for the feedback. The 3.3K (plus the 33Ω if you want to bother) set the tube bias. 3.3K plus 33Ω is 3.333K, not 3.303K.

[navdave]

The 4.7K and 33Ω form the voltage divider for the feedback. The 3.3K (plus the 33Ω if you want to bother) set the tube bias. 3.3K plus 33Ω is 3.333K, not 3.303K.

Mucho thank you for the explanation Sluckey.

[tubeswell]

You may think of the 3k3 bias resistor as 'bouncing up and down on top of' the 33R.

The amount of NFB at the junction of the 33r and the 4k7 is 33/(33+4700) = 0.007 of the VAC swing at the OT secondary. This is just enough to take the edge of the cathode voltage swing in the stage that is driving the LTP.

[PRR]

At max power we have 52V peak at 7027 grids, 40V peak at 8 ohm tap. Output stage has gain of 0.8. 12AU7 stage has gain near 14. 12DW7 with 3.3K under has gain like 30.

30*14*0.8= 336

The 4.7K and 33 define a gain of 143.

NFB is slim when loaded in nominal impedance; it mosty controls the loudspeaker bass resonance where impedance may be 5X-10X nominal. Here the forward gain could exceed 1,000, but the NFB loop restrains it to near 140.

The 3.3K under 12DW7 is _local_ NFB in the 12DW7 stage, making it more linear. We really do not care: the 12DW7 only has to make 52V/7= 7 volts peak, where its THD will be under 1%, far less than the 12AU7 driver or the 7027 outputs. 

If you bypass the 12DW7's 3.3K you double the gain in this stage, have more NFB overall. But clearly Ampeg was not aiming for "high NFB"; indeed other models omit NFB around the output stage.

Depending on your speaker, you may wish to increase the 4.7K to 10K, 20K, 50K, or open, to get a more "raw" sound. 22K would make NFB similar to many Fenders, but won't make it sound like a Fender. Uh, a 120W amplifier may be too raw. You could reduce the 4.7K to 2K or maybe 1K, be more clean, but also lower gain.

[HotBluePlates]

At max power we have 52V peak at 7027 grids, 40V peak at 8 ohm tap. Output stage has gain of 0.8. 12AU7 stage has gain near 14. 12DW7 with 3.3K under has gain like 30.

30*14*0.8= 336

The 4.7K and 33 define a gain of 143.

For all of our future reference, can you explain how you arrived at the 4.7k and 33 ohm resistors defining a gain of 143?

[PRR]

> how you arrived at the 4.7k and 33 ohm resistors defining a gain of 143?

Ratio. 4700:33 is 142.4, right? (Don't trust me, my slipstick is sloppy in the dry winter air.)

[printer2]

> how you arrived at the 4.7k and 33 ohm resistors defining a gain of 143?

Ratio. 4700:33 is 142.4, right? (Don't trust me, my slipstick is sloppy in the dry winter air.)

Probably used it so much the numbers are wore off.