Phase Question

[Grees01]

Hello there.
Hopefully I didn't miss this in my reading thus far, but concerning "phase"? when wiring up power tubes to the OT in a push pull configuration, is there a way to tell before soldering up the leads to the tube sockets, which way is correct? Other than leaving the leads long and possibly switching the two leads if there a hum problem.

"On the Output transformer, if you are not sure which plate wire goes to which plate, leave the wires full length for now so you can reverse them later if they are not correct
Any wire that is not connected to the turret board (except pot wires) can be wired up at this point"

Excerpt taken from the infrastructure wiring portion of the installation instructions for Hoffman boards.
Thanks in advance for the help.
G.

[shooter]

unless your OT has phase markings and you have a scope to set the phase of the PI, just leave 'em long and swap as needed.

[pdf64]

unless your OT has phase markings and you have a scope to set the phase of the PI, just leave 'em long and swap as needed.

I think that should be an ‘or’ rather than an ‘and’ 
It’s certainly nice to know it’s right, rather than have to wait to check circuit behaviour (a positive feedback loop doesn’t necessarily result in oscillation).

[Grees01]

What typically happens if the leads are reversed? Is hum correct? And is this because this portion of the circuit is "out of phase"?
G.

[pdf64]

Oscillation, probably at full power, and / or an unusual tonal response.
The definitive test is to disable the feedback loop and note whether gain increases (NFB) or decreases (PFB).
FYI the topic here is signal polarity rather than phase, the latter being a frequently / time dependent characteristic.

[tubeswell]

…
FYI the topic here is signal polarity rather than phase, the latter being a frequently / time dependent characteristic.

^This^


(Signal phase in an amp circuit can vary with frequency - and it’s possible to have global NFB (inadvertently) configured with LF getting a bit of NFB and potentially enhancing the fidelity, but FB at HF being almost completely positive, resulting in very high-pitched or ultra-sonic oscillation, which requires a different ‘fix’ than merely swapping the OT winding ends around).

[PRR]

"....leave the wires full length for now......"

A very wise and experienced man wrote that. If you doubt him, do as you wish.

My experience is, unless you get in a rut of SAME circuits SAME layouts SAME parts, you can bash your head against the bench and still come up wrong half the time. Humans are fallible. Murphy works against us.

How is leaving both wires long (long enough, not necessary full factory length) going to slow your build?

[Grees01]

"....leave the wires full length for now......"

A very wise and experienced man wrote that. If you doubt him, do as you wish.

My experience is, unless you get in a rut of SAME circuits SAME layouts SAME parts, you can bash your head against the bench and still come up wrong half the time. Humans are fallible. Murphy works against us.

How is leaving both wires long (long enough, not necessary full factory length) going to slow your build?

Not at all, I'm not doubting anyone, just trying to learn, and I don't think leaving the wires long would necessarily slow a build down by anything more than maybe 5-15 minutes. I'm just trying to understand. If the wires are backwards, the likely outcome is that you would hear an oscillation? And curious as to why the right wire to the right plate can't be figured out. No disrespect or doubt intended.
G.

[pdf64]

As mentioned, the correct polarity can be determined.
However, to do so may req info / equipment / competence / time not available to the builder. And if a mistake is made and the wrong polarity is determined, the wires might be cut short. 

[HotBluePlates]

... If the wires are backwards, the likely outcome is that you would hear an oscillation? ...

Yes.

Negative feedback is instead positive feedback, because the polarity of the voltage on the feedback loop is opposite what it should be.  Most of the time you get a full-power HOWL, but sometimes the positive feedback leads to early (not-good) distortion, or a weird raspy sound to the treble.

... And curious as to why the right wire to the right plate can't be figured out. ...

On rare occasions, the transformer manufacturer includes "phasing dots" on the wiring diagram.  That is, when the wire with the dot on the primary has a positive-going voltage, the wire on the secondary with a dot also has a positive-going voltage.

You can figure out the phasing ahead of time for an unknown transformer, but you need a signal generator to drive the primary, a dummy load for the secondary, and a 2-channel oscilloscope to monitor the voltages on the primary & secondary at the same time.  You can then observe whether the voltage waveforms are same- or opposite-polarity.

Since that's a bunch of equipment & some setup time needed, most just leave the wiring long and a hand on the power switch at first startup, in case the polarity is wrong & the amp howls.  No equipment needed, and testing time is quick.  Plus a 50/50 chance you got it right with the first guess.

[Grees01]

Thank you HBP. That helps quite a bit. I appreciate the detailed explanation.
G.