Looks normal to me. You did lower the XO frequency and
the ribbon will receive more power at lower frequencies.
the ribbon will receive more power at lower frequencies.
Build it with the cheapest parts you can and then do a full range of tests outside and off the ground by about 8 feet. Models are often "off" by a bit when compared to measurements and usually need some refinement. Also do some listening tests to see if it has the in-room power response/tonal balance you are looking for.
I've been changing all of the part values to try to do better - what do you think of this, better?:
Attachments
Much better, 5 dB less tweeter spl at 2 kHz than the former version.
This is probably 8,53 deg vertically off, so closer to the tweeter axis,
the individual driver phases will be even closer together.
This is probably 8,53 deg vertically off, so closer to the tweeter axis,
the individual driver phases will be even closer together.
..and from a different point of view:
-worse.
Horizontal directivity will result in rather substantial suck-out around 1.5-2 kHz, and on top of that you have added-in some additional "lift" between 1.5 and 2.5 kHz. Result: listened to off-axis won't be linear, and listened to on axis will be fatiguing and "forward".
On the other hand non-linear distortion will be a lot better, but really - it's tough to hear any non-linear distortion unless you are playing them quite loud on average. (..Ex. I think I rarely listen to my speakers at much over 83db on average (at a little over 2 meters) with rare transient peaks over 6db.) It really depends on your own use.
Last edited:
That was at 2.86 degrees I forgot to mention that. The last couple I've shown are 2.86 because that's where the deepest null is when the tweeter polarity is reversed.
That last one uses a few extra parts, a third order on the woofer and a notch instead of the contour, in an attempt to not have such a deep dip at 1.5k.
I've struggled to keep the rise around 1.1 - 2.4k in check while crossing the tweeter higher, it's hard to do! 🙂
That last one uses a few extra parts, a third order on the woofer and a notch instead of the contour, in an attempt to not have such a deep dip at 1.5k.
I've struggled to keep the rise around 1.1 - 2.4k in check while crossing the tweeter higher, it's hard to do! 🙂
Have you tried the filter in post #134? Better phase matching at
the same angle with improved linearity, at least in the simulation.
I don't think you need a 3rd electrical on the woofer. It introduces
an extra phase shift unnecessary.
the same angle with improved linearity, at least in the simulation.
I don't think you need a 3rd electrical on the woofer. It introduces
an extra phase shift unnecessary.
Isn't it even better to have the vertical phase lines right on top of each other? Sorry I'm not an expert at reading the phase graph.
Not really. When you plot phase in a 2D plane, it just looks like it 'wraps'. When in fact, it is just needing to be plotted on a tube. The vertical lines in fact do not exist, but are the plots shifting from -180 degrees shift to 180 degrees shift- ie- the same phase point.
If the red tweeter phase and the blue woofer phase are not spread very far in terms of plot, not even looking at the 'wrapping', then they are in fairly close alignment.
If you have higher math knowledge, the derivative (calculus) of the freq response is the phase plot.
Later,
Wolf
Crossover Showdown
This may be a little bit insane, but it does show my commitment to getting the best out of these. 🙂 Here are the ten crossover designs I have seen in this thread. Each of the ten images shows the system response at 0 degrees, 2.86 degrees, 5.71 degrees and 8.53 degrees vertically, respectively down the page. Each page is numbered. What I'm asking is for you guys to all take a look, save them or view them full size and pick one, then tell me the number of your favorite crossover based on this data. I'll then take the winner by consensus and build it and post the info for that crossover in the original post. Thanks in advance!
This may be a little bit insane, but it does show my commitment to getting the best out of these. 🙂 Here are the ten crossover designs I have seen in this thread. Each of the ten images shows the system response at 0 degrees, 2.86 degrees, 5.71 degrees and 8.53 degrees vertically, respectively down the page. Each page is numbered. What I'm asking is for you guys to all take a look, save them or view them full size and pick one, then tell me the number of your favorite crossover based on this data. I'll then take the winner by consensus and build it and post the info for that crossover in the original post. Thanks in advance!
Attachments
Me personally? I'd do a high-pass (only) and measure the tweeter's with various power inputs looking at non-linear distortion. (..actually I'd probably model this first, but barring that..) Comparing 5, 6, and 9.
