more important to match Q or box resonance when driver matching is less than ideal?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Just another Moderator
Joined 2003
Paid Member
OK the reason for the question is that I’ve measured my drivers (Morel MW144 X 4) and have a quite a large variance in T/S parameters resulting in (what I think) is a big variance in optimum enclosure volumes for each of the four drivers. What I’m unsure of is wether I should be optimizing the enclosure volumes to get as close to the same system Q for each driver, or as close to the same response curve for each driver. I haven't made sealed enclosures before and maybe I’m being overly paranoid ;)

These drivers are going to be crossed actively at between 200 and 300Hz so won’t be going too low (haven't decided on what order yet).

The speakers will be MTM with separate cavities for each midbass and the tweeter.

Unixbox’s “standard design” recommendation for each midbass is below.

Midbass_1 5.2L
Midbass_2 5.8L
Midbass_3 6.9L
Midbass_4 8.2L

1. Considering that these will only be going down to around 200Hz (without considering lower freqencies the driver will be producing depending on xover slope) Can I safely take the average of the optimum enclosure volumes and put each driver in a non optimum average size enclosure without too much affect on sound quality

2. Should I try and match the box Q’s to get them all as close to .7 (which is what I'm aiming for) as possible (I’m thinking different volume for top and bottom cavities, but making these the same for each speaker and varying stuffing to suit the individual drivers (one cavity optimised for midbass_1 and one cavity optimised for midbass_3) then increasing effective volume for 2 and 4 by using extra fill.

3. Is having similar box resonant frequencies (and hence more similar response curves) more important than matching the box Q’s?

4. Am I worrying about nothing!!! :) Considering that plugging in different volumes in Unibox seems to give minor variations like .2 or .3 db bumps or dips.

For those interested below are T/S details….

Measurements below were done in speaker workshop, and were done twice, once before and once after breaking in the drivers (approx 2 hours playing pink noise, approximately 2 hours playing frequencies between 50 and 55Hz and approximately ½ hour of playing 25Hz). The difference between before and after breaking in was minimal (figures shown are after break-in). I must say I was a bit disappointed that there was such a large variance between drivers, especially since I had been told that Morel had a very good reputation for driver consistency.

Measurements were taken using 24Khz sampling rate and VAS was calculated by adding a 20c coin scotch taped to the center of the midbass dustcaps (weight 11.7g). I also took impedance curves at 96Khz sampling rate but decided to use 24Khz as it was the lowest sampling rate that gave me clean curves, and has a resolution of .09 Hz compared to .37Hz resolution for 96Khz (due to SW’s fixed upper limit of 256K on sample size). I haven’t done a sanity check to see whether the ratios are the same when measured at 96Khz compared to 24Khz, but the T/S parameters are definitely different.

TS parameters below:

Midbass 1: FS 52.38
Vas 9.73
Qms 2.15
Qes 0.59
Re 5.28
Le 0.20


Midbass 2: Fs 52.06Hz
Vas 9.79L
Qms 2.21
Qes 0.55
Re 5.28
Le 0.20

Midbass 3: FS 50.21
Vas 10.61
Qms 2.29
Qes 0.56
Re 5.18
Le 0.20

Midbass 4: FS 54.57
Vas 8.842
Qms 2.53
Qes 0.62
Re 5.27
Le 0.20

I've attached a pic of one of the drivers mounted for testing.


Any input greatly appreciated :)

Tony.
 

Attachments

  • 280_8026.jpg
    280_8026.jpg
    52.8 KB · Views: 115
I pick answer #4.

If these are going to be actively crossed that high, (or even if they weren’t) I wouldn’t give the differences much thought. As I model them, they appear to have very small variations in frequency response as a consequence, even is driver #1 is modeled in driver #4 enclosure.

If it were me, I’d probably pick the mean volume and make them all that size, and worry about something potentially more important, like how to best mitigate the baffle edge diffraction effects.

C
 
Just another Moderator
Joined 2003
Paid Member
Thanks C.

I was kinda thinking that may be the response :) I tend to agonise over very minor details sometimes, although I must admit I was kinda suprised about the fact that the optimum enclosure size for driver four was almost 60% bigger than driver one and was thinking it wasn't exactly a minor detail ;). But as you say they model pretty close in frequency response when put in the same volume, so that is probably all that matters.

Tony.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.