Help please--Phase problem with Subwoofer (1 Viewer)

[MartyP]

I am having a problem with a subwoofer. When I adjust the phase control (without volume from the LR mains, the normal output seems to drop near the middle setting and then increase back to normal as I keep turning. What would cause this?

 

[Philip>L]

Negative cancellation from reflections off the walls, floor and ceiling?

 

[Kevin C Brown]

Theoretically, the output level should not change at all by changing the phase knob *without* power to the fronts. (Can't be cancellation from elsewhere in the room, because the effect would be the same as you are changing the phase of the source and there's nothing interacting with it, if that makes sense. ) How exactly are you testing this? I.e., are you sure there's no signal going to the surrounds/rears/center that also might be interacting with the sub?

 

[Philip>L]

Changing the phase of the sub by itself will change where the maxima and minima are in the room, with or without reflections involved. That's the whole point of changing the phase, after all. So if you sit in front of the sub and rotate the phase knob it makes sense that the volume will change too.

 

[Kevin C Brown]

Not true. Phase is always relative to something else. When you change the phase of a sub with no other speaker functioning, all you are doing is adding delay to the signal.


The whole point of changing phase on a sub is to line up the wavelength peaks and valleys *to* those of the main speakers. Does not change how the sub by itself interacts with the room. But it does however change how the sub *and* main speakers interact together.

 

[Philip>L]

Yes, in relation to its source. By turning the phase knob you're affecting where the beginning of the wave starts (the zero point on an X-Y plot) in relation to the speaker.

Phase alignment is relative to another speaker. If the wave form didn't shift in space when you rotated the knob, then there would be no point to having it.

 

[Kevin C Brown]

The wave form shifts in *time* when the knob is rotated.

 

[Philip>L]

And space. Rotating the knob has the same effect as moving the speaker closer to or further away from you.

The volume change could also have something to do with the way in which the electronics play into the phase shift.

 

[Mitch_J]

Are you confusing "Phase" with "Polarity"? Or am I just confused now?

 

[ChuckRG]

Polarity is phase but at 0 or 180 degrees.

 

[GrahamT]

Where did you hear this? It doesn't make sense to me, could you point me in the right direction.

 

[Kevin C Brown]

A phase knob on the sub is simply this: analog delay. It inserts a small amount of delay into the signal such that, like I mentioned up above, that the peaks and valleys of the signal sent to the sub can all line up (become "time aligned") with the peaks and valleys in the signals sent to the main speakers.

It is partly correct that rotating the phase knob changes the *perceived* distance of the sub from you. Adding "distance" does the same thing as a delay. 1100 ft/sec. But. That distance is not real in that it does not change how the sub interacts with the room. The sub is in the corner (or wherever it is). That is a physical fact that can not changed by a phase knob.

The sinusoid still starts from 0, it is simply delayed. If the signal started at any other point than zero, you'd be cutting off part of it, and that is not what's happening.

 

[LanceJ]

Don't forget about the sound waves the sub just previously launched into the room: depending on how far they traveled before they bounced off the walls/floor/ceiling, when they return to the listener's listening position, they could very well be 30, 90, 180, 270, etc degrees out of phase with the wave that just emerged from the sub's woofer, i.e. the wave shot directly towards that listening position (man I wish a I could post a drawing of this because it would explain this much more easily!).

In other words:

* At the listening position when those two waves finally meet each other, if the bounced wave is at 180 degrees RELATIVE to the direct wave, they will cancel each other out >>> no bass, because +1 plus -1 = 0.

* If the bounced wave is at 90 degrees RELATIVE to the direct wave, there will only be a partial cancellation >>> causing the volume to just be reduced rather than be cancelled all together.

* 270 degrees of phase difference will result in the same amount of cancellation as 90 degrees (which is why the volume starts going back UP as the two waves' phase difference moves away from 180 degrees).

* And at 360 degrees? This would mean the waves are finally 100% lined up with each other (0 degrees is the same thing).

In reality it doesn't work out quite this neatly since there are multiple waves meeting from every which direction but you get the idea.

This is why I hope my next receiver has a distance setting for the subwoofer channel......unless the sub I own has a variable phase control like Marty's and not just a simple "0/180" switch. But in the end they end up doing the same thing.

Bass: audio's problem child.

LJ

 

[Philip>L]

Also true, but you are changing where the peaks and valleys fall in the room which affects the perceived volume at any given point as you rotate through the whole range of the knob.

 

[GrahamT]

Lance, you know that all his other speakers are disconnected right? It is only the sub.

 

[LanceJ]

Like I said, a picture would explain this much, much better! :frowning:

Yes, I am aware the sub is the only speaker operating in this situation--my above post was written with that condition in mind hence the reference to the bounced and direct waves.

What we are dealing with here is TIMING issues. As far as explaining the changing volume phenomena we are talking about now, it doesn't matter how the wave's timing is changed but just so they ARE changed when they finally meet.* The term "phase" in the situation being discussed right now is simply the technical term for timing issues and is used to numerically specify how much two or more waves are out of sync with each other.

