DIY MTWM center channel crossover help

[runninkyle17]

Hello everyone-

This is my first DIY speaker and I have been playing around with the crossover for a little while. Here is what I have so far. I would appreciate any advice and criticism.

Crossover schematic:

http://pic4.picturetrail.com/VOL744/4324355/9108053/127721371.jpg

Component values:
3-Way Crossover Network
Low-Pass (LP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Corner Frequency: 500 Hz
Band-Pass (BP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Lower Corner Freq: 700 Hz
Desired Upper Corner Freq: 4000 Hz
High-Pass (HP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Corner Frequency: 3400 Hz
C1 = 7.524 µF, Polypropylene, 0.00589 ohms
C2 = 18.45 µF, Polypropylene, 0.00383 ohms
C3 = 32.03 µF, Polypropylene, 0.00294 ohms
C4 = 102.1 µF, Polypropylene, 0.0015 ohms
C5 = 10.1 µF, Polypropylene, 0.00508 ohms
C6 = 209 µF, Polypropylene, 0.00132 ohms
L1 = 0.161 mH, Air Core(#16), 0.268 ohms
L2 = 0.844 mH, Air Core(#16), 0.347 ohms
L3 = 0.43 mH, Air Core(#16), 0.299 ohms
L4 = 0.099 mH, Air Core(#16), 0.261 ohms
L5 = 0.879 mH, Air Core(#16), 0.351 ohms
L6 = 0.359 mH, Air Core(#16), 0.291 ohms
R1 = 0.645 ohms

Woofer
L-Pad: 1.00 dB
Rp1 = 0.222 ohms
Rp2 = 5.581 ohms



Response graph:

http://pic4.picturetrail.com/VOL744/4324355/9108053/127721362.jpg

[runninkyle17]

Anybody got any ideas on this for me?

[cjd]

Well... there's not really a whole lot we can say specifically about the transfer functions. To really have much feedback we would need to see how the transfer functions work on the actual response, etc.

A couple things though - why do you have ANY resistors on the woofers, particularly that 5ohm in parallel?

Are you accounting for baffle step?

Really, just looking at the graph I see rather lumpy response and I don't have any idea WHY it's that way. How did you arrive at these values?

C

[ThomasW]

I've never seen a X-Over Pro design. I assume you haven't measured the drivers in your intended baffle?

As Chris noted one shouldn't use a L-pad on a woofer, 1/2 the power will go to it

That particular PE dome has a pretty nasty looking plot.

[Jim Holtz]

This might help a little. This is a budget 3-way design that Rick Craig and I came up with about 3 years ago. It uses the PE 285-010 mid dome. Here's a link...

http://www.audiodiycentral.com/sd300_jh3.shtml

HTH

Jim

Hello everyone-

This is my first DIY speaker and I have been playing around with the crossover for a little while. Here is what I have so far. I would appreciate any advice and criticism.

Crossover schematic:

http://pic4.picturetrail.com/VOL744/4324355/9108053/127721371.jpg

Component values:
3-Way Crossover Network
Low-Pass (LP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Corner Frequency: 500 Hz
Band-Pass (BP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Lower Corner Freq: 700 Hz
Desired Upper Corner Freq: 4000 Hz
High-Pass (HP) Filter: 1 required
Type: 3rd-Order All-Pass (APC)
Desired Corner Frequency: 3400 Hz
C1 = 7.524 µF, Polypropylene, 0.00589 ohms
C2 = 18.45 µF, Polypropylene, 0.00383 ohms
C3 = 32.03 µF, Polypropylene, 0.00294 ohms
C4 = 102.1 µF, Polypropylene, 0.0015 ohms
C5 = 10.1 µF, Polypropylene, 0.00508 ohms
C6 = 209 µF, Polypropylene, 0.00132 ohms
L1 = 0.161 mH, Air Core(#16), 0.268 ohms
L2 = 0.844 mH, Air Core(#16), 0.347 ohms
L3 = 0.43 mH, Air Core(#16), 0.299 ohms
L4 = 0.099 mH, Air Core(#16), 0.261 ohms
L5 = 0.879 mH, Air Core(#16), 0.351 ohms
L6 = 0.359 mH, Air Core(#16), 0.291 ohms
R1 = 0.645 ohms

Woofer
L-Pad: 1.00 dB
Rp1 = 0.222 ohms
Rp2 = 5.581 ohms



Response graph:

http://pic4.picturetrail.com/VOL744/4324355/9108053/127721362.jpg

[runninkyle17]

Thanks for all the info and the criticism. I think I am going to forego the 3-way and stick with an MTM design.

I just received some updated T/S parameters on the woofers I am going to use, so that should help out.

