Building 3-way speakers

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
diyAudio Moderator
Joined 2008
Paid Member
Obtaining 4 ohms is not a very useful goal. Your two woofers is close enough in that direction. Fs is important but different drivers handle it differently. Try to stay around double or more away from it. The operating range shown is not very useful especially with these crossovers. The best way to find out may be to 'suck it and see'.
 
I suggest taking the opportunity to learn how to build an x-over.

See this website for diagrams of the wiring and value calculations that can be used as a STARTING point for an x-over.

Crossover Design Calculators

I have in the last couple weeks learned how to use x-over simulation software and must agree with the sentiment of the other more experienced diyers here that you really should sim out the x-over to ensure that it will work as expected. You may be able to build x-overs that are properly matched to your drivers for a similar price as those pre-manufactured units, while [possibly] using higher quality components.

The sealed back mid-range unit there should work fine, though, I would personally suggest that you use a regular open-back woofer and give it a nice little sealed box with plenty of damping material. The spike in output centered ~450 hz of the goldwood is something that could be factored into x-over design and "taken care of," whereas, the pre-made x-over will not account for this.

If you figure out what drivers you want to use, let me know and I'll be happy to sim up a simple x-over for you. It might not be the best x-over possible, but at least it would be closer to ideal.
 
I suggest taking the opportunity to learn how to build an x-over.

See this website for diagrams of the wiring and value calculations that can be used as a STARTING point for an x-over.

Crossover Design Calculators

I have in the last couple weeks learned how to use x-over simulation software and must agree with the sentiment of the other more experienced diyers here that you really should sim out the x-over to ensure that it will work as expected. You may be able to build x-overs that are properly matched to your drivers for a similar price as those pre-manufactured units, while [possibly] using higher quality components.

The sealed back mid-range unit there should work fine, though, I would personally suggest that you use a regular open-back woofer and give it a nice little sealed box with plenty of damping material. The spike in output centered ~450 hz of the goldwood is something that could be factored into x-over design and "taken care of," whereas, the pre-made x-over will not account for this.

If you figure out what drivers you want to use, let me know and I'll be happy to sim up a simple x-over for you. It might not be the best x-over possible, but at least it would be closer to ideal.
Yea after I slept on it I was thinking about getting single crossovers and putting them together then decided nah I guess I just have to learn how to build the correct one I will check out that link when I get home from work tonight hopefully I can figure this all out.
 
A few things to think about when picking drivers and choosing X-over design:

If you want to cross both 10" woofers over at the same point (traditional 3 way), it would probably be best to mount the drivers such that the mid and tweeter is equal-distance from each woofer. So.. a sort of triangle of drivers with those 10"ers side by side in your case with the tweeter above the mid. In the future, if you build the same drivers into a different box, it would probably be best to arrange keep a similar arrangement of drivers or place the mid and tweeter side-by-side and in-between the woofers. This will help keep driver spacing to each woofer inside of the size of the x-over point wavelength. IMO baffle step correction in those current wide baffle boxes is unnecessary as they are right against a wall and you have a sub that can more more than overcome any baffle losses.


Due to the combined sensitivity of the 2 10" drivers, your mid and high frequency drivers either need to be either 4 ohm and ~90dB@1W or better sensitivity, or 8ohm and 93dB@1W sensitivity or better. As far as dome tweeters go, this means you will almost certainly need a 4 Ohm dome, since very few domes are 93dB or higher sensitivity.

Dome ideas:
Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026
Vifa BC25TG15-04 1" Silk Dome Tweeter 264-1040

Mid-range ideas:
Dayton PK165-8 6" Professional Kevlar/Paper Cone Midrange 295-020
Eminence Alpha-6A 6" Midrange 8 Ohm 290-400
Celestion TF0615MR 6" Professional Midrange Speaker 50W 294-2054
 
A few things to think about when picking drivers and choosing X-over design:

If you want to cross both 10" woofers over at the same point (traditional 3 way), it would probably be best to mount the drivers such that the mid and tweeter is equal-distance from each woofer. So.. a sort of triangle of drivers with those 10"ers side by side in your case with the tweeter above the mid. In the future, if you build the same drivers into a different box, it would probably be best to arrange keep a similar arrangement of drivers or place the mid and tweeter side-by-side and in-between the woofers. This will help keep driver spacing to each woofer inside of the size of the x-over point wavelength. IMO baffle step correction in those current wide baffle boxes is unnecessary as they are right against a wall and you have a sub that can more more than overcome any baffle losses.


Due to the combined sensitivity of the 2 10" drivers, your mid and high frequency drivers either need to be either 4 ohm and ~90dB@1W or better sensitivity, or 8ohm and 93dB@1W sensitivity or better. As far as dome tweeters go, this means you will almost certainly need a 4 Ohm dome, since very few domes are 93dB or higher sensitivity.

