Speaker EQ curves projecting exaggerated back towards OT and output tubes

[Guitarist]

Freyette was saying that NFB (and maybe OT) can even affect your overall sound this way.
I am about to order a flatter-frequency Eminence EM12N as I have good EQ in the varied preamps and would like to steer away for peaky highs. More into Santana/Schon/Sykes natural highs!

And~ applying this to especially with two tubes pulled from a 100 watt amp, giant Hammond transformers, would you hear better response with the 4Ω speaker choice as there is higher Reference Efficiency? 2Ω combined parallel pair even better? From Traynor factory they used 100 watt iron so adding another cab might be a good idea. Would one half or double the FB resistor?

thanks so much, hope this makes sense enough

[pdf64]

Cant make much sense of that; got a link for Fryette?
The gain of the power valves couples with the load impedance. The application of NFB will tend to counteract that.

[Mick Bailey]

Given the varying impedance of a speaker load and the reflected impedance that's presented to the output tubes, you expect to see the speaker curve to be presented as the square of the turns ratio. This could easily be plotted as a simple relationship but would not account for the OT characteristics. This also ignores NFB, which acts to linearize the frequency response, so the more NFB the flatter the response, within the limits imposed by the amp.

A conjunctive filter can be quite good to take away peaky highs and level off the response.

[Helmholtz]

Freyette was saying that NFB (and maybe OT) can even affect your overall sound this way.

What is reflected to the power tubes is the impedance curve - not the frequency response.
Speaker frequency response is typically measured with a very low impedance (high damping factor) SS amp. As tube amps have higher output impedance, frequencies at higher impedance get emphasized.
Generally more NFB lowers the output impedance and thus allows to get closer to the published curves - but only in the linear range. As soon as the power tubes clip, NFB loses control and speaker frequency response changes.

And~ applying this to especially with two tubes pulled from a 100 watt amp, giant Hammond transformers, would you hear better response with the 4Ω speaker choice as there is higher Reference Efficiency? 2Ω combined parallel pair even better? From Traynor factory they used 100 watt iron so adding another cab might be a good idea. Would one half or double the FB resistor?

When you pull 2 tubes from a 100W amp, its nominal output impedance doubles.

Most quality OTs have a flat frequency response between maybe 40Hz and greater than 15kHz (sometimes up to 40kHz).

I can't see a difference in efficiency between different impedance versions from your graph. And speaker efficiency is not related to frequency response.

Increasing NFB can be done by lowering the value of series resistor. But more NFB might result in instability (HF oscillation).

[Guitarist]

I like those conjunctive filers too— also in one side of a Carvin rack amp that I removed the NFB in.
Hmmm, so a EV10ML flat-type speaker will ballast out those gains/frequencies. ..
Steve Fryette (of old VHT amps) has blogged closer to the page bottom: https://www.premierguitar.com/author...steven-fryette

And the concepts are getting clearer! Thank you

Also related: I have a 4x12 with a pair Peavey Blue Marvels (for now). Want to add a pair of ATC 12" with 86 efficiency vs BM' 96dB. The bass half won't be as loud as the other guitar speakers. So I like to insert a 25? Watt resistor before the 16Ω Peavey/guitar drivers. Any idea for a good value, 3Ω?s
Appreciated

[loudthud]

You can see a slight shift in the resonant frequency of the impedance curves. Higher impedance resonates slightly lower. This implies that the higher impedance voice coils weight more. This would affect efficiency, probably by a fraction of a dB.

Sometimes speaker manufacturers play a game where efficiency is related to some Voltage across the voice coil, not necessarily a given power regardless of impedance. Of course, power is hard to define when capacitance and inductance are part of the equation, so you really have to dig into any company's data sheets and measurement practices.

[pdf64]

...Steve Fryette (of old VHT amps) has blogged closer to the page bottom: https://www.premierguitar.com/author...steven-fryette

There’s a lot a articles there, please link to the relevant one.

...
Also related: I have a 4x12 with a pair Peavey Blue Marvels (for now). Want to add a pair of ATC 12" with 86 efficiency vs BM' 96dB. The bass half won't be as loud as the other guitar speakers. So I like to insert a 25? Watt resistor before the 16Ω Peavey/guitar drivers. Any idea for a good value, 3Ω?s
Appreciated

Consider a speaker’s impedance chart, eg nominal impedance is only achieved over the low mids, it spikes to maybe 10x nominal at bass resonance, and doubles beyond 2kHz. Hence a fixed resistor in series will create a frequency dependent potential divider.

[Helmholtz]

Hence a fixed resistor in series will create a frequency dependent potential divider.

Just like the amp's output impedance (which is essentially resistive) does. In fact a series resistor just adds to the source impedance as seen by the speaker.
Results will be somewhat augmented bass resonance and treble response.

[Guitarist]

Thanks gentlemen, that makes sense. If once the 300 Watters are installed I'll see how the pairs play together in XY. Might just upgrade the top slots and plywood over the lower two holes. Biamping might be the better scheme.