Who wants xover design practice?

[cobbpa]

I know it's a holiday weekend, and it really isn't urgent, but I have a combination of drivers that I would like to make into a pair of small towers or bookshelf speakers. I don't know how to use Speaker Workshop (I'm trying, but not getting very far..any kind of great tutorial around?) and I am guessing some people may want to toy around with it (maybe?). So this is an open invite to anyone who wants some practice and playtime with crossover design. I'm going to keep trying to learn, but if anyone wants to suggest a layout, I'd love it :T

Like I said, near-wall placement, would probably like to keep the total width (including roundover) under 10 inches. Other dimensions are totally open. These will be used solely for music, not really heavy listening..I'd be happy with extension somewhere into the 40's. No sub will be used. Doesn't need to be super loud. I'd like it to be a fairly budget-oriented endeavor; more specifically, if I could do this for < $50 per crossover, I'd be ecstatic. I do realize that isn't very realistic, but I'm not expecting outstanding quality, so maybe it's possible.

Parts:

Northcreek D25 tweeter

TangBand W4-616s mids

Dayton DA-175 woofers

[jonathanb3478]

I'd like it to be a fairly budget-oriented endeavor; more specifically, if I could do this for < $50 per crossover, I'd be ecstatic.

Good luck!

[cobbpa]

Good luck!

I know, maybe not possible for < $50, but seeing that one of the 3 way RS centers has a $78 network, I'm not needing anything that sophisticated. That's where the number came from. If it's more, so be it!

[Jed]

Save yourself the trouble and build one of the many designs available on this site and others. Now that I've said that here's why. You don't have measurements for us to simulate, just manufacturer responses on an IB. Yes, one can use the FRC tools to predict the acoustic slopes and then plug and play with the FRD and ZMA files to simulate a crossover in Speakerworkshop, but this is an almost neverending task of trial and error until things are just right. Now, I don't know how picky you are, but in my experience, the simulated speaker might sound ok for a bit, but then you will begin to get used to it and find things wrong with the tonal balance. In fact, this sometimes happens when you DO HAVE in box measurements and a good theoretical crossover. It takes a lot more than that to get it right. Thank goodness for folks on this board who have practical experience and know that a simulation is far from a finished product.

Here is some info on setting up speakerworkshop.

http://www.claudionegro.com/

Jed

[cobbpa]

Jed, I appreciate the comments and do realize that these speakers won't be optimal, and I don't expect them to be. It's an assortment of drivers on hand to be used for casual (2nd zone) listening in my bedroom. I do plan on building a couple of the designs on here for mains & center in my actual HT setup, but since I have these drivers I would prefer something just acceptable for my room, nothing to be listened to critically at all. They will be a step up from what I'm already using, and I'm fine with how far from great they will be. But I do appreciate the heads up, and I understand where you're coming from that one can't just throw things together and expect something great. Since I don't plan on testing my drivers on baffle or spending a lot of money, my expectations are low. I'll be pleased with some design to put out merely acceptable sound.

[Jed]

Jed, I appreciate the comments and do realize that these speakers won't be optimal, and I don't expect them to be. It's an assortment of drivers on hand to be used for casual (2nd zone) listening in my bedroom. I do plan on building a couple of the designs on here for mains & center in my actual HT setup, but since I have these drivers I would prefer something just acceptable for my room, nothing to be listened to critically at all. They will be a step up from what I'm already using, and I'm fine with how far from great they will be. But I do appreciate the heads up, and I understand where you're coming from that one can't just throw things together and expect something great. Since I don't plan on testing my drivers on baffle or spending a lot of money, my expectations are low. I'll be pleased with some design to put out merely acceptable sound.

Hi Cobbpa,

Now that we have this established. I tell you what... I'll take a stab at it for ya. It might take me a little bit of time because all the data has to be traced in for the FRD and ZMA files, normalized, phase added, and then crossover simulated. I'll post some results back here when I'm getting somewhere. Then maybe some of the speaker gods can critique the slopes. More than likely I'm going to target a 2nd order between the woofer and mid, and 4th order between the mid and tweeter. We'll see because I haven't even looked at the responses of these drivers yet. Any preferences for crossover slopes?

Jed

[cobbpa]

Firstly, thanks for the work you're doing. I realize it takes time, and though enjoyable, it's time that could be spent elsewhere. I have no preference for slopes; the tweet can play low enough to have no worries in a 3 way (Fs of 950) and the mid seems to have a pretty wide range in which it can be used, though I haven't come across a response graph that doesn't look smoothed. As I've stated, it doesn't have to be an amazing outcome; I'll be happy with whatever comes of it. I will be reading the content in the link you gave to try and do some playing of my own. Again, thanks Jed & whoever else may help steer me and my use of these spare parts.

