Crossover help

[czag]

I have built a hand full of other people's designs and have built a few with crossovers modeled for me by Madisound. I have been pleased with pretty much all of the projects that I have built, some much more than others. I have read a few books and thought about attempting to model a crossover, but I never have.

So, here I am, wanting to start a new project. I want to build a 3 way, or maybe more accurately a 2.2 speaker system, with a Scan-Speak 9900(4 ohm variant), an Aurum Cantus ac130/50ck for the midbass and an ac200mkii for the woofer. The good news is that I will just cover the bottom of the network with the stereo sub outputs on my preamp and I think I have the box alignment figured out, and yes they are actually stereo outputs. So basically what I need is a two way crossover for the ac130 and the SS 9900.

Someone replied back to me in another thread in this forum that someone here might be willing to model a crossover for me when I mentioned that I was probably going to get Meniscus or Selah to design the crossover.

So, I am asking for that help. I was thinking of an LR2 for the ac130 and a third order for the tweeter somewhere around 2.2khz to 2.4khz. I know this will also need a notch around 4.5khz and 9khz. These will be placed next to bookshelves that are just as deep as the speaker will be and within 3-5 inches of the back wall. I know this is not optimal, but that is where they have to go.

It seems like a lot of work to do this for me, so it hard to believe that anyone would be willing to do this. Any input or constructive criticism is also welcome, and if that is all I get then that is still a big help.

Thanks,
Chris

[Pete Schumacher ®]

If you really want to learn this, let me suggest that you download Jeff Bagby's modeling tools so that you can have a go at it yourself.

http://audio.claub.net/software/jbabgy/jbagby.html

Download the suite. If you have Excel, these are absolutely awesome tools.

I think all the driver curves are available at Zaphaudio.com. You can trace the data and create your own raw files. Then, you can import them into Jeff's tools and model both room and enclosure responses and export those files for use in his crossover designer.

[czag]

Thanks for the suggestion. Actually, I have traced the fr and impedance of both drivers. I have downloaded that program and a few other FRD tools, and have played around some. But I really enjoy the wood working aspect of this hobby more than anything. I do enjoy building known designs and have paid someone to design the crossover as a last resort. I will delve deeper into crossover at some point in the future, but I barely have enough time to just build and finish my enclosures.

I have been lurking around for years, but I have always been able to find the answers by just looking for them. I have noticed that about 95% of all questions asked in new threads have been answered already in multiple other threads. I never really thought of asking for this help until a well known senior member replied to me that someone probably would help me. As I said earlier, I understand if no one does. But, I have seen plenty of people, some just this week, post with crazy ideas of wanting to build a speaker with 3 tweeters and 5 midranges and 7 woofers using stock crossovers and getting defensive or snarky when someone points out the obvious pitfalls of their ideas and then still getting significant help. I have not ever wanted to impose on people before, so I just read a lot and have built some proven designs. But, now I do want to use these drivers and really do not want to pay 200 to 300 dollars for a crossover design, although I might have to. I know this means that this might put the project or for a few years, but that is where I am at.

Chris

[Pete Schumacher ®]

Thanks for the suggestion. Actually, I have traced the fr and impedance of both drivers. I have downloaded that program and a few other FRD tools, and have played around some. But I really enjoy the wood working aspect of this hobby more than anything. I do enjoy building known designs and have paid someone to design the crossover as a last resort. I will delve deeper into crossover at some point in the future, but I barely have enough time to just build and finish my enclosures.

I have been lurking around for years, but I have always been able to find the answers by just looking for them. I have noticed that about 95% of all questions asked in new threads have been answered already in multiple other threads. I never really thought of asking for this help until a well known senior member replied to me that someone probably would help me. As I said earlier, I understand if no one does. But, I have seen plenty of people, some just this week, post with crazy ideas of wanting to build a speaker with 3 tweeters and 5 midranges and 7 woofers using stock crossovers and getting defensive or snarky when someone points out the obvious pitfalls of their ideas and then still getting significant help. I have not ever wanted to impose on people before, so I just read a lot and have built some proven designs. But, now I do want to use these drivers and really do not want to pay 200 to 300 dollars for a crossover design, although I might have to. I know this means that this might put the project or for a few years, but that is where I am at.

Chris

Before anyone can even begin to offer a crossover for you, a completed cabinet drawing with driver placements laid out is a must. That is required for calculating diffraction signatures specific to the cabinet shape. That signature is what you use to modify the infinite baffle response plots to reflect on-baffle response. Those modified plots are used to design a crossover.

