Help with Bass/Mid Selection for Small Box

As a relative newcomer to speaker design I am looking for suitable drivers for a small 2-way.
Actually the speakers will be used with a 2nd order electronic crossover with a low pass frequency of 120Hz and I'm targeting a 2nd order low pass for the speakers themselves in the range 80Hz to 120Hz i.e. closed box. Maximum baffle width is 200mm and target volume is 9L for a rectangular box or 7.5L for a sloping baffle version.

The Satori range of drivers looks like a good mid-price solution. The frequently used MW16P actually goes a bit to deep for my application!
So I'm looking at the MW13P-8 or MR13P-8 along with the TW29R tweeter.

How do I choose between the MW and MR?
Any suggestions for drivers that might suit my application better?


2019-02-09 10:28 pm
I guess you're planning to use it with a sub?

So what you basically want is a 5-6" midbass that will work in a 10L closed box.

What tweeter are you planning to use?

You might still be able to find a NE180W somewhere; Seas U18RNX/P also seems to fit.

You should model any shortlist woofers for closed box and see which ones meet requirements for expected QTC and Vb.

Quick and dirty for QTC 0.707, NE180W has Vb = 10L, F3 = 77hz, U18 Vb = 8L F3 = 82hz
Last edited:


2019-02-09 10:28 pm
MR13 is only at 3L for QTC 0.7 F3 = 103hz so doesn't look suitable.

MR16 is better with Vb 10L F3=75hz.
MW16 wants a bigger box at 13L for QTC 0.7.
Scan 18W 8434 Vb 11L F3 = 82hz

Think there are some NE180Ws floating around on ebay for much cheaper than the Satori, though.
Last edited:
Planning to use SB Acoustics TW29R tweeter.
Yes the 2-way will be used with dual subs.

According to this calculator, using measured values from Audioexcite, the MR13P-8 works OK in a 7L box - maybe a bit overdamped, am I wrong in thinking this is not an issue?
Vas = 7 L
Qts = 0.46
Qes = 0.49
fs = 63 Hz
The result is f3 = 97.5Hz and Qtc = 0.651

For the MW13P-8, the figures are:
Vas = 12.5 L
Qts = 0.35
Qes = 0.39
fs = 44 Hz
The result is f3 = 92Hz and Qtc = 0.584

Data from here:


2019-02-09 10:28 pm
It depends how much you mind about humps above or dips below flat, and your room response, speaker placement, what the subs are doing.

In a 2-way, some prefer a QTC of > 0.7 to say, 0.9, for warmth; some would say under 0.7 and it's going to sound thin.

Not sure if this changes when used as pure sats.
Last edited:
I use the natural 2nd order rollof of the 2-way (let's call it X Hz) + active LR2 low pass at 120Hz. This "matches" with active LR2 at 120Hz and another active LR2 at X Hz.
For my current speakers with CA15RLY drivers X is 80Hz (measured -6dB point).

Plugging the CA15RLY numbers into the calculator gives a -6dB point of 70Hz and Qtc = 0.576

I realise that the above is not a true LR4, but sims show that the phase sums correcty and the amplitude error is <1dB at 100Hz. This is probably irrelevant when room gain is considered. And it sounds very good!

The simulated -6dB point with the MW13P-8 is 67.5Hz and Qtc = 0.584. This is very similar to the CA15RLY so I think I've answered my own question!!

I've also thought about the MR version and it looks to me like it has a shorter voice coil with finer gauge wiring (less power handling, less cone travel and lower inductance). This does improve the bandwidth slightly, but oddly the off-axis response is not as good as the MW.