[Jed]

Firstly, thanks for the work you're doing. I realize it takes time, and though enjoyable, it's time that could be spent elsewhere. I have no preference for slopes; the tweet can play low enough to have no worries in a 3 way (Fs of 950) and the mid seems to have a pretty wide range in which it can be used, though I haven't come across a response graph that doesn't look smoothed. As I've stated, it doesn't have to be an amazing outcome; I'll be happy with whatever comes of it. I will be reading the content in the link you gave to try and do some playing of my own. Again, thanks Jed & whoever else may help steer me and my use of these spare parts.

How many of those 175 woofers do you have? 2 per side would be nice because I have to cut the mids down quite a bit with just one. They are really low efficiency and not the best match for the mid and tweeter. I'll have a working crossover that I'll post tomorrow for you.

Jed

[Jed]

Firstly, thanks for the work you're doing. I realize it takes time, and though enjoyable, it's time that could be spent elsewhere. I have no preference for slopes; the tweet can play low enough to have no worries in a 3 way (Fs of 950) and the mid seems to have a pretty wide range in which it can be used, though I haven't come across a response graph that doesn't look smoothed. As I've stated, it doesn't have to be an amazing outcome; I'll be happy with whatever comes of it. I will be reading the content in the link you gave to try and do some playing of my own. Again, thanks Jed & whoever else may help steer me and my use of these spare parts.

I'll look through my Klang and Ton downloads to see if I can find a more accurate FR graph of the Tangband. They may have tested it. Right now I'm just using the published specs.

Jed

[cobbpa]

I only have 2, so 1 per side right now..I did notice the inefficiency while browsing specs. Go ahead & design for 2; the final price will still be less than that of building a pair of anything else.

[Jed]

I only have 2, so 1 per side right now..I did notice the inefficiency while browsing specs. Go ahead & design for 2; the final price will still be less than that of building a pair of anything else.

The front baffle layout is as follows: 40" tall and 9" wide with no roundover. If you want a 3/4" roundover just do it as it will smooth out the response dramatically. The tweeter and midrange centers are offset from the middle by .75". The tweeter center is located 3" from the top of the baffle. The midrange center is located 8" from the top of the baffle. The first woofer center is located 16" from the top of the baffle. The second woofer will be just below the first. Let me know if you want a sealed or ported enclosure and what volume works well for you (space constraits?). For sealed you can go anywhere from 21L to 35L depending on the Q. For ported you can go as large as 60L for the 2 woofers.

The midrange was modeled in a 3L sealed box- that should stay the same. Attached is a crossover using the specified drivers but adding another Dayton woofer. In the circuit diagram you will notice just a single driver is shown, but in reality it represents 2 Dayton woofers in parallel.

Jed

[cobbpa]

Looks good Jed, thanks so much :T . I have no real preference of ported vs. sealed, but I wouldn't mind sealed because I'm going to be pushing the limits of these power-wise, but if some ported alignment can be safely used, I'm all for it (I'll be careful). There is no real space constraint. I will probably end up rounding over the edges with a 3/4" bit. I could make them a little shorter if I make up for needed volume in depth and not cause problems with BSC, correct? I'll have to do some measuring height-wise to see if I can accomodate 40", but just in case. Seriously, I appreciate this and will probably start cutting the boxes next weekend. Exciting stuff. :B

[Jed]

Looks good Jed, thanks so much :T . I have no real preference of ported vs. sealed, but I wouldn't mind sealed because I'm going to be pushing the limits of these power-wise, but if some ported alignment can be safely used, I'm all for it (I'll be careful). There is no real space constraint. I will probably end up rounding over the edges with a 3/4" bit. I could make them a little shorter if I make up for needed volume in depth and not cause problems with BSC, correct? I'll have to do some measuring height-wise to see if I can accomodate 40", but just in case. Seriously, I appreciate this and will probably start cutting the boxes next weekend. Exciting stuff. :B

You can make them shorter, just keep the drivers spaced the way I simulated them on the baffle with their respective distance to the top of the enclosure. Sealed will give you an F3 of around 60HZ in 1cuft with the walls lined with dampening material. Then, just put a "shelf" in the tower to get the right volume. The one shortcoming of making the tower shorter is of course having a tweeter below the optimum listening axis.

