X-over design (Formerly Help with LEAP sim)

[Forte_II]

Chris, Yes the bass bins will be crossed actively. With a cross over point around 280hz. To a 10" Daton RS woofer.in a sealed box.
So how much BSC?

[cjd]

I have all these good intentions and sometimes they take a while to play out. If max sent files I didn't get any, but that's not a show-stopper. Just a big time-saver.

To put what we see as wrong into something your ears might pick up on - there are a few possibilities.

Most obviously, being partially out of phase, as you move off-axis you may hear a lot of the key determining sounds diminish - suck-out through the crossover frequency. With speech that may translate to a lack of intelligibility. You may also lose character on a lot of instruments. It is actually possible that the opposite will happen: output will increase off-axis in this range. The result will be added harshness and over-emphasis.

Second, you may find that it will smear imaging/soundstage - exact placement of sound depends on phase relationships and reflections - subtle timing differences in the same basic sound arriving at your ear relative to, well... everything around you. So your ears may hear a slightly delayed signal from the tweeter (or midrange, depending) in that critical region causing your brain to process its place in space differently. Or perhaps just making it ambiguous.

I'll have to check into the crossover you have (I assume you're using that to cross the sub in) to see if it does uneven slopes. You'll also need to consider phase relationship of the woofer (since you're crossing it way above sub duty, we can call it a woofer, right?).

In fact, you should be able to play with phase already, if the crossover supports it. See if you can hear differences as you change the phase on one driver.

C

[Forte_II]

The Behrenger I have doesn't do anything with phase other than a 180 degree flip.
The Datons are true woofers. they won't go below 80Hz. Below 80Hz I will have a LLT Tempest X.
Edit:
The behringer is a CX2301 2 way active with 24db. L-R alignment.
This will run the MTM and w. The reciever will cross the mains and sub at 80hz. I will be useing a Elemental Design PQ2 sub eq/hp filter.

[Maximiliano]

Chris, Yes the bass bins will be crossed actively. With a cross over point around 280hz. To a 10" Daton RS woofer.in a sealed box.
So how much BSC?

If you use something like my xo and cross it with your bass bin at 280 Hz, you won't need more BSC on the current 2-way. With a suitable active xo setup, they will integrate well.

Max

[cjd]

If you use something like my xo and cross it with your bass bin at 280 Hz, you won't need more BSC on the current 2-way. With a suitable active xo setup, they will integrate well.

Max

That's got what, a couple dB? Then I'd agree with this.

C

[Forte_II]

Thanks, Max.
What do you guys think of raising the x-over point? Considering the FR of the Vifa and the Daton RS 10 what if anything would improve by raiseing the x-over point? As I understand it: 300-3000Hz. is the criticle range. But I read another opinion (Earl Geddes( that 1k to 10k is the most important.

[cjd]

Proper implementation matters more. Above 3k won't happen with these drivers well IMO. 2500 is probably where I'd top out and I'd very probably push it a good deal lower considering center to center issues, the ability of the tweeter to play lower, and the breakup artifacts in the mids.

800Hz and up is where most of the important cues we get (spatial, speech) exist - 80-800Hz is where a lot of the core fundamentals in our music exist, particularly in speech. Of course, some things cover most all that range (a violin, for example, is 200Hz-3kHz give or take).

It gets crazy when you have a room full of zealots discussing the various compromises in audio, and which ones each of us choose.

C

C

[Forte_II]

Oh, I wasn't thinking of running the mids that high! I chose the 280 x over point based on the first premise above. I guess I can play with the freq. and see what happens.

[Maximiliano]

I think 300 Hz xo is about right. If you move that down, say, to 200 Hz, then you'll have lower distortions from the RS 10" but higher distortions from the PL14's. If you move that up to 400 Hz, then you'll probably have lower distortions from the PL14's but higher distortions from the RS 10". But this is just an educated guess. We can't be sure until we see actual distortion measurements.

