2-way OB crossing to woofers question:

0 Members and 1 Guest are viewing this topic. Read 11824 times.

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #20 on: 4 Apr 2010, 01:32 am »
Despite neobop's apprehension, it's pretty straightforward once you have a bit of knowledge about the drivers in question.

The series attenuation makes our 2ohm helper woofer a 6ohm driver. The ToneTubby has an Fs around 90hz and the 4x Kappa-12A have a resonant frequency around 45hz. As the Eminence drivers are rolling on, the ToneTubby is going high-Z. As we move lower in freq. the ToneTubby approaches Re, but the 4x Eminence is going high-Z.


Here's a simplified network:


Here's the resulting impedence:


Here are the individual curves (ToneTubby=Blue, 4x Kappa-12A=Red) overlaid:



The series attenuation makes a voltage divider with the driver's impedence. So, lot's of attenuation when impedence is 2ohms, not much attenuation at the resonant frequency. You could also look at it as transforming a low-Qts driver with requisite big honkin' magnet and compliance into a high-Qts driver. Here's the transfer function for the series attenuation + series inductor for the 4x Eminence Kappa-12A:


Would this work with any random group of drivers thrown on a baffle? Certainly not. But that doesn't mean it can't work. And, I'll suggest, they're awesome from 50hz and upward. Sensitivity around 98dB/W, it's a piece of cake to drive -- my 7W Class A PP 2A3 is overkill.
 
In case anyone wants to know what they look like:


John

neobop

  • Full Member
  • Posts: 3448
  • BIRD LIVES
Re: 2-way OB crossing to woofers question:
« Reply #21 on: 4 Apr 2010, 12:53 pm »
Apprehension is about attempting that WITHOUT a network.

If this is still contemplated as a full range OB, I'm curious about the integration with the Pioneers - whether they'll wind up doubling on bass, or if a high pass network will be necessary. I'm getting ready to build a couple of OB spks and some of this is new territory for me.

panomaniac

Re: 2-way OB crossing to woofers question:
« Reply #22 on: 4 Apr 2010, 02:13 pm »
Ahhhh....  Now I see!  In you first schematic it's Ra3 that's doing it.  Is that the DCR of the 12mH coil?

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #23 on: 4 Apr 2010, 03:19 pm »
You got it, Pano.

No, it's a physical resistor. I'm using the Erse SuperQ steel core inductors, and DCR is low -- around .4ohms. It's a little more flexible arrangement in terms of dialing in series resistance.

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #24 on: 4 Apr 2010, 04:26 pm »
That is really clever. So, for me it will be vital not to try to bring the 15's in too high. I am beginning to see now that building multi-way speakers is completely a shot in the dark without measuring capability.

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #25 on: 4 Apr 2010, 04:44 pm »
Back on topic...

Absolutely, Erik. There's no way I would have gotten to where I did without some sort of measuring gear.

For you, if the Pioneers are 103dB/W, I don't think you have any alternative to biamping. No hyperbole, it really would take a wall of 15" drivers to keep up @ 40hz.

If they were nearer 94-95dB, on a really big baffle because you want the coax covering as much ground as possible/ rolling on the helper woofers as low as possible, you could maybe do 2x Eminence Delta-15A with a 2ohm series resistor and a big choke. But that's a WAG and really you'd need to take a leap -- order the drivers, build the baffle, then try to make them work -- if it were me I'd just build one channel as a proof of concept. I don't have any of the drivers in question in house to measure, so I really couldn't assign a probability of success. But I wouldn't say that there's no chance of it working.

John Busch, Panomanic's cohort, is the master when it comes to this stuff. Maybe Mike would be kind enough to point John in the direction of this thread...

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #26 on: 4 Apr 2010, 10:06 pm »
Well, I had the silly idea of using 2 pair of massive alnico Pioneer 15"s that I have. The boxes they came from are rated 99dB/W and I thought that they could possibly match, not taking into account loss below Fequal.

But it is looking very much like I have two options right now due to the efficiency of my coaxes:

1. Biamping with active XO.(John, what did you end up doing for active XO?) I have a pro amp that I could do this with but wonder if perhaps something like a BASH amp would provide better cone control. Biamping would also allow for using Graham Maynard's TBass circuit, which has intrigued me for some time now.

2. Integrate a sealed sub @ roughly 80Hz. This looks the easiest and most elegant. And it fits with what my friend Dmason has been urging. The thing is, I really really like the quality of OB bass. But Dan hasn't been wrong in anything he's recommended to me thus far. I already feel like I have a King's system on a Pauper's budget.

BTW, with just my 12"s I'm getting some good reproduction of bass fundamentals. Cleaner and more defined bass than I had with a middling quality JBL 12" powered sub, and just as deep. And fewer room modes. It mystifies me, as the cones are only 4" from one baffle edge.

(I actually have 3 pair of the 15's, only one pair short of having 4 per side to match the 12's. What are the chances that my wife would approve of that baffle monstrosity coming into the house from my shop??????? I can only laugh at the thought.(But the potential for chest slamming bass actually has me considering it. I must be nuts))

Erik

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #27 on: 4 Apr 2010, 11:40 pm »
If you already have the drivers and a spare amp, I'd think almost anything is worth a try. You'd only be out the cost of wood and glue plus some of your time. Just for giggles, I think you have to try 3x 15" per side.

I built my own active crossover. It's volume control -> open baffle compensation -> 2nd-order lowpass -> 2nd-order highpass at ~28Hz with variable Q -> phase correction. There's a bit about it here, under the 20 April update.