[czag]

Thanks again for the input, will get to that as soon as I can. Like I said, even if I do not get a design, any thoughts are very welcome and appreciated.

Chris

[jkim]

Hi Chris,

Your system has a potential to be a really good one if designed properly. I'd be willing to help with this design, but these days I don't have much time to spend on full modeling. Although I agree with Pete that you can do it from scratch, designing a system with accuracy requires some knowledge and skills, and there are a few critical steps where you could make mistakes.

Among people who can help here, I'd trust only a few for accurate modeling. They are Pete Schumacher, Jeff Bagby, Roman Bednarek, and dlr (Dave L Ralph)---please don't be resentful if I dropped your name here It seems that Jeff and Roman aren't doing much free modeling work for others these days. dlr can certainly help if he wants to. So, it seems that Pete is pretty much the only one.

However, if you're more of a perfectionist and willing to spend a bit more, it will be best to send your cabinet and drivers to a person you can trust, and have him measure and design a crossover for some amount of money. I heard very good things about Rick Craig at Selah Audio. Another person I'd trust will be Dan Neubecker. You may want to contact him, too, if you want to go this route.

Jay

[Wolf]

If I didn't already have 3 pending models, I'd help him...
Later,
Wolf

[rogoll]

I haven't dealt with them, so I cannot actually vouch for them, but here's someone you might want to contact --> http://www.pirateacoustics.net/crossovers.html

[czag]

Thanks Jay, Pete and everyone else. I have built designs from some of the people that you listed and have been quite satisfied with the results. Those are most of the people I was hoping to get help from. Believe me, I can understand being to busy to help. I have two four year old twin girls and one always like to help daddy "work". She even has her own tools, while this is very enjoyable, it does make for slow progress sometimes.

I have done some fooling around with FRD tools and I was tempted to try this one as my first. What stopped me is that I have been wanting to do something with these drivers and since they are such high quality drivers I wanted this to be done well. I figured that if I did a little modeling and then put the results up here to ask for help, I would be starting with a cruddy design and with some help it would probably turn into only a semi cruddy design. This would be fine for a less expensive design, but I have high hopes for this one. I have always put projects before this one, but I have some interesting ideas for a cabinet and really want to get out of my comfort zone with the wood working and get this started. I am getting bored with rectangles.

I'll take Pete's advice and put up a cabinet design, if the least I get is feedback on the cabinet, alignment and layout that is still a big help.

If worse comes to worse, I do have the RB Reference from madisound to start. My wife asked if I need speakers in every room. I told her no, of course not. I don't need them in the closets, but everywhere else, yes. I have to say that this forum is great and that I have learned quite a bit from searching around. I appreciate you guys sharing your knowledge. I do have a little knowledge myself and will try to chime in on other people's questions.

Sorry for the lengthy replies, I just do not want to come off as lazy when asking for help, and want my motives understood.

Chris

[czag]

Before anyone can even begin to offer a crossover for you, a completed cabinet drawing with driver placements laid out is a must. That is required for calculating diffraction signatures specific to the cabinet shape. That signature is what you use to modify the infinite baffle response plots to reflect on-baffle response. Those modified plots are used to design a crossover.

Here are the cabinet drawings. The Tweeter is truncated and measures 130 x 116 mm., it is 3.5 in from the top and centered horizontally on the baffle. The mid-woofer is 157 mm in dia. and is centered horizontally 9 in. from the top. Measurements from the top are to the driver center.

The baffle is 7.25 in. wide at the top and 8.6875 in. at the bottom by 12.25 in. deep and 37.75 in high. The wall thickness will be 1.125 in on all sides.

It will be ported out the back near the top, I will be 3 in. flared port with a 90 degree bend. The woofer volume will be .7747 ft3 tuned to 37.74 Hz. The mid-woofer and the tweeter will be separated internally from the woofer.

Does anyone see any potential problems or any ideas to improve this?

I was wondering if 4-6 inches space between the wall and the back of the speaker is enough?


Chris

[Pete Schumacher ®]

Here are the cabinet drawings. The Tweeter is truncated and measures 130 x 116 mm., it is 3.5 in from the top and centered horizontally on the baffle. The mid-woofer is 157 mm in dia. and is centered horizontally 9 in. from the top. Measurements from the top are to the driver center.

The baffle is 7.25 in. wide at the top and 8.6875 in. at the bottom by 12.25 in. deep and 37.75 in high. The wall thickness will be 1.125 in on all sides.

It will be ported out the back near the top, I will be 3 in. flared port with a 90 degree bend. The woofer volume will be .7747 ft3 tuned to 37.74 Hz. The mid-woofer and the tweeter will be separated internally from the woofer.

