Tweet
I was playing with my dcx2496 with the deq as a rta. I have a coaxial speaker and used a log tone and a linear tone sweep to see how each driver responded individually and as a set. Basically I want a first order crossover. As a system, the flattest response was achieved by using a 6db low pass to the mids at 2k, and a 6db high pass to the HF driver at 4k. Does this make sense? I'm new to this, most crossover designs pick a point, not 2 different points. Thanks. Jim.
-
-
It makes sense to me
With a first order network, the drivers will be down 3dB at the xo point, and 90degr out of phase, thats in theory.
If you use a flat baffle, the drivers are not time alligned, and the drivers are out of phase before you connect the xo.
In your system it may be that your drivers are about 90 degr out of phase because of the baffle, with a first order network connected, the 90 degr phase shift in the network will make your drivers in phase again.
With drivers in phase, you need -6dB at the xo point to make them sum flat.
Thats why you have to spread the xo points.
Measure the drivers one by one, with the xo connected, and you probably will see that they are 6 dB down at the xo point.
Try to push the tweeters xo point even higher, to gain a BBC dip at 3-4khz. This will often sound very well, and your tweeter will be even better protected against the lower frequencies.
If it sounds good to your ears, be happy with it
-
Thanks for the explanation! It's been a great learning process with the dcx/deq combo, but I think I'd like to buy some real measuring devices. I'm thinking of the woofer tester pro. Can't find much reviews on it though. Can anyone comment on this device?Comment
-
You will get much more out of a decent microphone/pre-amp and one of the measurement software packages, either free or commercial. Since you already have the ability to create test crossovers "on the fly", and the ability to do FR correction as well, an ability to measure the result is what you need next.Originally posted by jimluuComment
-
How far do you want to take this?
If you want to do more than just experiment by ear, you'll be best served to do as Deward described. Nothing can replace measurements, but it takes some effort to get to that point, though IMO it's well worth it.Originally posted by jimluu
I disagree with some of the post by norcad. The basics on Fc and phase at the crossover are correct insofar as some of the theory is concerned, but the parts related to your situation are not, I believe. You don't have multiple drivers on a flat baffle, so there's little if any significant misalignment in a coaxial unit unless it's the kind with a tweeter stuck out on a post at the front. If it's a coax with the tweeter in the pole piece vent area, then alignment isn't much of an issue, if at all. My comments assume the latter.
As for the separation, your target is first order, therefore spreading has nothing to do with needing to be -6db since your target is BW1, -3db at Fc, in phase quadrature and can be connected in positive or negative polarity. With an (assumed) aligned coax unit, the polarity is immaterial for SPL summing of a BW1, since the target has phase quadrature throughout the crossover area and will be in quadrature with either connection. The overall group delay is altered, but I doubt that it's audible for this type of coax. There won't be a lobe issue.
You can't be precise about SPL levels without measurements, though you may get close. A likely reason for needing the spread as you have described would be that baffle step, part of diffraction, usually causes a peak in the response in and around 2K for the tweeter. It could easily be 2-3db. The upturn may start above 3K. The woofer will see little peaking due its directionality at that point. In spreading the Fc of each driver, you're likely attempting to compensate for this peaking, unless you've rolled it off in another manner to achieve the target highpass desired for the tweeter. One can only guess with the limited info available. There may also be woofer peaks/dips entering into it as well. Even the best coax units have some issues related to the nature of a coax.
Measurements will be a real benefit to you. There are quite a few systems available. Here's an old link that has some limited info, but with links.
Measurement systems info
DaveComment
-
Im sorry! I was answering without reading your post good enough. I didnt see that you are using an coaxial unit.
So Im totally agree with dlr, who disagree with me!
I also took it for granted that you actually was measuring the flat respons.
When tweaking by your ears, you often will get the BBC dip I mentioned. But like dlr says, this is just guessing.
Without any information on the driver and setup, its hard to give advice, but IMO 1.order network is in most cases, not the best way to go. And I have never seen or heard a coax that will do well with it.Comment
-
I actually have a mic and preamp already. What's attractive about the woofer tester pro is that it seems to be a relatively simple all in one solution. It's got its on mic, preamp, measurement software and crossover simulator. My other choice would be soundeasy. I'm trying to find some useful reviews of woofer tester pro if some one knows of any.Comment
-
When I looked around I could find little end user information posted about WT Pro. I imagine that's a function of so few units being sold.
I am surprised when I occasionally see plots that were made with WT Pro.
Their measurement mic is likely some variant of the ECM-8000 design. That's the only realistic option given the price of WT Pro and the price of real test mics ($1000 or more).
IB subwoofer FAQ page
"Complicated equipment and light reflectors and various other items of hardware are enough, to my mind, to prevent the birdie from coming out." ...... Henri Cartier-BressonComment
-
Then get SoundEasy. It's used by more people around here, and John K's tutorial, combined with the manual, will get you started quickly.Originally posted by jimluuComment
Comment