If it were me I would not modify a Dr. Z amp...I believe that his designs are solid and should not be tampered with.
The NFB is mainly there to reduce distortion in the output stage and secondarily it affects the output impedance of the output stage and alter speaker damping.
There are schematics floating around on the net of Dr. Z's designs, mostly from reverse engineered attempts...an accurate schematic is necessary for accurate analysis.
Here's a well-written article on the subject of NFB by Randall Aiken:

What is Negative Feedback?

Guitar amplifiers

www.aikenamps.com

 

5F4 shows 1.5k cathode resistor to ground and 56k FB resistor (pin 3) to spkr out, looks good but I can't quite make out the value of the FB resistor. 5F4 has only two FB resistors to my understanding, the 56k here and the 4.7M in the preamp.

Good old Mike Z, copying everybody's designs and calling them his own.

 

So I've been looking a little longer and researching since I posted this, It seems like the 1.5k was connected with the filter caps and now it has its own separate path to ground so not sure what effect that has.

If it were me I would not modify a Dr. Z amp...I believe that his designs are solid and should not be tampered with.

My intent if anything would be to bring it back to stock as it was modified before I got it.

 

I don't think the location of the resistors matters in this case as far as function but yes, 5F4 layout shows them on the board as you've indicated. Who moved them and why? What is the value of the FB (stock 56k) resistor?

 

I'll verify the resistor value tonight, someone had commented on FB that it was probably just done to minimize noise by isolating the Cathode resistor so it does not drain with the filter caps. However instead of leaving it on the turret board they made the change p2p instead most likely to minimize wire.

 

I've dickered with lifting the feedback resistor on my 5F2-A, to get more "bite". Possible someone removed the feedback resistor for the same reason and simply reinstalled it in a different physical, but not electronically, different location.

The cap that normally connects cathode to ground provides greater gain for frequencies above what is dictated by the cap value. You can see 25uf caps to ground in parallel with 1k5 resistors in earlier stages. Those large-value caps provide gain for the full guitar spectrum. Placing a resistance in series with any of those caps will reduce the amount of gain provided, especially if it is greater than the 1k5 resistor. The "Presence" control puts a small-value cap in series with a 5k pot, and puts that in parallel with the 1k5 fixed resistor. So the 5K pot adjusts the amount of gain for high frequencies rather than the entire spectrum.

BUT, since there is negative feedback from the output side of the transformer, the control essentially adjusts how much of that negative feedback (for top end only) is "bled off" to ground. If one used a slightly larger value cap (e.g., 0.22uf), the presence control would bleed off some of the upper mids as well. If the Presence pot was 10k instead of 5k, there would be less bleed-off of the negative feedback.

I hope that all makes sense.

 

@mhammer that makes perfect sense thanks for that write up!

 

I would put it back the way it was and experiment with the NFB values to lower (or increase) the amount. You already know how it sounds with it lifted.