Update: New schematics with final values and parts list for crossover
Part PE # Qty Price ea. Total
Hammond 1591ESBK ABS Project Box Black 7.5" x 4.4" x 2.4" Part # 320-718 2 $7.46 $14.92
3.3 Ohm 5W Resistor Wire Wound 5% Tolerance Part # 015-3.3 2 $0.68 $1.36
12 Ohm 5W Resistor Wire Wound 5% Tolerance Part # 015-12 2 $0.68 $1.36
5.6 Ohm 10W Resistor Wire Wound 5% Tolerance Part # 016-5.6 2 $0.64 $1.28
10uF 100V Electrolytic Non-Polarized Crossover Capacitor Part # 027-340 2 $0.67 $1.34
22uF 100V Electrolytic Non-Polarized Crossover Capacitor Part # 027-348 2 $0.86 $1.72
Audyn Cap Q4 5.6uF 400V MKP Metalized Polypropylene Foil Crossover Capacitor Part # 027-115 2 $2.99 $5.98
Audyn Cap Q4 3.9uF 400V MKP Metalized Polypropylene Foil Crossover Capacitor Part # 027-113 2 $2.29 $4.58
Jantzen Audio 0.80mH 20 AWG Air Core Inductor Crossover Coil Part # 255-044 2 $7.28 $14.56
Jantzen Audio 0.80mH 18 AWG Air Core Inductor Crossover Coil Part # 255-242 2 $9.14 $18.28
Jantzen Audio 2.0mH 18 AWG Air Core Inductor Crossover Coil Part # 255-266 2 $14.12 $28.24
ERSE 0.22mH 18 AWG Perfect Layer Inductor Crossover Coil Part # 266-808 2 $5.39 $10.78
Parts Express Dual Binding Post 1 Red 1 Black Part # 090-475 2 $3.88 $7.76
Part PE # Qty Price ea. Total
Hammond 1591ESBK ABS Project Box Black 7.5" x 4.4" x 2.4" Part # 320-718 2 $7.46 $14.92
3.3 Ohm 5W Resistor Wire Wound 5% Tolerance Part # 015-3.3 2 $0.68 $1.36
12 Ohm 5W Resistor Wire Wound 5% Tolerance Part # 015-12 2 $0.68 $1.36
5.6 Ohm 10W Resistor Wire Wound 5% Tolerance Part # 016-5.6 2 $0.64 $1.28
10uF 100V Electrolytic Non-Polarized Crossover Capacitor Part # 027-340 2 $0.67 $1.34
22uF 100V Electrolytic Non-Polarized Crossover Capacitor Part # 027-348 2 $0.86 $1.72
Audyn Cap Q4 5.6uF 400V MKP Metalized Polypropylene Foil Crossover Capacitor Part # 027-115 2 $2.99 $5.98
Audyn Cap Q4 3.9uF 400V MKP Metalized Polypropylene Foil Crossover Capacitor Part # 027-113 2 $2.29 $4.58
Jantzen Audio 0.80mH 20 AWG Air Core Inductor Crossover Coil Part # 255-044 2 $7.28 $14.56
Jantzen Audio 0.80mH 18 AWG Air Core Inductor Crossover Coil Part # 255-242 2 $9.14 $18.28
Jantzen Audio 2.0mH 18 AWG Air Core Inductor Crossover Coil Part # 255-266 2 $14.12 $28.24
ERSE 0.22mH 18 AWG Perfect Layer Inductor Crossover Coil Part # 266-808 2 $5.39 $10.78
Parts Express Dual Binding Post 1 Red 1 Black Part # 090-475 2 $3.88 $7.76
Attachments
Last edited:
What's the project box for?
I would also question why you went for a Q4 across the woofer (good call!) but not in series with it? I see why the NPE for the shunt notch- that's fine.
A Q4 22 only runs you about $9 last I checked.
Later,
Wolf
The crossover is external so it goes in a small box. I'll show some pics when it's done.
The np cap is much smaller. The film caps get pretty silly in size at a certain point.
1.45" thick by 1.7 long to be exact but I'm seriously considering getting them if I can fit them, just wish I had ordered all together to save an extra shipping fee.
- Status
- Not open for further replies.