Odd OT primary backwards wiring symptom

[DudeNiceAmp]

I finished a build last week. Fender deluxe reverb. Had an Alnico speaker in the combo. Amp sounded great upon first fire up. Thought I was lucky with the OT primaries first try. Those are usually 50/50. Wanted to try some other speakers today. As soon as I put a new one in, had the craziest low frequency feedback. Not necessarily the symptom I’m use to with backwards primaries. Tried another speaker. Same thing. Switched the primaries. All went away. Why did the Alnico speaker mask the problem?

[Willabe]

Why did the Alnico speaker mask the problem?

Some times the leads on the speakers are labeled backwards.

Take a 9v battery, and just click it, quickly, don't leave it connected, on the speaker leads, just to get the speaker cone to move. If their labeled correctly the cone will move forward with the batteries + going to the speakers +. Check all 3 speakers. 

[DudeNiceAmp]

Good theory. I checked all speakers. All correct.

[HotBluePlates]

... Amp sounded great upon first fire up. ... Wanted to try some other speakers today. As soon as I put a new one in, had the craziest low frequency feedback. Not necessarily the symptom I’m use to with backwards primaries. Tried another speaker. Same thing. Switched the primaries. All went away. Why did the Alnico speaker mask the problem?

"Impedance" is the vector-sum of Resistance and Reactance.

Reactance varies with frequency, and also causes phase-shift (which also varies with frequency).

It seems one speaker had enough phase-shift to cause the feedback to be marginally-stable, and the other had a phase shift that turned the feedback unstable & positive.  This is fairly common.

The trick is to always go for disconnecting the feedback.  If the unstable amp suddenly becomes stable, you know the polarity of the feedback is backwards.

A very few transformers have a winding issue such that they are unstable either way.  This happened to output transformers for a specific vintage Fender model for a couple years.  The only cure is a different OT.

[DudeNiceAmp]

Very good info HBP.

I have not thought about disconnecting the NFB. That seems like it would be a good habit to have moving forward on everything. First time using one of these model Alnicos. Very odd occurrence. Glad I decided to try some other speakers.

[Merlin]

As soon as I put a new one in, had the craziest low frequency feedback.

Pretty unusual to get *low* frequency oscillation; I presume the speaker resonant frequency occured in the most inconvenient spot. I wish speaker mfr's would publish the impedance curves of their speakers, then we could answer these sorts of questions more rationally!

[pdf64]

Just to expand, the gain of a pentode stage (ie the output stage here) is directly related to the load impedance.
So a very high peak impedance at bass resonance is more likely to result in enough gain in the positive feedback PFB loop to cause free running oscillation, most likely at the bass resonance frequency.
Whereas a speaker with a not so exaggerated bass resonance impedance peak (perhaps partly due to a lower strength magnetiv field?) may not have enough gain for the PFB loop to oscillate.
It would probably sound weird though, with a boomy bass etc.

Combine the above with the phase shift thing mentioned previously, and it's not unusual for an amp with issues to be stable into a dummy resistive load, but unstable with a real speaker / reactive load.

[DudeNiceAmp]

Pretty unusual to get *low* frequency oscillation

I thought the same thing. Not necessarily the symptom I am use to.