I am looking to copy the 1 cu. ft PE box and set the drivers at the same distances like the Modula MTM design that is posted on this forum.

I will be using 2 Peerless 731773 6.5" woofers and a Vifa D25AG-35-06 tweeter (this particular tweeter was used on a McIntosh design, so it has a rectangular faceplate instead of a circular one).

These are the parameters I have for the Peerless woofers:
Z = 8 Ohms
Re = 4.9ohms
Fs = 52Hz
Qes = .6509
Qms = 2.3624
Qts = .5103
Zmax = 22.8 Ohms
Le = .4224 mH
Sensitivity = 88dB 1 watt / 1 meter
Vas = 19.27 liters
Freq. Response: 50Hz - 20000Hz
Pe = 100 W

This is the info I have on the tweeter:

Re = 4.6 Ohms
Fs = 850 Hz
Qes = 0.677
Qms = 1.128
Qts = 0.423
Le = 0.0521 mH
Z= 6 Ohms
Freq. Response: 1500Hz - 35000Hz
Sd = 7.1 sq.cm
Rms = 1.42 kg/s
BL = 3.3 Tm
SPL = 90 db 1Watt /1 Meter
Pe = 100 W

That is all I have so far. Hopefully I can get the enclosure made and then start working on the crossover soon. I am tempted to send this info to Madisound and have them model a crossover for me, but I would like to do it myself just so I can get my fingers dirty. Again if anyone has any advice I would be very happy to hear it.

-Kyle

[cjd]

If you can not take acoustic measurements, send it somewhere. T/S parameters will help you optimize the box, but if you're just picking a box and not worrying about how it works out... will also allow you to see if porting would be any value. That box size *should* work well enough though I suspect.

Madisound does decent work from what I know. There are others around that do this as well.

Randomly picking crossover components doesn't usually work well. Using values from crossover calculators doesn't do much other than get you started, because real drivers never work the way the calculators assume (purely resistive, perfectly flat response, and no baffle step to worry about!)

You *can* start looking at other designs that use these drivers and very possibly arrive at something that works out well enough for you. It probably won't be optimal.

It's not something to do just to get your hands dirty, 'cause you get sucked down the rabbit hole. And it's a deep hole.

C

[runninkyle17]

That'a good to know. I guess now my question is, is it worth it to get the equipment to test the drivers out? I will be making more speakers later, but I really have no idea how much the testing equipment costs.

Sorry if this is a terribly newbie thread. I guess I should go out and buy some of the loudspeaker design books before bothering you guys about it.

[ThomasW]

This will get you started

http://home.comcast.net/~thomasw_2/SubwooferSetupandEQpage10.html

[dawaro]

Try downlaoding Speaker Workshop . Add this Behringer mic and this Behringer mixer to your pc with Speaker Workshop and it will give you an entry measurement set up for $100.

The hard part is learning how to do the measurements and what to do with them. There is a pretty good manual that has been written for Speaker Workshop that will help get you started.

[cjd]

SW still requires the cables or jig. Not quite a $100 setup, but close-ish. It also takes a good sound card - onboard sound generally doesn't cut it.

If it's something you're going to continue doing over time for "fun", it's a great investment. If it's for one project? Find a project that does what you want and build it.

C

[runninkyle17]

I will definately continue doing this for "fun" and just because I am interested in it. I will check out all the suggestions everyone made and probably get a new soundcard for my laptop (SB Audigy comes in a PC card version that I have heard good things about). Thanks again for all the suggestions and I will keep everyone up to date on my future purchases. My wife isn't going to be happy :argue:. Oh well, some things just can't be helped :T :T :T :T !!!!

[runninkyle17]

I was just looking around at mixers. Does anyone know if the Behringer UB502 would also work for my situation or should I spend the extra money to get the UB802?

[cjd]

The UB502 does not have phantom power. The mic needs it. So, you want the UB802.

You'll also need a mic cable that works with it.

I believe NADY offers a less expensive version of the mic. And some sound card options can provide phantom power. So do a little research and maybe save some!

The M-Audio Transit is also generally very well received as a laptop soundcard option.

C

[Dennis H]

I have a small problem with using the Behringer mixer as a mic preamp for measurements. It has tone controls with a detent on the pots for the "flat" position. Personally, I wouldn't want to trust that the pot was really giving a flat response. For a few bucks more, you can get a Rolls MP13 Mini-Mic. Google finds prices around $60.

[runninkyle17]

I just bought a Behringer UB1002 off of eBay. Got a great deal (cheaper than buying a new UB502). I will be looking to get either the NADY CM100 or the Behringer ECM8000 mic soon. The Nady is cheaper and it is the same thing, so I will probably go with that unless I find something on ebay.