If you want to go ported, and I see no problem in power handling with this, just the added possibility of bass boom if too close to the back wall. A 50L box with flared port on both ends 3" inner diameter by 10" in total length will give you an f3 below 30HZ.

Since you will need another two dayton woofers- you should know they are backordered for a bit.

Jed

[cobbpa]

Thank ya Jed, I hope you've enjoyed this at least a little. My only problem is that I can't find inductors with the DCR in the schematic. The closest ones are the foil ones, is that the intended component?

[Jed]

Yeah, it was fun. . . but remember it is still just the beginning. You might want to buy a few extra resistors to experiment with to get the tweeter Lpad perfect or at least to your liking. Also, the midrange series resistor can be modified to a lower value if you need more midrange presence. As for the DCR, I had iron core in mind for the larger values since you said you are on a budget. Unlike some other designs, it's not really all too critical in this one except for keeping the resistence low on the woofer inductor. However, if you have time, let me know the inductors that meet your price range and I'll model it up for you with the exact values before you purchase anything.

Jed

[yousuredo2]

"LPad" ?
someone explain what this is please ?
I thought it was a Volume control for tweeters...
I am so confused...

[Jed]

"LPad" ?
someone explain what this is please ?
I thought it was a Volume control for tweeters...
I am so confused...

Technically you are correct, but I'm referring to the series parallel resistors on the tweeter circuit that cut the SPL down to the correct level and contribute to the overall phase of the speaker. Depending on the room and other factors, these values can be adjusted to one's tastes.

Jed

[cobbpa]

PE has an iron core with value of 4.7mH and DCR of .231 ohms or one with the value of 4.5mH and DCR of .41 ohms. Which is better; the closer value of mH or the low DCR? (talking about the woofer section) I'll probably use another iron core for the 3.9 mH in the mid section and small (18 or 20 gauge?) air cores for the rest of the network.

Any advice on a way to minimize the cost of the 120 uF caps? Right now, it's a little over $100 a side, without the other mid. That's more than I'd like to spend, since the Modula MT isn't much more and CJD's mtm isn't much more than that. I'll probably still end up building it to use the parts, but may wait until the PE tent sale to do it as cheaply as possible.

[Jed]

PE has an iron core with value of 4.7mH and DCR of .231 ohms or one with the value of 4.5mH and DCR of .41 ohms. Which is better; the closer value of mH or the low DCR? (talking about the woofer section) I'll probably use another iron core for the 3.9 mH in the mid section and small (18 or 20 gauge?) air cores for the rest of the network.

Any advice on a way to minimize the cost of the 120 uF caps? Right now, it's a little over $100 a side, without the other mid. That's more than I'd like to spend, since the Modula MT isn't much more and CJD's mtm isn't much more than that. I'll probably still end up building it to use the parts, but may wait until the PE tent sale to do it as cheaply as possible.

Use the 4.5mH and 3.9mH, but madisound carries some low priced iron core. Check them out too. The large value caps can be electrolytic bypassed by a poly. That will save you a bunch of money and sound fine for this purpose.

Jed

[yousuredo2]

Technically you are correct, but I'm referring to the series parallel resistors on the tweeter circuit that cut the SPL down to the correct level and contribute to the overall phase of the speaker. Depending on the room and other factors, these values can be adjusted to one's tastes.

Jed

:T
thanx, as I learn, I feel like I am Lossing my mind...
I got alot of catching up to do...

[cobbpa]

Good deal; thanks again for all of this!

[Jed]

Good deal; thanks again for all of this!

Anytime... just let me know if you have any more questions. As you are seeing, it's pretty hard to make a cheap crossover (with good components) for a 3-way.

Jed

[Jed]

:T
thanx, as I learn, I feel like I am Lossing my mind...
I got alot of catching up to do...

Speaker building is a lot like Alice in Wonderland. You focus on opening up one door, only to reveal another 100 doors after it.

[cobbpa]

Hey Jed, if it's easy, could you investigate what it would look like if you bumped up the tweeter/mid xover to about 5k and changed to a 3rd order slope? I'm thinking cost reduction, but if it screws up power handling or the expected resopnse, I won't mess with it. Just curious. It may worsen the bump at 5k, if that's caused by the mid. Impedance could also be an issue. I dunno, just thinking, and if it only takes a minute, it may be an option.