So the WM13P-8 is looking like the favourite... unless somebody tells me to look elsewhere!?!
FWIW, other drivers I've considered:
Monacor SPH145-HQ (slightly concerned by the 3rd HD shown in one of the pics here: (click the right arrow on the pics to scroll through the measurements)
SS8530K00/01 (no need for all that bass extension and Xmax)
Faital 5FE120 (unknown distortion preformance)

Just read your caption "Don't procrastinate..." That's me at the moment :)

Actually I have just noticed the textremely expensive version:

This seems to have less HF breakup that the MW13P. I might be able to get away without a notch filter with the TX.
I'm aiming for a time aligned LR2 design. 2.5KHz crossover seems to "fit" the MW13 and TW29 according to my sims.
Last edited:
When crossing at 120Hz, a lot of discussion about tuning to the right Qtc becomes moot. Rather target for a clean and neat highpass function by combining the acoustic and the electric slopes. Choosing the optimum box volume with that in mind is quite different from choosing the acoustic suspension tuning alone.
Last edited:
I'm now wondering if it wuold be better to look at "smooth" drive units rather than the latest super-wide bandwidth units.

This has led me to the Vifa PL14WJ09-08, SS 15W8424G00 and SB13PFCR25-8.

These inexpensive drive units look easier to integrate with a nice tweeter than the expensive stuff. Obviously they don't have the bass perfomance of the expensive variants - but I don't need this.

Have I lost the plot? Please tell me what's wrong with these drivers.


2019-02-09 10:28 pm
I find the shape of the SB13PFC ugly and it has a plastic frame compared to most competitors.

Smooth response drivers are just easier to work with.

Certain drivers will have been used in published designs, so you can crib good XO points, schematics and complementary tweeters from those.

Otherwise, one can look through datasheets and independent measurements (sensitivity, on and off-axis responses, distortion plots, breakup points). Or take a view on cone material (are you a paper, coated paper, poly or metal person).

My impression is that you want some perfect decision-making process, so you could call that overthinking cause all drivers have compromises.

Although you are high-passing these as sealed box sats, I'd still eyeball closed box sims in WinISD or whatever to see if a particular volume causes any peaks above 100hz or so depending on desired Qtc.

Draw up your shortlist, decide how much you want to spend, eliminate from there based on the criteria above.

I don't think you'd go wrong with the 15W Discovery - there's also the 8434 version.
  • Like
Reactions: 1 user
The 15W8424 looks a touch smoother than the 8434. I don't know what the physical difference is.

When I compare the fabled 15W8530K01 to the 15W8424, the 8530 goes deeper than I need, has low efficiency and has a nasty bump at 2KHz which exacerbates the baffle step bump with my 200m wide baffle. The combined bump at 2K really doesn't help with crossover integration.

I just wanted somebody with more experience to confirm that my thinking was along the right lines. It's unusual for a cheaper part to do the job better (unless we are talking "audiophile" components).


2019-02-09 10:28 pm
You haven't given a clear definition what doing the job better is.

One pays for a Revelator for prodigious bass, higher power handling than e.g. 40w of the SB13PFC, high transparency, trading off sensitivity and more filtering.

8434 seems to have a bigger magnet but otherwise seems equivalent if used as a sat.

If we can be caveman about sensitivity uber alles, then I still think why not go for a 6-7" (with smooth response, reasonable price); it will still fit your baffle width constraint. They will also still be "alright" when used as standalone closed speakers (unless you really want a tiny box).

If you really want a smaller box and easy build, well, you could even go back and look at small fullranges.
Last edited:
I'm not 100% sure what a better job is. I suppose I'm hoping to make a better speaker than my CAOW1s.

Somebody suggested the best way to do this is to swap the CA15RLY for a 15W8530, which then means I'm heading for something like Paul Carmody's Carrera. But I don't need that bass capability, so I'm wondering if I can get better midrange.... and back round the loop I go.

I was thinking of staying with 5" because it should have better off axis response in the crossover region than 6-7".

My only box limitation is 200mm maximum width. That's the width of my bass extenders, which double as the speaker stands.
If you want, you can create a spreadsheet with Vb/F3 and for closed boxes Qtc to get a rough idea using standard formulas using only Fs/Qts/Vas as input, here's the drivers I've loaded and their theoretical alignments under 11L Qtc 0.7 volume.

Quicker to get a feel than having to load up WinISD etc...