* I tend to agree with Kevin that some sort of a delay system is used for a sub's continously variable phase control. Analog systems for that have been around for years (I think they are called "bucket brigade" systems). But as far as the simple "0/180" switch goes that is found on most subs, AFAIK that switch simply reverses the sub's polarity, the same as if you physically reversed the wires on the sub's woofer driver itself. Very cheap & easy to implement compared to the variable phase control.

LJ

 

[Kevin C Brown]

LJ's explanation is right on. But the interaction of the direct waves and bounced waves does not change based on the phase knob. *Both* signals are affected, hence the *interaction* is not changed.


False. It is timing alone. Nothing at all changes spatially.

The reason why a sub's volume changes with the phase knob *while the main speakers are also playing* is the constructive and deconstructive nature of the two signals as the phase of the sub is changed. Nada changes with just the sub alone.

I have a sub with a phase knob. I'll whip out my Radio Shack meter and a test tone disc tonight and report back.

 

[LanceJ]

Kevin: LanceJ made a logic boo boo. :b

I see what you mean by the direct and bounced waves being the same wherever the phase control is set. Because while they can be out-of-phase like I previously described, their phase relationship will not change since both waves are being adjusted at the same time.

(where's that melting-brain smiley face? )

Thanks!

But now I 've got another theory which literally just popped into my head as I wrote the above:

(Taking into consideration how large bass waves are, I think up to @20 feet long in our case.......) Since the location of the wave's peaks and valleys are physically being moved in space because of the delay system, could it be that since Marty isn't moving, it is the null and non-null(?) bass nodes that are moving towards and away from him, in turn causing the volume level at his point in space to be reduced and increased as he turns the knob???

LJ

 

[Kevin C Brown]

Lance:


Say it with me now: nothing is moving in space. The only change that is occurring is the delay in the signal sent to the sub. Time domain, not space. (Think about it this way. I initiate a signal to the sub at time = 0 sec. Now I repeat that at t = 360 sec. No change right? Now let's say I change that to t = 20 msecs. There cannot be any change. Only time is changing.)

Anyway. I tested this myself with my sub. Here's the particulars:

o Vandersteen V2W sub
o Has a phase knob, and
o an inverted input (basically, I can cover 0 to 360 deg.)
o analog Radio Shack meter on a tripod at the listening position
o all speakers unpowered, only the sub is powered

I have measured this sub in my own room. Flat to 21 Hz, and - 6dB at 20 Hz. (Radio Shack meter compensated.) Like any sub-room interaction, I have peaks and valleys in the response due to this interaction. (Sub in a corner.) I have my sub running through a BFD eq to flatten its response. This, I did by using my PC, the Radio Shack meter, and www.etfacoustic.com software. I have had up to 5 filters implemented, but like a lot of things in life, less is sometimes more, and I'm only using 1 filter right now. (I attenuate a 65 Hz peak.)

I also have a discrete test tone CD (Autosound 2000, test CD 101). Has discrete frequencies from 10 to 98 Hz. I specifically picked 30 Hz because it's comfortably below the lowest crossover I use (60 Hz), and because from the previous plots I've done, I know it's in a repeatibly flat regime.

So, what did I get?

I tested the folowing conditions:

o the phase knob goes from about 8 o'clock to about 4 o'clock
o I tested 9 o'clock, 12, and 3 o'clock
o I tested using the straight input and the inverted input

(My baseline setup is 3 o'clock, normal input.)

I set up the test so that I got 72 dB for my baseline setup. And ... that is exactly what I got at all the other test conditions. No change at all. I was actually a little bit surprised by this. I figured I'd get *some* difference in repeatibility at least, but nothing. 72 dB every time. And ... I have an analog meter so I can spot down to 1/4 dB differences.

Sometimes theory does match with reality. I do know that my phase knob functions, because if I use the inverted input, with sub and main speakers playing, the response drops by 7 dB at 60 Hz. And, I can see that 7 dB change with rotating the knob itself.

 

[Kevin C Brown]

Hey, no one ever asked me what I thought is going on with MartyP's setup.

1) All speakers are not unpowered, and he's still getting interaction between the sub and the speakers that are receiving power.

2) There is a dirty little secret of subs: there can be unintended interactions between the phase knob and low pass freq knob (if present, and if not defeatible). So I would suggest that if it's defeatible, to turn it off. If it's not defeatible, then adjust it to its highest setting. (This is not the case in my setup. The V2W is meant for home theater and the line level connection from a pre/pro. No low pass filter at all. No speaker level inputs that would need to be low passed.)

3) There is something wrong with either how the test is being done, or there's something wrong with the sub. (Discrete frequency? Has it been determined that the level does not change if the phase knob is *not* turned? I.e., with low freq white noise, we've all seen the Radio Shack meter jump around.)