Floorstanding 3-way speaker: woofer choice for a newbie

Hi,

I’m still working on my 2-ways, but in my “spare time” 🙂 I’m thinking about a floorstanding 3-way speaker for “general music listening”.

Concept so far:
  • Floorstanding tower shape.
  • Sealed or ported, no preference really
  • Cabinet size can be large, but not coffin-size large.
  • Should have a really good low-end. I will not be adding a subwoofer.
  • Woofer 8'' or 10'' probably. Need to keep the baffle width acceptable.
  • Passive XO
  • Crossover woofer at BSC range
  • Crossover midrange 3k or 4k or so
  • Sensitivity doesn't need to be super high for the ensemble, I'm probably not going to double the W to go super high-sensitivity.
Because the low end is important, I wanted to start with the the woofer selection. The more I read and look at graphs, the more I’m confused.

These are some candidates I found:
  • SEAS Prestige L26RFX/P
  • SEAS Prestige L22RN4X/P
  • SEAS Prestige CD22RN4X
  • Scanspeak 25w/8565
  • Scanspeak 26w/4558t00

Sensitivity: not much difference between them. 88dB, 89dB,..

Material: alu or paper or some paper-hybrid. Word on the street is "alu is more precise, clinical. Paper is warmer." Even if that is true, does it matter with a low-crossed woofer? I can still use a paper midrange.

LF response: the SPL of some of these drops faster than others. E.g. CD22RN4X (only 8 inch). But others remaing stronger for longer like the L26RFX and the 25w/8565.

The 25w/8565 is a bit rocky in the low end though. Is that audible?

1661247537536.png


So given my concept, how on earth do I choose between these drivers? If I were to search more, I'd probably find even more drivers than seem to fit my wishes.

They are too expensive to just buy and test.

Thank you
Vincent
 
There are quite a lot of candidates actually. Most of us start the narrowing down by simulating the low end response in an app like WinISD or Basta! Other apps will do fine too. But... and don't mind me saying this, but did you consider how to design the whole system? Like 'designing an enclosure that is near optimal, measuring real output from drivers and designing an appropriate crossover'? In other words, do you have the knowledge and means to fulfil the task or are you willing to learn on the job? If not, my advice would be: build a proven design. The OSMC on this forum is a good design. But there are plenty of others.
 
There are quite a lot of candidates actually. Most of us start the narrowing down by simulating the low end response in an app like WinISD or Basta! Other apps will do fine too. But... and don't mind me saying this, but did you consider how to design the whole system? Like 'designing an enclosure that is near optimal, measuring real output from drivers and designing an appropriate crossover'? In other words, do you have the knowledge and means to fulfil the task or are you willing to learn on the job? If not, my advice would be: build a proven design. The OSMC on this forum is a good design. But there are plenty of others.
I don't mind and it's a just question.

I'm no expert on anything related to speaker building. But I did finish a two-way speaker (only a two-way) with a self designed crossover, which looks pretty good in REW and the speaker sounds pleasing to me. Took me some time, but I can now say that I can make usable and repeatable REW measurements. The crossovers I created in VituixCAD and then measured in reality were almost identical to what was simulated.

Disclaimer: I'm using XMachina for my initial crossover. While I always do tweaks myself, mostly because I don't have this coil or that inductor in my box o' components, it really helps me with getting started. I'm a bit surprised there's not more talk about this tool, it's fantastic.
 
A few ideas - look at measurements like from hificompass or dibirama, search for threads on here that mention the drivers on your shortlist, look for published designs using said drivers, check the simulated bass capabilities (the enclosure tool in vituixcad has a bunch of driver data in its database) sealed and ported. That should narrow things down. Also, do you not have a preference for cone material? I generally like coated paper and poly - of which there are tons - and therefore don't really bother with aluminum, at least when creating shortlists in my head.

Then you'd have to do a similar exercise for a mid and tweeter...then look again if these seem to make sense all together.

Look at Troels' various 3 way classics - he has one each for Seas, Scanspeak, SB Acoustics, and more. They could be made floorstanding by closing off the original box and just extending down to the floor.
 
You too wrong with the SB WO24P woofers. If budget it tight the 29NRX-8 (not the 6 used in Jeff Bagby’s Kairos 3way bass module) gives almost the same performance in the same box size with this a bit more overhead and probably a touch more distortion.
 
  • Like
Reactions: Vinne81
What is the reason you don't like aluminium? I ask because I want to learn.

I don't have preference for cone material at this time, because I'm too new to this. But from what I read, aluminium cones are clearer (too clear/clinical?) and have tighter bass. They also have a nastier breakup. But this won't be a problem in a 3-way as far as my woofer is concerned.

