That's a different kind of tertiary winding. The approach (maybe) used in your amp is simply a separate secondary winding from the speaker windings. It helps isolate the feedback circuit from the PA speaker outputs, at the cost of a more expensive output transformer.
After reviwing the Marantz 5 and reading the posts I still dont have a very good understanding and I dont really think that they do either ? , A cap to the nfb resistor or direct to the splitter cathode, lower nfb resistance and ground or other lead to nfb resistor ? . DC bypass ? Its all pretty confusing to me being new to it and no one really seems to be an expert on it. I understand how the old fc speakers bypassed current and voltage to make emf and how that properly regulated could create some really beautiful tone , it seems to me that may be the basic idea here with the two winding but I dont know enough about it to decide if it should go to nfb res. and or spkitter cathode, or to ground on one side, or to go to one of the el34s instead. The only thing I know for sure is that the 80 ohm nfb resistor in it is most likely way too big. 3 winding transformer ; 1. It reduces the unbalancing in the primary due to unbalancing in 3 phase load. 2. It redistributes the flow of fault current. 13 ohms is what the tertiary winding reads in my OT unhooked from end to end. I dont fully understand its operation other than comparing it to something like a field coil electro magnet , if that is even similar ? but See the attachment. There has to be a better approach than trial and error to hook it back up. The windings read 13 ohms across but are not the same, the white one is 15 ohms to chassis or comm. And the red or orange is 2 ohms to chassis or comm. They also make y and delta windings as well and I cant find any info on audio tetiary windings, almost like it is some kind of a secret or something ?
Tertiary windings for negative feedback were a relatively uncommon -- and now arguably obsolete -- technique probably used more on PA amps than anything, and they've largely fallen out of favour because it made for expensive output transformers without much justification unless you're constructing a tube PA amp. These days, nobody is going to construct a tube PA amp.
Use of a tertiary winding is exactly the same as tapping the negative feedback from one of the speaker taps, which is more common.
Here's an example of a tertiary feedback winding used in a vintage ham transmitter:
https://faculty.frostburg.edu/phys/latta/ee/ranger/schematic/rangerschematicfullres.jpg
Ham transmitters for voice work are modulated by the audio (voice) signal via an audio amplifier. In the diagram, T2 is the audio output transformer. Note the tertiary negative feedback winding is grounded on one side and on the other side injected into the grid of the final preamp dual triode.
Injecting negative feedback into the final preamp, instead of earlier in the preamp chain, is reasonable -- a PA amp isn't striving for the utmost in frequency response or distortion-free operation. Also, a PA amp is likely to be divided into separate channels prior to the final preamp stage anyway, so there's no common audio path into which to inject negative feedback
until the final preamp.
In your case, to wire up negative feedback from the tertiary winding, a simple arrangement might be to ground one of the tertiary winding leads and connect the other to the 80 ohm negative feedback resistor on the phase splitter 12AX7. If that kills the output, try swapping the tertiary winding leads. If it oscillates, swap them back. You may need to add additional resistance.
However, I've reviewed your output transformer resistance readings. Because they're scanned I find them hard to read. It's difficult to tell whether the unattached output transformer leads are a true tertiary winding, or simply part of the same secondary as the speaker outputs. If they are taps on the same secondary as the speaker outputs, you
don't want to ground one of the two disconnected leads -- leave it floating -- and the other will be used to obtain negative feedback.
A complete table of lead-to-lead output transformer resistances -- typed in a post here rather than handwritten and scanned -- would be helpful.