[Maximiliano]

If max sent files I didn't get any, but that's not a show-stopper. Just a big time-saver.

Chris, I did send the files. Check out your emails, perhaps in a spam mail folder?

[cjd]

sigh. if it got filtered, it wasn't on my client side. I'll check the server.

[Forte_II]

Hi, I got a FR file of my speaker with the Madisound designed xover...tell me what you all think. I think it is just as Chris and Max predicted... Lumpy response due to phase issues. But it doesnt seem to drop in the midrange as much as predicted. Definintly need some BSC. :P

[Forte_II]

BSC???d some help with a formula.
0.18 is baffle width in meters.
L1 = .18 x 4 / 1.021 = 1 / 1.021
how do I resolve this? I get the first part... it equals 0.705 But what is with the last part = 1 / 1.021?

with a 8.315" (0.18 meter) baffle I get into full space at a f3 of 638hz. right?
f(3) = 115/0.18

with the above formula I need a 4 ohm resister in parael with a .7mh inductor... right I really don't get the last part of that formula!

[Forte_II]

Max would you mind including a zobel for the woofers? [QUOTE]
Can someone write out the signal path for Max's crossover, so that you can assemble it without a picture or scematic?

[Forte_II]

Someone please help these things sound lame with the madisound x overs! Just like Cris and Max suggested... bad imaging and too bright as well as that special something they had with the active x-over... This design has way more to offer than the xovers are giving.
Thanks

[Maximiliano]

Forte,

I think I can do a little better than I did before. I'll do another sim for you today.

Max

[Maximiliano]

Okay, I finished the crossover work. Unlike my previous version, this xover uses full, 6 dB baffle step compensation. You don't need to add a separate BSC circuit. The xover point is about 2.25 kHz. In the previous version, I used an LCR series notch filter connected in parallel with the XT25. But this time I removed it. With 2.25 kHz LR4 slopes, I'm convinced that this resonance suppression is not necessary. A Zobel is not necessary, either, in the woofer network. Many people have a misunderstanding that a Zobel circuit is always needed when the woofer has a bit rising impedance. That's not true. A Zobel was needed in old days when many people had to rely on textbook formulas to select xover components. This is because the formulas worked better with flattened woofer impedance. But with today's CAD available to people, a Zobel is not necessary in almost all cases.

Now I'll describe the designed network for you. The low pass section uses a 2nd order electrical filter to achieve a LR4 acoustic rolloff. The primary series inductance is 1.2 mH with DCR between .2 and .3 ohm. The shunt cap is 16 uF. That's it for the woofer network. The high pass section uses a 3rd order electrical with an L-pad. The primary series cap is 10 uF, and the shunt coil is 0.2 mH (DCR around .2 ohm), and the secondary series cap is 25 uF. The L-pad is located between the filter and the tweeter. The series resistance is 1 ohm (you can use two 2 ohm resistors in parallel if you can't find a 1 ohm one), and the resistance in parallel with the tweeter is 10 ohms.

Attached is my simulation result for other people.

Let me know if you have any questions.

Max

[Dennis H]

There you go, Forte. Max's simulation looks great. Build it and let us know how it sounds.

[Forte_II]

Thanks Max!

[Forte_II]

Just to make sure I got it...

+ = positive signal path.
- = ground signal path.

+1.2mh+ >> +16uF- >> +woofer-
+10 F+ >> +0.2 mH- >> +25 uF+ >> +1 ohm+ >> +10 ohms- >> +tweeter-

Pretty simple really. Cool
How do I adjust BSC?
How do I treat the grounds? seperate legs through the network or can I just use 1 ground leg for both HP and LP?

+ HP +
- -HP,LP-
+LP+
or
+hp+
-hp-
+lp+
-lp-

[cjd]

Adjusting BSC is not done independently of the rest of the circuit. When the measurements include baffle "loss" as any live measurements will do, it's inherent in the design. This should be a very comfortable and pleasing sounding setup.