I had a question though. What is the best way to simulate an infinite baffle? I was thinking just a 4 x 8 piece of MDF with a square cut out in the middle so that I can put inserts in that I can custom fit for the driver I am testing.

Also, does anyone know a good test box design? I have been looking around, but I have not really found anything that makes sense to me.

[cjd]

Absolutely no reason to simulate an infinite baffle: You want to measure IN your final cabinet with all diffraction and step loss and everything right there in the numbers. Same goes with test box.

C

[runninkyle17]

Great that makes it easier on me. I can start working on the box now and by the time I get the box done I will probably have all the test equipment I need. Sweet!

[cjd]

Remember to consider how the ultimate series impedance may alter box design attributes.

C

[runninkyle17]

Once I have the baffle of the enclosure I intend on using made, should I test the speakers in that baffle and not in the enclosure to get some baffle difraction analysis?

As far as the impendance goes for the series, will the enclosure dramatically affect the impendance and is there anyway I can model this in a software program?

[cjd]

diffraction is best modeled before you start building baffles. Measuring a live speaker will not really give you diffraction results unless you have something to compare it against (infinite baffle measurement).

FRD Consortium plus Windows and Excel get you the tools you need.

And, series impedance will effect the system Q and how a woofer models in the box. It's less critical how the box changes impedance, since it is less likely that you'll be crossing in that region (which is where accurate impedance is most critical). Usually, the more series impedance, the bigger the box for the same results. Which is why we like big gauge inductors on woofers. 14ga or 12ga. Were it not for the problems, iron core would be lovely here. For me, it's gotta be air core.

C

[runninkyle17]

Thanks for the advice. I will check out the FRD site and get the software from it.

I had another question though. I have been looking for a midrange speaker that has a flat response from about 500Hz to 4.5 kHz and has a sensitivity around 90 dB. Does anyone have any suggestions? I have looked at the fostex FE127E, but I read one review on this driver and it said it was crap. I like the look and the buzz on the Hi-Vi B3S, but the sensitivity is only around 80 dB and I think that is too low. However, since I am not an expert this may be perfectly all right. I usually do not play my music all that loud and my room is only 17 x 12 x 8; so if the sensitivity of the speakers do not match is there anyway to make them match. I thought about L-Pads, but I was told that putting an L-Pad on a woofer is a bad idea. So you can see I need a little help.

I would like to match this up with my Peerless woofers and the Vifa tweeter so that I can get improved off-axis reponse from my center.

[cjd]

Why do you want to go so high in a midrange? That introduces a slew of other issues including center-to-center and whatnot.

If you push the lower limit to 700Hz or so, the B-G Neo8 may fit the bill (since I do not consider it a tweeter at all). Some of the high end dome midranges may also pull this off. Not a whole lot else, at least in the sensitivity range you're talking about.

I can't find data on the Peerless drivers you have selected, but another option would be one of their 4" drivers. It won't do 4500 by any means, but should handle 2500 quite happily and allow for a nice snug design.

C

[runninkyle17]

When I got to looking at all my options for a midrange, I realized that I would be hard-pressed to find a midrange with that sensitivity.

I guess I should have asked if it is necessary for the different drivers that I am going to have similar sensitivities?

I have thought about using the Audax AP130ZO 5.25" midbass because there is a lot of literature on it and it has a fairly smooth curve. As far as the frequency response from the midrange, I looked at the D25AG's frequency response and it is useable down to about 2000Hz, so if I could get a midrange to run smooth from 700Hz to 2500Hz I think I would be in business.

This opens up my choices quite a bit, so I guess I will have to do some research and see what I end up with.

CJD thanks so much for all your help so far. I am sure I will be posting more questions later.

[cjd]

Tweeters almost always get padded down.

With full baffle step, a pair of woofers in parallel will end up with a net sensitivity the same as if you were just using one (gain6dB by parallel and doubling Sd, lose 6dB by baffle step).

In general then, your target sensitivity is going to be limited by what you can pull off with the woofers.

Also, you may want to target a lower woofer-mid crossover but use the woofers to cover most of the baffle step losses that will hit the midrange. Maybe 400Hz-ish.

A midrange *can* be padded down, but I really try to avoid padding circuits at all when I can.

I still think a quality 4" driver may be the ticket here. Something ~88dB should be quite do-able and flexible.

C

[runninkyle17]

I have looked around and I think I am going to go with the Vifa K10MD-19-08 4" woofer. They are on buyout at PE and really cheap. The sensitivity is rated at 88dB and the frequency response seems reasonable. I may go ahead a buy a few of them when they come back in stock to have extras when I start on some other speakers.

Here is the link to them: link

Hopefully they will be back in stock soon or else I might try and find some other woofers that will do the trick.

