• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Help me to design a stereo pair of 3-way speakers from scratch (well, almost from scratch)

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
I've just acquired a pair of Seas Excel W26FX001 10" drivers, which I want to use as the basis for a new pair of 3-ways.

In terms of general design objectives:

1. I'll use these with some form of active DSP, but I'll leave the decision whether that's going to be a pre-pro/multichannel amplifier or a Hypex plate amp arrangement until later.

2. I'm open to creative cabinet designs as long as I can make them and they're up to WAF standards. Some aspects of the cabinet design will emerge as I make the other design decisions.

3. They'll be used for stereo listening only (no HT) and I have very varied tastes. I tend to gravitate towards good recording quality.

4. My listening room is 15' wide x 22' long x 8' high. The speakers will be 8' to 10' apart and 3' to 4' away from the end wall. My listening arrangement is the classic isosceles triangle. Of course, this may change in the future.

5. In addition to my current active DSP speakers, I have a pair of passive 3-ways based on Dennis Murphy's MBOW1 design. This design uses an OW1 tweeter, GR Research M130 midrange, and Peerless 850146 10" woofer in a sealed box. I like this speaker, but would like bass that's a little tighter and deeper. I could (probably) replace the Peerless woofers with my newly acquired Seas woofers, but where's the fun in that?

Unibox tells me that the Seas woofers should give me an F3 of 41 to 42 Hz in a sealed box of 45 to 66 litres. This seems like very wide volume range, but that's what I get. I'll go with the bottom end of that range, so 45 to 50 litres, in order to keep the cabinets relatively small. This F3 is about 5 dB lower than the slightly more sensitive Peerless woofers which are in about 50 litres.

The Peerless woofer is a good performer and was excellent value for money, the M130 midrange was terrific value at its price, and the OW1 is a well-respected tweeter, but perhaps we can achieve a modest improvement? Replacing the Peerless with the Seas and then converting the speakers to DSP may be the way to go, but I'd like to explore the potential for a new design first.

My preliminary thoughts for mids and tweets are the Seas Prestige MCA12RC mid and the Vifa/Peerless XT25TG-30-04. Both are available at modest prices. If we want a mid that's a visual match with the woofer, there's the Excel W12CY001, and if we want to go all-Seas there's the Prestige 27TDFC tweeter. These options add significantly to the budget, but that's acceptable if they will produce a better result.

I must emphasise that my knowledge of how to pick compatible drivers is thin, so my first objective in this thread is to ask for guidance on how to do this, whether any of these are good choices, and whether there are better ones. Price is always a consideration, but not necessarily a show-stopper within reason. I prefer to avoid the more "exotic" stuff.

So there we are. . . your ideas, please!
 

DVDdoug

Major Contributor
Joined
May 27, 2021
Messages
3,016
Likes
3,966
Unibox tells me that the Seas woofers should give me an F3 of 41 to 42 Hz in a sealed box of 45 to 66 litres. This seems like very wide volume range,
Sealed boxes aren't as finicky as ported designs.

I like this speaker, but would like bass that's a little tighter and deeper.
UniBox should allow you to optimize that. With a sealed box and a given driver, box volume is pretty-much the only variable. (Plus whatever you can do with DSP.)

My preliminary thoughts for mids and tweets are the Seas Prestige MCA12RC mid and the Vifa/Peerless XT25TG-30-04.
Of course the frequency ranges should overlap so you can get good response through crossover frequencies. And they should have approximately the same sensitivity. If the sensitivity is too-high you can use a resistor/pad in your crossover but the last thing you'd want to do is pad-down the woofer. Or if you are talking about tri-amping, sensitivity matching isn't that important. (I wasn't clear what you were going to do with the multi-channel amp so I was guessing tri-amping.)
 

ppataki

Major Contributor
Joined
Aug 7, 2019
Messages
1,241
Likes
1,389
Location
Budapest
I have a pair of passive 3-ways based on Dennis Murphy's MBOW1 design. This design uses an OW1 tweeter, GR Research M130 midrange, and Peerless 850146 10" woofer in a sealed box. I like this speaker, but would like bass that's a little tighter and deeper.
Since it is a sealed box, have you tried adequate DSP to get that deeper bass?

Are you using Dirac Live or equivalent?

