3-Way Design Help

[Rockfan]

Hi Everyone,

In the past I've built a couple of speakers using designs, but this time I've decided to dive in a build mine from the ground up. Any advice would be greatly appreciated!

Drivers:
HF - Beston RT001A Ribbon Tweeter
MF - Tang Band W3-315E 3" Aluminum/Magnesium Full Range Driver
LF - Dayton Audio RS150-8 6" Reference Woofer

If you see any glaring issues with the components I've selected please let me know, or I'll find out once the project is completed. Note: I may have jumped the gun, but I've already ordered these.

One thing I can't seem to wrap my head around is the placement of the drivers. I.E. Does it matter if the drivers location is HF-MF-LF or MF-HF-LF (from top to bottom) ? This mainly has to do with the visual aspect I'm going for. Thoughts?

[Jay1]

Re: 3-Way Design Help

How to do plan on designing the crossover? That would answer your question about driver placement

[Rockfan]

Re: 3-Way Design Help

I just used a design calculator on the inter-web. I've attached a photo of the design. At the cross over points the wave lengths are 4.5" and 27".

[lowpolyjoe]

Re: 3-Way Design Help

Hey Rockfan.

Crossover design is a deep, complicated topic. Simple parts-calculators ignore a lot of factors. I've never heard a design produced from one of those calculators and my ears are made out of wood anyway :p. But the experienced designers on this forum often advise people that the results will not sound so great.

Passive Crossover Designer by Jeff Bagby is a very popular tool used by many on this forum. It's an Excel file full of macros. It requires measurement files (impedance sweeps and frequency response) for the speakers to be used. It also takes a bit of reading to figure out how everything works. It's a commitment.

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


One thing I can comment on is that tweeter. I recently used it in my first serious attempt at a design from the ground up: The Lemons.

The tweeter should be flush-mounted, which is fairly difficult for non-round speakers. If you try to surface mount it, you would still need to dig out a divot behind it because the back is not flat. I happen to think it's a very attractive speaker, but it's an unpleasant physical specimen to work with . If you are a seasoned woodworker this probably won't bother you. If not, it's something to keep in mind.

Good luck with whatever you end up doing.

[Jay1]

Re: 3-Way Design Help

Those calculators dont account for anything like actual impedance and frequency response. You have a metal cone woofer with severe breakup and related distortion. A single inductor is NOT going to get the job done. That's only one problem... 3 ways take a lot of finesse

[Rockfan]

Re: 3-Way Design Help

Hey Jay,

I read on your site you help people design cross overs. I would love to take you up on that is the offer still stands!

[Rockfan]

Re: 3-Way Design Help

Lowpolyjoe, what do you think of the drivers? I read a lot about them before choosing them.

[Jay1]

Re: 3-Way Design Help

I'm not a fan of doing sims with manufacturers graphs because they can be erratic, especially tang band. What I come up with would still be 10x better then that generic crossover though. Figure out your box dimensions and driver locations (keep them tightly spaced) then shoot me a PM

[Herman Trivilino]

Re: 3-Way Design Help

I just used a design calculator on the inter-web.

One thing I know is that those calculators don't work and those off-the-shelf crossovers that you can buy just about anywhere don't work either. They'll produce something that's listenable but you won't be getting anywhere near the sound quality you deserve for the money you're paying for the drivers and crossover components, not to mention the time and effort you'll spend on the build. Plus the cost of the cabinet materials and any tools you might have to buy.

Not to worry, though, as the folks who frequent this board can help you. All you have to do is listen to them.

Here are three things for you to consider:

1. You can return the drivers and get store credit to replace them with other drivers or crossover parts, if need be.

2. Designing a crossover for a three-way system is much more difficult than for a two-way system.

3. There are plenty of designs for speakers already available, made by folks who really know what they're doing, and all the information you need to copy them is readily available.

[Jay1]

Re: 3-Way Design Help

One thing I know is that those calculators don't work and those off-the-shelf crossovers that you can buy just about anywhere don't work either. They'll produce something that's listenable but you won't be getting anywhere near the sound quality you deserve for the money you're paying for the drivers and crossover components, not to mention the time and effort you'll spend on the build. Plus the cost of the cabinet materials and any tools you might have to buy.

Not to worry, though, as the folks who frequent this board can help you. All you have to do is listen to them.