[cjd]

I wouldn't count on those being available again. It's a rare thing with the buyouts.

C

[runninkyle17]

Okay I have been looking around for a good 4" midrange with a decent SPL (at least 87dB), but so far I have had no luck.

I have seen some of the Fostex drivers that look good, mainly the FE103E, 107E, and 127E. Unfortunately I do not know if the Fostex will match up with my Peerless woofers and my Vifa metal dome tweeter (D25AG3506).

If anyone has any experience with the fostex drivers can you give me some feedback. Thanks

[runninkyle17]

Well after getting fed up looking for a good budget midrange, I opted out for the Vifa K10MD-19-08. There was a pair on eBay for $17 and while I had to kick myself because I could have bought 5 of them for that price off of PE if they were in stock I went ahead and did it anyway.

I figured I would just get the pair because right now I have the two Peerless woofers (831773), two Vifa tweeters (D25AG3506), and the two Vifa midranges (K10MD1908). So if in the future I decide to upgrade the drivers in my center channel I will have the drivers to make two 3-way bookshelves.

Anyway, I am still going to make the box about the size of the PE 1.0 cu.ft. box. I think I figured it out that the dimensions are going to be 8" tall, 22" wide, and 14" deep. I will make a small chamber for the midrange and tweeter (not sure what size yet because I haven't looked at the Vas on the midranges in detail).

After looking at the Audax center channel that is so popular I think I will try to copy that box design, but I do not want the speaker to be that tall. I am hoping to be able to get away with only an 8" tall speaker, but I may have to go up to 9" (it is amazing how much bigger the extra inch makes the speaker look).

So that is my plan so far. I will start making the box late next week after I get done with all my tests (for some reason my med school likes to torture us with a full week of tests at least three times every semester). On that note I will be off.

I probably will not be able to update my progress for a week or so, but if anyone out there has any suggestions please feel free to post and give me all types of advice.

I am still a major noob at this whole thing, so please bear with me. Cheers!

-Kyle

[Dennis H]

I will make a small chamber for the midrange and tweeter (not sure what size yet because I haven't looked at the Vas on the midranges in detail).

T/S parameters aren't really relevant for a mid because you are usually crossing well above Fs. Generally, the more depth you can give a mid, along with light stuffing, will help remove the "box" sound (front/back reflections radiating back out the cone). Some people even leave the back off the mid enclosure, along with stuffing, to help remove those reflections.

PS that's a problem I noted with the RS WMTW center posted here -- a tiny midrange enclosure. Not good for the ultimate in midrange clarity.

[runninkyle17]

I will have to keep that in mind. I know it sounds trivial, but my biggest concern with the box is the amount of bracing I need to put in it. I may do something like the PE MTM box where the brace is from side to side or I may just figure out something else.

I know bracing is important so I want to make sure and put enough in the enclosure.

[runninkyle17]

Okay I am back!

I have the enclosure pretty much finished except for one last piece of bracing. I also got my UB1002 mixer and NADY CM100 mic in the mail so I am ready to start testing.

I did have a few questions though.

I want to use bannana clips to connect the drivers because I do not want to solder anything just yet. What cables should I get for all of this to work? I know I need a mic cable and I am going to pic one up today, but what other cables to I need?

Thanks guys.

[ThomasW]

Look at the connection pictures in this thread

http://www.avsforum.com/avs-vb/showthread.php?t=572477&highlight=TrueRTA+for+dummies

[runninkyle17]

Thanks for the link to the thread. There is a lot of good information in there.

To acurately test my drivers, what is the best way to hook them up? Should I run the output from the amp I am using straight to the driver I want to measure or hook up a jig or something in between?

Thanks in advance.

[ThomasW]

This depends in the software you're using and what you're measuring. A jig is usually required if you're making measurements for crossovers/speaker design

[runninkyle17]

I do intend on making measurements to design crossovers. So I guess I will need to build or find a jig to use.

Does anyone have any recommendations on a good jig? I have looked around and seen the Wallin jig, but have not been able to find plans for v2 of the jig.

Also, is there any things that I could buy instead of making my own jig?

[ThomasW]

Google is your friend.....

http://www.vikash.info/audio/sw_jig/

[runninkyle17]

Initial crossover design is done!!!

Okay,

So now I have the initial crossover design done and have all the cap and inductors. I will be making this a tri-wired design (three sets of binding posts on the back) just because I like to be different.

Does anyone have any recommendations on the best way to hook this all up? I have 14 AWG wire to run from crossovers to the drivers.

I guess my real question is about what type of wire to use in the actual crossover circuit? Most of the connections will not need extra wire, but the input and output terminals will need extra wire.

I know these are really newbie questions, but someone has to start somewhere.