Based on measurements made at the main listening position I would apply a low shelf filter first (playing with the frequency and the Q value to get the results you want)

You can actually get mind-blowing results with sealed boxes when using simple low shelf filters. And you can even combine that with Dirac Live (or equivalent) the results will be stellar
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
Sealed boxes aren't as finicky as ported designs.


UniBox should allow you to optimize that. With a sealed box and a given driver, box volume is pretty-much the only variable. (Plus whatever you can do with DSP.)


Of course the frequency ranges should overlap so you can get good response through crossover frequencies. And they should have approximately the same sensitivity. If the sensitivity is too-high you can use a resistor/pad in your crossover but the last thing you'd want to do is pad-down the woofer. Or if you are talking about tri-amping, sensitivity matching isn't that important. (I wasn't clear what you were going to do with the multi-channel amp so I was guessing tri-amping.)
According to Unibox, the F3 and step response are the same from 42 to 55 litres. This range makes a difference of about 1 dB at 30 Hz, which I wouldn't hear even if there were anything down there. Still, it would be fairly easy to throw a sandbag in there to see what happens.

The published FRs for the drivers I mentioned suggest XOs of around 400-500 Hz and as high as 4 kHz for the MCA12 (less for the W12CY) based on the rule-of-thumb of an FR being flat for one octave beyond the XO frequency. The sensitivities seem fine, although the tweets are higher than need be. Three channels of amplification per speaker, either by a remote six-channel amplifier or by 3-way plate amps (or active analogue crossovers).
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
Since it is a sealed box, have you tried adequate DSP to get that deeper bass?

Are you using Dirac Live or equivalent?

Based on measurements made at the main listening position I would apply a low shelf filter first (playing with the frequency and the Q value to get the results you want)

You can actually get mind-blowing results with sealed boxes when using simple low shelf filters. And you can even combine that with Dirac Live (or equivalent) the results will be stellar
I haven't tried using DSP crossovers instead of the current passives yet. There's a limit to how many non-passive speakers I can deal with, but it's on the list as an experiment. My assumption is that DSP will bring an improvement anyway; the first question is whether I can improve on the basis design, particularly the drivers and their arrangement.

No room correction yet (but it's coming). This is another variable which I'd rather not introduce to the current question yet - optimise the design of the speaker first. Good comment about using a shelf filter though - I've added a note to my list.
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
I have three initial choices:

1. Don't bother. Just re-sell the Seas woofers.

2. Replace the Peerless woofers in my current speakers with the Seas. The passive crossovers would need reworking (which I'm not going to attempt), so I would bypass them and go DSP - in which case it would make sense to try DSP with the current arrangement first as @ppataki suggests.

3. Build a new pair of speakers, in which case I need to choose mids and tweets. This option is obviously much more fun, but is viable only if the end result improves on the current speakers (when converted to DSP).
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,201
Location
Northern Virginia, USA
Pertinent thread:

 

ppataki

Major Contributor
Joined
Aug 7, 2019
Messages
1,241
Likes
1,389
Location
Budapest
The passive crossovers would need reworking (which I'm not going to attempt), so I would bypass them and go DSP - in which case it would make sense to try DSP with the current arrangement first as @ppataki suggests.

...and even before that you could try a low-shelf filter to shape the low-end of your current system (with the passive crossover)
You might get surprised what a difference a simple low-shelf filter (when playing around with all parameters - frequency, gain, Q) can produce

and if you are still not satisfied you could convert from passive crossover to fully active crossover (using your current drivers)

and then you can start swapping drivers - if needed

In any case I would very highly recommend getting Dirac Live or equivalent (it is free for 14 days, nothing to lose) - that will make a huuuuge difference even with your current passive system (been there, done that.....)
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
...and even before that you could try a low-shelf filter to shape the low-end of your current system (with the passive crossover)
You might get surprised what a difference a simple low-shelf filter (when playing around with all parameters - frequency, gain, Q) can produce

and if you are still not satisfied you could convert from passive crossover to fully active crossover (using your current drivers)