You can use one ground (-) for both HP and LP but most of us like to build each network out on a separate board. If you use one, I would imagine it to be + HP - LP +

I don't see if the tweeter is wired in phase or out, but that won't change how the crossover gets wired up at all.

Max: A zobel is another tool in the toolbox, and one I find quite useful often. Not required, but I usually find them on mid networks. When I go all out, elsewhere. Thanks for stepping up.

I know I was going to help out here and I have not. I apologize for that. Sometimes I bite off more than I can chew and, while I could have fit this in, I did not. Ahh well. I'll learn and not volunteer so freely where I'm not sure I'll have the energy.

C

[Maximiliano]

You can use just one ground leg unless you want to biamp them. Both drivers should be connected "in phase" because this is a LR4 design. So, your signal flow schematic is correct.

You can adjust BSC---I think in this case of full BSC you may want to reduce it---by unwinding the primary series inductor in the woofer net and adjusting the tweeter L-pad accordingly. But try this full-BSC version first. If you need to reduce BSC, let's talk about it later. Another possibility is that you may need to simply adjust the tweeter level, not the BSC. In this case, the tweak will be easier.

cjd: I guess you mean a Zobel-like CR circuit. Of course, it's quite useful as sort of dampened shunt capacitance. I view this as one of many 'response shaping' tools, not as a traditional Zobel.

Max

[Maximiliano]

BTW, Forte, you haven't changed anything about the baffle and driver layout from what you told us before. The same MTM on the same sized baffle. Correct?

[Forte_II]

Max, Same setup. Finally got a finish on them...Oak on top and three sides. satin black paint on the baffle. People say they look good...LOL Who knows they may just being nice. :P
Cris, don't worry about it. It is nice that people are willing to help at all.
Thanks.

[Maximiliano]

Max, Same setup.

Good. That's what I assumed.

[Forte_II]

I have the parts ordered... How do I combine caps to get a specific value? say a 1uf and a 15uf to make 16uf...do I paraelle them or put them in series?
24.7 close enough to 25uf? If I need to add the .3uf would a higher end cap make a difference? I have read where people use a small value high end cap as a "bypass".

[cjd]

parallel to add value up, yup.

24.7uF is close enough to 25uF

I think Jon and Thomas have found you really need to go half value or greater if paralleling "high end" caps.

C

[Forte_II]

Thanks Cris.

[Maximiliano]

Forte,

Where do you shop for xover parts? What kind of caps did you purchase?

[Forte_II]

Parts Express.
I got Jansen CrossCaps, inductors (15ga. and 18ga. to get the right dcr) and Mils 12 watt resistors.

QTY PartNumber Product Price Ext. Price
1 370-096 Kester Electronic Solder 60/40 .040" .5 oz. Tube $2.95 $2.95
1 370-340 Variable Wattage Soldering Iron 0-50 Watts $13.69 $13.69
2 005-1 Mills 1 Ohm 12W Non-Inductive Resistor $3.50 $7.00
2 005-10 Mills 10 Ohm 12W Non-Inductive Resistor $3.50 $7.00
2 255-210 Jantzen 0.20mH 18 AWG Air Core Inductor $3.26 $6.52
2 255-424 Jantzen 1.2mH 15 AWG Air Core Inductor $15.48 $30.96
2 027-910 Jantzen 1.0uF 400V Crosscap Capacitor $1.17 $2.34
2 027-918 Jantzen 2.7uF 400V Crosscap Capacitor $1.87 $3.74
2 027-932 Jantzen 10uF 400V Crosscap Capacitor $4.44 $8.88
2 027-936 Jantzen 15uF 400V Crosscap Capacitor $5.11 $10.22
2 027-940 Jantzen 22uF 400V Crosscap Capacitor $6.07 $12.14

[Maximiliano]

Nice choices! Jantzen Crosscap is my favorite. Using 22 + 2.7 uF for 25 uF and 15 + 1 uF for 16 uF, connected in parallel, will have no problem.