[Jed]

Hey Jed, if it's easy, could you investigate what it would look like if you bumped up the tweeter/mid xover to about 5k and changed to a 3rd order slope? I'm thinking cost reduction, but if it screws up power handling or the expected resopnse, I won't mess with it. Just curious. It may worsen the bump at 5k, if that's caused by the mid. Impedance could also be an issue. I dunno, just thinking, and if it only takes a minute, it may be an option.

When I get a chance I'll play around with it more. May even raise the crossover point between the woofers and mid. That will require smaller inductors and caps- hence cheaper. I'm not even sure if that bump at 5K exists 10 degrees off axis. Remember I simulated for straight edges, so the roundover will help significantly in this regard. The impedance dips down to 3.1ohms and there's not much I can do about that in this configuration.

Jed

[cobbpa]

The impedence should be fine, I'm pretty sure my receiver will handle it, especially if it's just a dip and not across many octaves. I will probably end up rounding over all 4 edges. It was just a thought I had, for cost, but hopefully it doesn't hurt response too badly. That mid seems to be decent across a fairly broad range, at least according to the manufacturer's response graph. But don't take it too seriously, have a good holiday!

[Jed]

The impedence should be fine, I'm pretty sure my receiver will handle it, especially if it's just a dip and not across many octaves. I will probably end up rounding over all 4 edges. It was just a thought I had, for cost, but hopefully it doesn't hurt response too badly. That mid seems to be decent across a fairly broad range, at least according to the manufacturer's response graph. But don't take it too seriously, have a good holiday!

This one should work just as well if not better. I like the looks of the tweeter response a lot better than the last version. Fewer components too. Happy Holiday!

Jed

[cobbpa]

That response does look nicer; good work! I'll probably be using an iron core inductor in the woofer section, 20 ga air cores for the mid, and an 18 gauge one in the tweeter section. The only problem I see is the higher resistance of these small inductors; what are the effects of this?

[Jed]

That response does look nicer; good work! I'll probably be using an iron core inductor in the woofer section, 20 ga air cores for the mid, and an 18 gauge one in the tweeter section. The only problem I see is the higher resistance of these small inductors; what are the effects of this?

Hardly any change was observed when I plugged in the slightly different DCR values. You could even use a 2.0mH inductor in place of the 1.8mH and .4mH in place of the .39mH if you want to use 20 gauge air core inductors. I see partsexpress has some inexpensive iron core inductors. Let me know if you were able to make the $50 goal per side.

Jed

[cobbpa]

Looks about $60 per xover, using the air cores where I mentioned. Another $30 for woofers & terminals & such, and I'm sure I'll be happy with some decent TWMM towers for < $100 per side Thanks Jed..I'll probably have enclosure questions in the next couple days, then build pics to come!

[Jed]

Looks about $60 per xover, using the air cores where I mentioned. Another $30 for woofers & terminals & such, and I'm sure I'll be happy with some decent TWMM towers for < $100 per side Thanks Jed..and build pics will begin in the coming weeks.

Looking forward to it!

[cobbpa]

If you want to go ported, and I see no problem in power handling with this, just the added possibility of bass boom if too close to the back wall. A 50L box with flared port on both ends 3" inner diameter by 10" in total length will give you an f3 below 30HZ.

50L before the port or after? A 7x8x7" enclosure for the mid is 3.22 L, but takes up (as I calculate) 6.4 L from the larger box's volume. I'm building the xover in a seperate compartment, so no worries about that displacement. But after the two larger woofers and the mid's enclosure, not considering the port displacement (guessing you meant 50L, then put in a port) I'm thinking 9Wx40Hx14D is good; just over 59 L. Some bracing will be added also, of course. But just let me know if I should change this because of the port. Thanks!

[Jed]

50L before the port or after? A 7x8x7" enclosure for the mid is 3.22 L, but takes up (as I calculate) 6.4 L from the larger box's volume. I'm building the xover in a seperate compartment, so no worries about that displacement. But after the two larger woofers and the mid's enclosure, not considering the port displacement (guessing you meant 50L, then put in a port) I'm thinking 9Wx40Hx14D is good; just over 59 L. Some bracing will be added also, of course. But just let me know if I should change this because of the port. Thanks!

59L sounds about right. The 50L is not including volumes taken up by port, bracing, driver baskets and magnets, so 59L should do it by my estimation. I'd build the Xover so that it is easily removeable just in case you want to tweak it by ear later.

Jed

[cobbpa]

Which resistors would I play with? All of them? Or just the first one of each speaker's circuit? And assuming a lower value = more presence of that driver?