Here are three things for you to consider:

1. You can return the drivers and get store credit to replace them with other drivers or crossover parts, if need be.

2. Designing a crossover for a three-way system is much more difficult than for a two-way system.

3. There are plenty of designs for speaker cabinets already available, made by folks who really know what they're doing, and all the information you need to copy them is readily available.

This would be ideal if you're dead set on picking drivers for a speaker. My simulation will be a lot more accurate if all of the drivers were measured under the same conditions. I refer people to look over www.zaphaudio.com and pick drivers measured there (check the blog too).

#3 is a great point too, but I understand the fun in picking out everything

[Rockfan]

Re: 3-Way Design Help

Thanks Herman, I've done plenty of builds using other peoples designs. With this build I really want to dive in and get my hands dirty. Make some mistakes and learn from them!

However that being said, I'm keeping a very open mind and can't wait to read everyones advice!

[Rockfan]

Re: 3-Way Design Help

Jay,

Thank you, you have some good points. Hopefully this build will help me learn for my future builds. I don't think I want to return any of the drivers, but I will be sure to take notes on these for my next build!

And if I'm not happy with the sound, I will give them to a friend as a gift.

[Chris Roemer]

Re: 3-Way Design Help

You'd THINK a 3-way XO might cost 1.5 times what a 2-way does (three divided by two = 1.5, right?), but that's not how it goes.

For a 2-way, you need parts for 2 filters, an "HP" (high pass - for the tweeter) and an "LP" (low pass for the woofer). A VERY common Fc (crossover Frequency) for a 2-way is around 2000Hz.

For a 3-way, you need an HP for the tweeter AND another one on the midrange. You also need an LP on the woofer and another on the midrange (also). The mid's LP and HP combined are called a "bandpass" filter (BP). That's FOUR filters you need (instead of the 2 for a 2-way), so now you'd think that a 3-way XO might cost TWICE as much as a 2-way, but it ends up being more than that. It's VERY common to cross a 3-way near 3000Hz and near 300Hz. The parts to cross the mid to the tweeter (near 3k) will cost almost as much as the entire XO for a 2-way (at 2k), but the parts to cross the mid to the woofer down around 300 will consist of some LARGE caps (which could get pricey) and some VERY large coils ( - and, you've probably heard that copper keeps going up). So, in reality, a 3-way XO could end up costing maybe triple what a 2-way does.

It's for that reason, that most guys stick with 2-ways unless they're really going to get their money's worth out of the 3-way, which (IMO) would NOT include going with just a 6" woofer. Most would go with a 10" or 12" woofer (although Wolf's "Attitudes" only used an 8" woofer - albeit a VERY robust 8"er, AND they sounded excellent).

Your internet calculator hasn't even made ANY attempt to level the sensitivity variations between your chosen drivers (not even a LAME attempt). And, as has been mentioned, metal cones quite often need "special treatment" (meaning even MORE extra XO parts) to make them behave.

Sorry for the pep talk.

[Herman Trivilino]

Re: 3-Way Design Help

Thanks Herman, I've done plenty of builds using other peoples designs. With this build I really want to dive in and get my hands dirty. Make some mistakes and learn from them!

It seems to me then, that the next step for you is to learn crossover design.

Hopefully this build will help me learn for my future builds. I don't think I want to return any of the drivers, but I will be sure to take notes on these for my next build!

And if I'm not happy with the sound, I will give them to a friend as a gift.

That makes me think you'd want to start with a two-way or maybe a MTM. That would be a better way to achieve the goals you've set for yourself. Either use some of the drivers you've got, or save them for a future a build.

If I were you I might not be so inclined to listen to me. But Jay and Chris really know what they're talking about. They are both accomplished DIY designers. I just copy stuff that people like them do, but I do know the physics and I know how people learn.

[Pete Schumacher]

Re: 3-Way Design Help

I don't intend to step on Chris's toes here, but with small drivers, you're typically not going to cross the woofer at 300Hz. With this choice of drivers I'd be pushing the woofer as high as 800Hz. Probably something like 500-600Hz. The parts cost for them certainly would be more, but the mid/woof XO can easily be made 2nd order with only a single cap as the HP filter on the mid and probably only 2 parts on the woofer. I've done some 2-ways that have 9 components in the crossover. I've done some 3 ways with only 8.

The TB will need a notch to deal with the breakup peak but those will be really inexpensive and small parts. Otherwise, he's got a nice, wideband driver as a mid which should keep things simple. Sensitivity looks adequate to keep up with the RS150 especially since the RS will be handling all the baffle step duties.