[Forte_II]

Hey, I am about to put these together... Does it matter how you wire the inductors? Should the signal start on the outer wire and come out the inner wire or does it matter? Do caps have a in and out?
On the L-Pad... The 1ohm resister is in series before the parael 10 ohm. correct?
Thanks

[ThomasW]

None of the components in your crossover are polarized, so it doesn't matter which lead is hooked up to the hot or ground.

The 1 ohm is in series and 10 ohms is the shunt (parallel)

[Forte_II]

Hey I got the xovers built! They sound better than the Madisound...But They have too much BSC. How do I change this from the full 6db BSC to 3db or so? If I change the series resistance on the L-pad to .5 and unwind the primary inductor some what would be the result? With a 15ga 1.2mh inductor how many turns should I take off at a time?
Thanks Max and everyone for the help.

[ThomasW]

Color me confused.....

http://www.avsforum.com/avs-vb/showthread.php?t=966913

[Forte_II]

Confused how? That a week ago I said they sounded great :P or the second tweeter experiment? that was just with the exovers cobbled together. I screwed up somehow when I made circuit boards from blanks. After my son and I soldered the components to the boards I couldn't get a signal all the way through so I had to redo them. As for the other post I am just experimenting.

I think I figured out how to calculate how many turns to remove to get a given value on a inductor. for example: to change the 15ga 1.2mh inductor to a 1.0mh I have to take out 7 turns. ... Can someone tell me what value to shoot for to get 3db of BSC?

[cjd]

You probably can't reduce BSC just by unwinding the inductor... hard to be sure.

C

[Forte_II]

I was told that useing less padding on the tweeter and reducing the primary inductor on the woofers would reduce BSC.

[Dennis H]

Forte, reducing BSC 3 dB isn't a simple change. It has to be modelled in software to make sure the woofer and tweeter blend correctly -- phase and all that. Since Maximiliano designed the crossover you are using, you'll probably have to wait for him to tell you what is needed for less BSC if he's still around and has time to model the change.

[Forte_II]

I talked to Max and he is modeling it for me. I hope it doesn't require too many new parts...

[Forte_II]

OK. I finally got what I was looking for! Thanks to Max and everybody that helped! I went through several adjustments then figured out my son and I cooked some caps... So I got new caps and then built a solderless peg board solution. Before it was done my granddaughter thought the nice little dust cap in the center of each woofer looked better pushed in!!! I tried to push them back out with a dowl and some compressed air. ;( Not much of a improvement...So I thought I might as well do some experimenting... Now I have a set of Open Baffle speakers with dented cones that just sound awesome. perhaps the shape of the center cone is not as criticle as I thought. And with theTempest X in a huge sealed enclosure these sound awesome. The Tempest is just off center and a foot behind the mains I can cross at 120 without any issues. I love the way the tempest can go low and still sound good above 120hz! I reworked the xovers. LP is now .92 mh series and 20 uf shunted. HP is 10 uf (was 11 at one point( series and a .2 mh (copper foil) shunted then a 25 uf series cap. The L-pad was removed. I am very impressed so far. I am going to do some testing tomorrow and post up some measurements. and some pics of the finished speaker. I am very happy .

[Dave Bullet]

So I got new caps and then built a solderless peg board solution. Before it was done my granddaughter thought the nice little dust cap in the center of each woofer looked better pushed in.

Little kids eh? - we love them

Congrats on your new crossovers.

Try a vacuum cleaner to suck out the dustcaps, preferably one you can set the power level on. Haven't read the whole thread so don't know the size of the woofers / dustcaps. If dustcaps are larger than the vacuum nozzle, use a kitchen funnel or similar and cup your hands around it (or tape) for a seal.

David.

[Forte_II]

Hi Dave. They are 5.5" Vifa treated paper cone woofers, they have good response out to 6k and I think the shape of the dust cover has something to do with that. They are fairly hard and once dented never come back out quite the same.I will keep the Vac trick in mind though...