Does anyone see any potential problems or any ideas to improve this?

I was wondering if 4-6 inches space between the wall and the back of the speaker is enough?


Chris

Well, one point that needs addressing. You've positioned the tweeter at the most symmetrical possible location for the worst possible diffraction performance. You've put both sides, as well as the top edge, at ~3.5" from the dome center. My suggestion, if you want to keep it centered, is to move it down another inch or so, or move it as close to the top edge as possible. The baffle diffraction will be smoothed better that way.

If you can create an angled enclosure for the mid, it will help suppress standing waves.

Do you have the actual TS parameters? The difference between what PE has posted for the 200MKII and the linked PDF spec sheet are quite substantial.

[czag]

Pete, thanks for answering my questions, it's a big help.

OK, the tweeter is moved up to 2.78 in from the top and it is 3.625 from either side. Is that better, or should I just move it down 1 inch from the original placement. I have planned on angled enclosure for the mid and tweeter.

The T/S specs are from Zaph, those were my ac200mkii's that he tested last summer. He told me the specs should be good for box modeling.

Is 4-6 inches space in the back enough between the wall and the back of the enclosure to vent it out the back?

Thanks,
Chris

[Pete Schumacher ®]

Pete, thanks for answering my questions, it's a big help.

OK, the tweeter is moved up to 2.78 in from the top and it is 3.625 from either side. Is that better, or should I just move it down 1 inch from the original placement. I have planned on angled enclosure for the mid and tweeter.

The T/S specs are from Zaph, those were my ac200mkii's that he tested last summer. He told me the specs should be good for box modeling.

Is 4-6 inches space in the back enough between the wall and the back of the enclosure to vent it out the back?

Thanks,
Chris

Moving the tweeter up is probably your best bet. Looks better than moving it down too.

Instead of a standard ported box, you might go with a tapered, partially stuffed line. I modeled a 50" long line, with a starting area 3 times the area of the woofer (~100sq inches) and the open end 1/3 the area of the woofer (~10 sq inches). The woofer 10" is from the closed end. First half of the line is stuffed with about 16oz of fiber. System F3 is upper 30s. Total enclosure volume is 1.6 cubic feet. It's larger than what you have for a vented box, but there is no port specifically to deal with. The opening could be a slot of 1.5" x 7", plenty of area for a terminus. A 3" port would need to be almost 2 feet long to tune where you want to. That's over .1 cubic feet alone and it will produce a pipe resonance at 300Hz, affecting response. The tapered line approach has a better managed passband response in comparison but the cabinet would be somewhat larger.

[czag]

A TL is something I have thought about. That is an interesting proposition.
Are you using the TS specs from Zaph. with an fs of 32.57hz, a qts of .324, a Qes of .34, and vas of 40.7 lt. I get a vent of 14.75 in. from a .774 cu ft box tuned to 37.74hz. This gives an F3 of 37hz and can go to 108 db with no excursion problems if a 30 hz HP is used. I was thinking of using a 30hz 12db FMOD hp. I used unibox to get these results, how accurate is unibox? It also shows the pipe resonance at about 425hz, by that time the response will be down 24-36 db. Am I making a mistake somewhere? I admit that I know little to nothing on the subject of pipe resonances, I'm just reading the unibox graph.

Thanks,
Chris

[Pete Schumacher ®]

A TL is something I have thought about. That is an interesting proposition.
Are you using the TS specs from Zaph. with an fs of 32.57hz, a qts of .324, a Qes of .34, and vas of 40.7 lt. I get a vent of 14.75 in. from a .774 cu ft box tuned to 37.74hz. This gives an F3 of 37hz and can go to 108 db with no excursion problems if a 30 hz HP is used. I was thinking of using a 30hz 12db FMOD hp. I used unibox to get these results, how accurate is unibox? It also shows the pipe resonance at about 425hz, by that time the response will be down 24-36 db. Am I making a mistake somewhere? I admit that I know little to nothing on the subject of pipe resonances, I'm just reading the unibox graph.

Thanks,
Chris

I was using Zaph's parameters. i was also assuming a 3" diameter port and 31.5Hz tuning. I must have read it wrong when calculating.

The Fmod is great idea.

[czag]

I think the vented option is the way to go to keep the speaker smaller. The speaker will only be placed 4-6 inches out from the wall is that enough space to run the vent out the back?

Thanks,
Chris

[Pete Schumacher ®]

I think the vented option is the way to go to keep the speaker smaller. The speaker will only be placed 4-6 inches out from the wall is that enough space to run the vent out the back?

Thanks,
Chris

That should be enough distance.