I have found another candidate, the Dayton DSA270-8. Modelled 63 liter vented, f3 34Hz. But it's alumium. On a positive note,, it's cheap. Unfortunately it also has a thin frame and only 5 mounting holes 🙂 Sensitivity is listed as 90.4dB, which is a plus compared to the others.
 
Last edited:
You too wrong with the SB WO24P woofers. If budget it tight the 29NRX-8 (not the 6 used in Jeff Bagby’s Kairos 3way bass module) gives almost the same performance in the same box size with this a bit more overhead and probably a touch more distortion.

This the she SB WO24P. When I look at this SPL chart I think that a flat response would be really hard to obtain with this woofer. That's a drop of 5dB. Isn't that a problem?

Edit: probably because of the IEC baffle measurement or so.

1661276879221.png


It looks better here: https://hificompass.com/en/speakers/measurements/satori/satori-wo24p-8

When simulation this driver in an enclosure, it looks very impressive. If the non-flat response in the graph above has an explanation then this woofer might be very high on the list!
 
Last edited:
Drivers are measured on standard baffle. This makes for FR variations. If you are worried about resonances, these would show as a blip in the impedance trace. Checking out som other sizes and cone material, shows that it s a common feature. Common bad design from a top producer or a common measuring aspect.... I think you can safely ignore that.

//
 

Attachments

  • SB23MFCL45-8.jpg
    SB23MFCL45-8.jpg
    86.6 KB · Views: 42
  • MW19P-8.jpg
    MW19P-8.jpg
    82.6 KB · Views: 46
  • Like
Reactions: Vinne81
I'll probably go for the SB WO24P, which looks great and is still affordable. To not go the all-satori way (I need to save some money for gas, I'm from Europe), I was thinking about choosing the MCA15RCY as my midrange. Any thoughts on that? It has nice response dips around natural crossover points and seems like a quality midrange.
 
I don't know your depth of expertise, budget or listening preferences, but have you considered an existing and proved design? For example, Paul Carmody's "Tarkus" 3 ways have Peerless and Vifa drivers, readily available in Europe: Peerless 10" woofer (830668) and 6" mid (830657) and Vifa DX25 tweeter.

https://sites.google.com/site/undefinition/floorstanding-speakers/tarkus

Many builds, many happy listeners. Although designed as a twin-cabinet design, some people have made it on one cabinet, there are photos at the bottom of the web page.

Geoff
 
What is the reason you don't like aluminium? I ask because I want to learn.

I don't have preference for cone material at this time, because I'm too new to this. But from what I read, aluminium cones are clearer (too clear/clinical?) and have tighter bass. They also have a nastier breakup. But this won't be a problem in a 3-way as far as my woofer is concerned.

I have found another candidate, the Dayton DSA270-8. Modelled 63 liter vented, f3 34Hz. But it's alumium. On a positive note,, it's cheap. Unfortunately it also has a thin frame and only 5 mounting holes 🙂 Sensitivity is listed as 90.4dB, which is a plus compared to the others.
paper cone is natural sounding. many purist only like paper cones.
metal mids i would avoid them!
 
  • Like
Reactions: Gill.T
Metal cones can sound natural if the resonances are managed with the croosover.

I will follow the good advice above about the Tarkus.

Be aware when you look the datasheet the big normalized bafle is nearer for the low end from a 2 Pi radiation i.e. infinite baffle than a woof on a small cabinet that radiate 4 Pi in the low : substract few dB, add few but very less for the room gain and anyway you will see you will not be at 89 dB with current hifi drivers but more 85 dB or less imho. The good news is the hifi drivers not having big efficienty go low.
According your roadmap, a good bet is a woofer with the lowest Fs you can and the max sensivity, with a Qts around 0,4 for a BR load; avoiding metal because it is harder and more expensive to fix the metal resonance. Notice some paper can have resonance too.
 
I have found another candidate, the Dayton DSA270-8. Modelled 63 liter vented, f3 34Hz. But it's alumium. On a positive note,, it's cheap. Unfortunately it also has a thin frame and only 5 mounting holes 🙂 Sensitivity is listed as 90.4dB, which is a plus compared to the others.

So? No big deal since with rare exception all woofers should be braced/preloaded to a sufficiently rigid cab (Fs ~ > an octave higher), ergo only needs two location pins or ideally not rigidly mounted to the baffle like pioneers originally did.
 
The Dayton rs-225 metal black cone is smooth till relativly high frequencies for instance. There are many designs with two of it as it its spl is very low, but cheap too according the part of the world you'live in.
 
Last edited:
  • Like
Reactions: krivium
  • Like
Reactions: krivium and GM