[Jed]

Which resistors would I play with? All of them? Or just the first one of each speaker's circuit? And assuming a lower value = more presence of that driver?

Changing the series resistor values by slightly different values should do the trick. Both on the tweeter and the midrange. If you feel like the sound is too warm, change the resistor on the mid to a lower value-- say about 4ohm or so. If the tweeter is dull, lower the series resistor a bit on that. A bit of trial and error might be necessary to suite the room and your listening habits.
Jed

[cobbpa]

Can someone look at this layout for the crossover? I cut the piece of paper down to the size I have to mount onto; I'd like to keep it all on one board. The part within the red line is the woofer section; between the red & blue line is the mid section; outside the blue line is the tweeter section. Pink pipe cleaners are wires to speaker :P White circles indicate they go to ground. I have a couple questions:

Are the inductors correct? The 3 air core ones are close together, but I'm thinking they're ok as long as they don't touch and don't 'point' at each other. Yes / no?

I couldn't get the site listed on Zaph audio that helps find out how much to unwind inductors to work. The 4 is supposed to be 3.9; I unwound 1 ring. The inductor you can't see the value of is 2mh but supposed to be 1.8; I unwound 3 times. These are just guesses, I didn't want to do too much. Do these sound close?

I wound 2 resistors in series to get 12 ohm; yay or nay?

[Amphiprion]

I would rotate the middle inductor so that it were pointing east-west along its axis. Keep everything at 90 degree angles if possible.

[cobbpa]

Ok...I'll make that change. I did the angle because one tip said that if the core is an eye, none should 'look' at each other. But, I'll trust ya :W

[opt-e]

Are the inductors correct? The 3 air core ones are close together, but I'm thinking they're ok as long as they don't touch and don't 'point' at each other. Yes / no?

It would be even better if you could separate the inductors by as much distance as possible.

[Jed]

Can someone look at this layout for the crossover? I cut the piece of paper down to the size I have to mount onto; I'd like to keep it all on one board. The part within the red line is the woofer section; between the red & blue line is the mid section; outside the blue line is the tweeter section. Pink pipe cleaners are wires to speaker :P White circles indicate they go to ground. I have a couple questions:

Are the inductors correct? The 3 air core ones are close together, but I'm thinking they're ok as long as they don't touch and don't 'point' at each other. Yes / no?

I couldn't get the site listed on Zaph audio that helps find out how much to unwind inductors to work. The 4 is supposed to be 3.9; I unwound 1 ring. The inductor you can't see the value of is 2mh but supposed to be 1.8; I unwound 3 times. These are just guesses, I didn't want to do too much. Do these sound close?

I wound 2 resistors in series to get 12 ohm; yay or nay?

Wooooooooh... I see a bunch of errors. I'm sorry for the bad news but you ordered some wrong parts. I see 2 inductors with the incorrect values. You bought a 4.0mH when it should be a .39 mH and the tweeter circuit is a .33mH to ground. Also the lead to the tweeter should be after the first resistor not before it. I can't tell which resistors you put together to make the 12 ohms. This is critical so let me know.

Jed

[cobbpa]

No problem on the inductors; just interpreted the uH mH wrong :cry:

I was planning on a trip back to PE this weekend anyway; bought a couple too many things at the sale.

I used a 10 & 2 to get 12. Good/bad? And if I have to switch, would the better (Mills) resistor arguably make a difference?

[Jed]

No problem on the inductors; just interpreted the uH mH wrong :cry:

I was planning on a trip back to PE this weekend anyway; bought a couple too many things at the sale.

I used a 10 & 2 to get 12. Good/bad? And if I have to switch, would the better (Mills) resistor arguably make a difference?

A 10 and 2 in series will work but it looks like you have them in parallel--- which won't make the necessary 12 ohms. I see PE has a 12.5 ohm resistor, so I'd just go with that (close enough for now). Don't bother with Mills for this budget project- you won't notice a difference.

Jed

[Jed]

A guide to inductor layout.

[cobbpa]

Cool little guide Jed; thanks again!
I'll probably just use the 10 & 2 ohm resistors in series if it's the same; I have them on hand.

Meanwhile, building is going well...little janky on the weight to push down on the tops & bottoms, but it works! Bracing to come; I'm gone all week for work, so time is short, they're coming along though. A couple mistakes / imperfections so far, so hopefully the paint helps me out at least a little.

[Jed]

A couple mistakes / imperfections so far, so hopefully the paint helps me out at least a little.

Nothing a bit of wood fill can't fix. Lookin good. :T