XO Modification Question

[jim1961]

(measure at the listening position with 1/24th oct smoothing, 250ms gate)

Green = all drivers +sub
Brown = sub +woofer
Purple = sub + mid + tweet



Crossover

Firstly, I have already modified C2011 to 130uf to try to extend the mid lower.


Question. To fix the small gap in the 150hz - 250hz area, I was thinking of XO the woofer higher. Instead of a 8.2mh coil, maybe a 6.8mh one (with corresponding different capacitor).

To make a long story short, the gap is primarily caused by the midrange (in my reading of the graphs). Go ahead and discount room, setup, or treatment changes for I have tried as many as is possible to limited avail (its primarily a room issue that I cant address). Is this the right approach? Is the dip worth fixing? Can I unwind the existing "c" coil some or do I need to purchase another?

Thoughts?

[JonMarsh]

would want to simulate this with at least some vaguely representative data before offering any opinion, much less real advice! Should be able to do that some time next week, I think...

[5th element]

Is this measurement an averaged response of the mic in various positions, or just in the one place? Sometimes you can pick a bad location to measure at, where the dip goes away if you move the mic a little. Certainly when talking about modal interactions, you cannot come to any conclusions with a single, fixed, mic position.

That said this could either be modal, floor bounce or a poorly integrated mid and bass driver.

If you take a number of measurements at different lateral off axis positions, this should let you see if it's modal or not. It if is the dip will go away, if it isn't, it will remain.

As the dip is still present, even with the woofers turned off, it is unlikely to be poor driver integration. Floor bounce is just physics screwing with your perfectly flat frequency response. You could measure again, with the woofer off, at different microphone heights, or try measuring closer to the cabinet. This should change the frequency that floor bounce occurs at and will let you see if it's likely to be the culprit. The dip also appears to be missing with the mid and tweets turned off, which points towards floor bounce possibly being the issue.

The only way to fix this is to cross over higher up between the mid and the bass driver.

[JonMarsh]

Measurements of the individual drivers on the cabinet would indicate if this is the case. Similarly, if drivers which measure well in their cabinet location don't show this dip with the crossover, then it's likely to be floor bounce issues... will take some sleuthing.

[jim1961]

Is this measurement an averaged response of the mic in various positions, or just in the one place? Sometimes you can pick a bad location to measure at, where the dip goes away if you move the mic a little. Certainly when talking about modal interactions, you cannot come to any conclusions with a single, fixed, mic position.

That said this could either be modal, floor bounce or a poorly integrated mid and bass driver.

If you take a number of measurements at different lateral off axis positions, this should let you see if it's modal or not. It if is the dip will go away, if it isn't, it will remain.

As the dip is still present, even with the woofers turned off, it is unlikely to be poor driver integration. Floor bounce is just physics screwing with your perfectly flat frequency response. You could measure again, with the woofer off, at different microphone heights, or try measuring closer to the cabinet. This should change the frequency that floor bounce occurs at and will let you see if it's likely to be the culprit. The dip also appears to be missing with the mid and tweets turned off, which points towards floor bounce possibly being the issue.

The only way to fix this is to cross over higher up between the mid and the bass driver.

The dip moves in depth and in frequency as you take readings in different places. My avatar is my speaker, so you can see the mid is up pretty high. When I put OC703 on the floor to lessen the bounce the dip actually gets worse believe it or not (overall). More specifically, the woofer bounce seems to help in the frequency range in question, while the primary culprit is the midrange floor bounce. When you move the mic towards the speaker, the dip progressively disappears.

The curves I have shown are a single location. The mic tip where ones right ear opening is measuring the right speaker.

I agree moving the XO point higher seems the best direction to go given, as you pointed out, the woofer is unaffected. Looks like 40hz or 50hz higher would do (220hz to 270hz or so).

[jim1961]

Measurements of the individual drivers on the cabinet would indicate if this is the case. Similarly, if drivers which measure well in their cabinet location don't show this dip with the crossover, then it's likely to be floor bounce issues... will take some sleuthing.

The dip isnt there at near field measurements (1M)

[jim1961]

Lets look at the notes within the frequency dip reference to 69.5db:

A#3 233hz -2.5db
A3 220hz -5.5db
G#3 208hz -5db
G3 196hz -3.5db
F#3 185hz -3db
F3 175hz -2.3db

In essence, there are (2) notes (A3 & G#3) that are the most need of fixing. That said, I am not going to spend a fortune to fix this problem. Buying (2) new "C" coils, having looked up their price ($53 x2) is too much. What I am prepared to do is unwind the 8.2mh coils already in the XO. BTW, is it tricky to get the plastic surround on those coils off and then back on again in order to unwind them?

In the mean time, I will look for ways to absorb or redirect the midrange bounce. But I am not too hopeful given I have been trying to do this for a year now.

[jim1961]

Tried something new. Got it from the green line to the red one. A modest difference, but now the two "bad" notes are:

A3 220hz -4.4db
G#3 208hz -3.2db

[Jonasz]

Can you actually hear the dip or is it just a measurement-ocd situation?

What happens if you put a bunch of bedcovers on the floor where the floorbounce should be? Does the dip change? A table placed at the same place should also change the dip, if it is a floorreflexion that is.

[jim1961]

Can you actually hear the dip or is it just a measurement-ocd situation?

What happens if you put a bunch of bedcovers on the floor where the floorbounce should be? Does the dip change? A table placed at the same place should also change the dip, if it is a floorreflexion that is.

Mostly measurement OCD :B But it does seem at times, especially playing classical, that what Jon calls the "power spectrum" seems like it should be a bit more robust.

I already have 5 layers of rug on the floor. Additional flooring absorption makes it worse (a previous post explains this more thoroughly)