and then you can start swapping drivers - if needed

In any case I would very highly recommend getting Dirac Live or equivalent (it is free for 14 days, nothing to lose) - that will make a huuuuge difference even with your current passive system (been there, done that.....)
Unfortunately, my DSP box is on its way back to the manufacturer to determine whether it can be repaired and I don't use any computer-based software that will work with Dirac. Quick and easy experiment, but it's going to have to wait several months.
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,151
Likes
4,838
Location
Portland, OR, USA
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
You could rework a Seas active design with the W26 woofers instead of the L26 units:
The W26 would likely work very well in similar sealed box, you could use Seas' DSP filters for the midrange and tweeter, just have to rework the woofer's filters. And not exceed maximum excursion since the W26 is not happy overdriven.
This is a very interesting idea. I understand the obvious benefit of having the mid and tweeter co-axial, but isn't there the inevitable trade-off? I seem to recall that co-axial drivers are frowned on. . .

While exploring this idea further, I found this: https://www.kartesian-acoustic.com/copie-de-cox120-vhp. The response and distortion curves are better than for the Seas, although they're the company's figures, and they're cheaper too (although this may not be the case if they have to come from France).

This idea could work very well for me, ceteris paribus.
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,151
Likes
4,838
Location
Portland, OR, USA
This is a very interesting idea. I understand the obvious benefit of having the mid and tweeter co-axial, but isn't there the inevitable trade-off? I seem to recall that co-axial drivers are frowned on. . .

While exploring this idea further, I found this: https://www.kartesian-acoustic.com/copie-de-cox120-vhp. The response and distortion curves are better than for the Seas, although they're the company's figures, and they're cheaper too (although this may not be the case if they have to come from France).

This idea could work very well for me, ceteris paribus.
Coaxial can be great. There are plenty of poor coaxial implementations. This is a pretty good one. You will need DSP. It is slightly pricey.
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
Coaxial can be great. There are plenty of poor coaxial implementations. This is a pretty good one. You will need DSP. It is slightly pricey.
The Seas is is indeed expensive: US$553 at Madisound and nearly twice that in Canada. The Kartesian is much more reasonable.
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,151
Likes
4,838
Location
Portland, OR, USA
The Seas is is indeed expensive: US$553 at Madisound and nearly twice that in Canada. The Kartesian is much more reasonable.
Ouch, I haven't checked the price lately...
It's one of the few concentric drivers available to consumers I would consider. Mostly out of ignorance of other units.
Perhaps a simple midrange and tweeter might be better.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,201
Location
Northern Virginia, USA
All the SEAS drivers have risen dramatically. My W18s have more or less doubled since I bought them.
 
OP
O

Old Hi-Fi Guy

Active Member
Forum Donor
Joined
May 3, 2020
Messages
103
Likes
52
Location
Western Canada
At the moment, I'm leaning towards the MCA12 and 27TDFC tweet. This tweet has lower distortion and slightly better off-axis performance, although the latter isn't very important in my room. For the mid, the W12CY001 is billed as having a particularly stiff cone for precise bass, but it's also billed as a mid/woofer. Since I will using a separate woofer, I don't need this bottom-end performance and the big, heavy magnet that comes with it. This mid weighs 50% more than the MCA12 and costs 2.4 times as much. The MCA12 has a lower moving mass, smaller X-max, and stiffer suspension, which I interpret as making it more responsive in this critical midrange. The FR is also flatter for longer and the power handling is better. If I'm out to lunch on my interpretation of the numbers, do please let me know.
 

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,151
Likes
4,838
Location
Portland, OR, USA
All the SEAS drivers have risen dramatically. My W18s have more or less doubled since I bought them.
Yeah, a bit hard to take.
They are good drivers, but the prices got absurd. I got my W26 and L26 woofers for around $200 each back in the day, that was expensive but are $600 now.:facepalm:
I'm probably done with buying these overly-pricey Seas drivers.
 

fpitas

Master Contributor
Forum Donor
Joined
Jul 7, 2022
Messages
9,885
Likes
14,201
Location
Northern Virginia, USA
Yeah, a bit hard to take.
They are good drivers, but the prices got absurd. I got my W26 and L26 woofers for around $200 each back in the day, that was expensive but are $600 now.:facepalm:
I'm probably done with buying these overly-pricey Seas drivers.
Hopefully mine will continue working and I'll never need drivers again.
 
  • Like
Reactions: MAB
Top Bottom