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #28 on: 5 Apr 2010, 12:28 am »
John,
You are right. Our ancestors were not fearful men. I'll try it. Can you imagine how big these baffles are going to be? So, straight baffle or H-Frame, since matching efficiency isn't going to matter now? And John, where did you begin to learn about all the stuff you've been building? Seems pretty daunting to me, albeit fascinating. You seem to have a good handle on the big picture stuff that makes this audio thing fit together and work.

There will be a bit of a hiatus before any build updates, of course. And I've promised my wife I will convert half of our family room into a master bedroom before the baby comes(one month away). So, I'm going to be pretty busy. And incidentally, that same family room used to be my listening room. Not sure what is going to happen now.

Erik

roscoeiii

Re: 2-way OB crossing to woofers question:
« Reply #29 on: 5 Apr 2010, 12:48 am »
Viridian,

What was the reasoning behind the choice of the 80Hz crossover point recommended by your friend?

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #30 on: 5 Apr 2010, 12:55 am »
Hi,
Well, it could be higher but I just don't want the sub to be audibly locatable. That was the only reasoning. I understand that it can go higher without too much problem but the 80Hz is just a round guess.

Is that what you are referring to?

Erik

Viridian,

What was the reasoning behind the choice of the 80Hz crossover point recommended by your friend?

roscoeiii

Re: 2-way OB crossing to woofers question:
« Reply #31 on: 5 Apr 2010, 01:00 am »
Yeah that's what I'm referring to. But I was actually wondering about why 80Hz and not a bit lower than that. A lot of crossovers give high pass options of 80 and 50 Hz, and a lot of passive high pass filters are at 50 & 80 Hz.

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #32 on: 5 Apr 2010, 01:19 am »
Sure, but this is all still just theory at this point. One major factor though is that with OB, as frequency goes below Fequal it takes a huge amount of EQ/power in order to keep response flat. My initial purpose here was to relieve my main drivers of bass.

Erik

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #33 on: 5 Apr 2010, 01:21 am »
First off, congratulations on the upcoming addition.

My preference would be a flat baffle with side supports just big enough to provide some stability, kind of like the JELabs open baffles.

I'm as hooked on the DIY aspect of audio as I am on listening, so it's pretty natural for me to just kind of surf audio forums looking for projects that might be interesting or educational. Right now, the only thing in my system I didn't make myself is my laptop. I do have a Crown K1 that comes out every once in a while, but even that will soon be replaced (hopefully) with an ESP P101, complete with variable output impedence. My belief is that there's nothing I'm thinking of doing that someone else hasn't done already, done a bang-up job to boot, and posted about in some forum somewhere...

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #34 on: 5 Apr 2010, 03:16 am »
Thanks John.

I once tried a full-sized JELabs baffle and then a smaller one. The side wings don't make much of a u-frame really, but I did find that they affected the sound more than I liked. MJK and others say that a u-frame gives no cavity resonance, Ok, but it does something. I've cut up plenty of plywood trying to find a shape that works well. Driver mounting is equally important. I ended up going with an L-frame baffle which has worked very well. But for the frequencies we are talking about I doubt that it will matter much and will give better low end extension.

Actually, my baffles won't be much bigger than your eminence woofer ones. I noticed that you extended the baffle on one side. Why?

Erik

rjbond3rd

Re: 2-way OB crossing to woofers question:
« Reply #35 on: 5 Apr 2010, 01:25 pm »
Hi Viridian, pardon my jumping in, but do you recall where MJK said that a U-frame has no cavity resonance?  Thank you in advance.

nullspace

Re: 2-way OB crossing to woofers question:
« Reply #36 on: 5 Apr 2010, 01:54 pm »
Actually, my baffles won't be much bigger than your eminence woofer ones. I noticed that you extended the baffle on one side. Why?

The offset helps smooth the ripples from front-to-back cancellation in the frequency response for the ToneTubby. I used MJK's MathCad worksheets to simulate the driver's response on that size baffle and pin down a more optimal location.

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #37 on: 5 Apr 2010, 02:28 pm »
Makes sense, I thought that it had something to do with the woofers. Offset is good.

Erik

The offset helps smooth the ripples from front-to-back cancellation in the frequency response for the ToneTubby. I used MJK's MathCad worksheets to simulate the driver's response on that size baffle and pin down a more optimal location.

John

Viridian

Re: 2-way OB crossing to woofers question:
« Reply #38 on: 5 Apr 2010, 02:30 pm »
Hi,
Thanks for jogging my memory. As I was driving in to work I remembered that what was mentioned was that for the intended frequencies it wasn't a problem. Above that it does become an issue as far as I recall. All depends on how deep the uframe and all is though. It might account for why I didn't like the supporting wings on the JELabs baffle in fullrange though. I don't remember where it was said. You should look over the quarterwave site to see.

Erik

Hi Viridian, pardon my jumping in, but do you recall where MJK said that a U-frame has no cavity resonance?  Thank you in advance.

rjbond3rd

Re: 2-way OB crossing to woofers question:
« Reply #39 on: 5 Apr 2010, 03:40 pm »
Hi Erik, okay your comment threw me because according to MJK, a U-frame does indeed have cavity resonance because it is a short TL.  (It would also tend to act as a lowpass.)  So the depth of the U-frame should be based on a quarter wavelength to shape the output.  This has to be modeled though, in my limited experience.

Cool project by the way!