Dome ideas:
Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026
Vifa BC25TG15-04 1" Silk Dome Tweeter 264-1040

Mid-range ideas:
Dayton PK165-8 6" Professional Kevlar/Paper Cone Midrange 295-020
Eminence Alpha-6A 6" Midrange 8 Ohm 290-400
Celestion TF0615MR 6" Professional Midrange Speaker 50W 294-2054
Thank you for all the suggestions. So put both 10" woofers on the bottom side by side then the mid above them in the center and then the tweeter above the mid also in the center? Also those mids are not sealed backs is it ok to put them in the same box? The first 2 mids are 100wrms the woofers alone are 160wrms and the amp we have is only 100wrms a channel so I'm really not sure what to do. Originally I was going to setup each 3-way to be 4 ohms so that the amp would be 150 wrms a channel but it looks like there is no way to do that so they have to be ok with 100 wrms at 8 ohms.

sigh, every time I come back to this project I think I have it all set then it all comes crashing down. He really wants to use the large boxes he currently has but they are around 8.31cuft and it's hard to find a couple 10" woofers that need that big of box and are only 100wrms.

He does have a couple 8" 3-way boxes on top of the large ones with mids and tweeters already in them maybe we could just go with a couple 8" woofers. I measured the boxes I think they are around 0.88cuft.
 
Last edited:
Due to the combined sensitivity of the 2 10" drivers, your mid and high frequency drivers either need to be either 4 ohm and ~90dB@1W or better sensitivity, or 8ohm and 93dB@1W sensitivity or better. As far as dome tweeters go, this means you will almost certainly need a 4 Ohm dome, since very few domes are 93dB or higher sensitivity.
What do you mean due to the combined sensitivity? What is the sensitivity after they are wired together? I was considering getting a horn tweeter what do you think about that idea?

This site is only 2-way calculators is there a site for 3-way?
 
Last edited:
I added the Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026 to my cart so just need to decide on a mid range. I don't really like those pro audio mids but yea your right I can't find a mid with the sensitivity we need.

What do you guys think about this crossover design? My dad wanted to go with 4th order because the sub amp also has a 24db crossover.

PRjic.png
 
Last edited:
Every time you double the number of drivers, the sensitivity increases by 3dB. The 10" drivers you have selected, have a sensitivity of about 88dB@1W. When paired, that increases to about 91dB@1W or about 94dB@2.83V. A horn tweeter can be a fine choice, just keep in mind that you generally have to spend a lot more to get a decent quality horn compared to a dome. There are many relatively inexpensive domes that have remarkably flat response and plenty of dynamic range for this build.

A 3 way is nothing more than two 2-way x-overs. The mid-range driver will have both the high pass from the bottom 2-way and the low-pass from the top 2-way wired back to back.

You should download speaker workshop and SPLTrace and read some of the tutorials around the web about how to use them to simulate an x-over.

Or like I said, if you can pick out some drivers, I'd be happy to do a sim for you.

Eric
 
I added the Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026 to my cart so just need to decide on a mid range. I don't really like those pro audio mids but yea your right I can't find a mid with the sensitivity we need.

What do you guys think about this crossover design? My dad wanted to go with 4th order because the sub amp also has a 24db crossover.

PRjic.png

Without simulation there is no way to know if this would work well or not... you could have a great result, or a terrible one...

I don't suggest taking a 5-6" driver up to 6K. I would try to keep that X-over below 3-4K. I would also want to get the X-over from the 10" drivers lower as well. ~500-700hz would be nice but that may not be practical as component costs would be higher.

You may want to look at the cost of inductors before setting your mind on a 4th order network... that network built with air cores could cost more than the drivers easily.
 
Every time you double the number of drivers, the sensitivity increases by 3dB. The 10" drivers you have selected, have a sensitivity of about 88dB@1W. When paired, that increases to about 91dB@1W or about 94dB@2.83V. A horn tweeter can be a fine choice, just keep in mind that you generally have to spend a lot more to get a decent quality horn compared to a dome. There are many relatively inexpensive domes that have remarkably flat response and plenty of dynamic range for this build.

A 3 way is nothing more than two 2-way x-overs. The mid-range driver will have both the high pass from the bottom 2-way and the low-pass from the top 2-way wired back to back.

You should download speaker workshop and SPLTrace and read some of the tutorials around the web about how to use them to simulate an x-over.

Or like I said, if you can pick out some drivers, I'd be happy to do a sim for you.

Eric
Ok so far I have:

(2) Dayton Audio DA270-8 10" Aluminum Cone Woofer 295-334
(1) Vifa BC25SG15-04 1" Shielded Dome Tweeter 264-1026

I can't find a mid range that is 4 ohm and 90db/spl the closest I can find is Goldwood GM-85/4 5" Heavy Duty Sealed Back Midrange 4 Ohm 280-110


I can't go lower on the woofers because the mids only go down to 800hz as far as I can see that is the reason I am building a crossover and not just buying the 700hz one off the shelf. The 3-way calculator I was using would only let me specify one of the frequencies then it auto populated the 2nd so when I put 800 it put 6400 automatically.

Wow inductors are not cheap ok 2nd order it is. Now I see that the inductors that parts-express sells are not exactly matching what I need for example one inductor that I need is 1.13mH now can I just use something close lets say 1.0mH or does it have to be exact? Same with the capacitors I need 35.16uf and closest I can find is 33uf.
 
Last edited:
Cap and Inductors are sold with only particular values available. This is not a problem because slight changes in value won't impact the x-over enough to be a huge concern. More importantly, if you learn to sim the x-over, you can purposely only use only available values in the simulation and therefor have a good idea of response based on component values that are available. You can usually adjust various components in the X-over to compensate for others.

The calc you used probably forces a certain number of octaves between drivers. There is no reason that you HAVE to follow this arrangement. In fact, you can use your mid-range driver to cover just a couple octaves in a custom design, and IMO, will often get a better result by doing so, again, simulation simulation simulation!

buying a pre-made car-audio x-over is going to be just as bad as buying a pre-made "home-audio" x-over. There is no way to know if the results will be even remotely close to flat without simulation.

What is it you don't like about those "pro audio" mids? Those would be superior to that goldwood mid range unit IMO. However, you can technically use the goldwood sealed back 4ohm if you really want to, it's just going to result in the mid-range being a couple dB quieter than the mid-bass/bass stuff... no big deal really, many people run their EQ with a similar arrangement anyways so your father might actually like that. I just don't feel like a $8 sealed back mid-range is going to deliver a low distortion performance.... I think the sealed back design could, on a cheap design, result in some nasty characteristics that are not evidenced by the usual SPL charts and specs.

Eric
 
Cap and Inductors are sold with only particular values available. This is not a problem because slight changes in value won't impact the x-over enough to be a huge concern. More importantly, if you learn to sim the x-over, you can purposely only use only available values in the simulation and therefor have a good idea of response based on component values that are available. You can usually adjust various components in the X-over to compensate for others.

The calc you used probably forces a certain number of octaves between drivers. There is no reason that you HAVE to follow this arrangement. In fact, you can use your mid-range driver to cover just a couple octaves in a custom design, and IMO, will often get a better result by doing so, again, simulation simulation simulation!

buying a pre-made car-audio x-over is going to be just as bad as buying a pre-made "home-audio" x-over. There is no way to know if the results will be even remotely close to flat without simulation.

What is it you don't like about those "pro audio" mids? Those would be superior to that goldwood mid range unit IMO. However, you can technically use the goldwood sealed back 4ohm if you really want to, it's just going to result in the mid-range being a couple dB quieter than the mid-bass/bass stuff... no big deal really, many people run their EQ with a similar arrangement anyways so your father might actually like that. I just don't feel like a $8 sealed back mid-range is going to deliver a low distortion performance.... I think the sealed back design could, on a cheap design, result in some nasty characteristics that are not evidenced by the usual SPL charts and specs.

Eric
I was going with the goldwood because its less wattage than the ones you posted and sealed because the ones you posted were not sealed don't I have to make some sort of separate enclosure for them? He only has 100wrms @ 8 ohms / 150wrms @ 4 ohms to work with so having the woofers which are 160wrms and a mid that's 100wrms seemed like way too much. I need to get that program you keep talking about to simulate it is it free?
 
Pretend like watts and wattage ratings don't exist for this project. It is practically irrelevant. Now, read that again. Seriously, ignore watts! If you have a configuration idea that has a "watts" problem I'm sure that someone here or myself will speak up on the matter.

There is nothing wrong with using a 1000W amp to power a 10W rated speaker, and there is nothing wrong with using a 10W amp to power a 1000W rated speaker. Please ignore the watts!

Any non-sealed back design will need it's own enclosure.

speaker workshop is free, just google it. I am running it under linux with wine (windows compatibility layer) and it runs well enough. It's a little tricky to get used to but, I'd only been using it for about 2 weeks and I feel pretty comfortable with it now.

Eric
 
If a driver is rated/measured ~90dB sensitivity @ 1W, then it doesn't matter if it is rated 1W, 10W, or 1000W, it will produce ~90dB at 1W. Assuming it is capable of handling 1000W, then at 10W, it will be 100dB, at 100W it will be 110dB, and at 1000W it will be 120dB. (assuming no power compression, but you should ignore this concept for now).
 
If a driver is rated/measured ~90dB sensitivity @ 1W, then it doesn't matter if it is rated 1W, 10W, or 1000W, it will produce ~90dB at 1W. Assuming it is capable of handling 1000W, then at 10W, it will be 100dB, at 100W it will be 110dB, and at 1000W it will be 120dB. (assuming no power compression, but you should ignore this concept for now).
Wow ok I always wondered why it gave ratings for spl that way that is crazy ok then I guess it wont matter too much then. So far I am going with this mid Celestion TF0615MR 6" Professional Midrange Speaker 50W 294-2054 but I would like to find a 4 ohm to match the tweeter and the woofer load and cheaper if possible my dad is on a tight budget and is already having to charge this all.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.