When I was working on a new passive crossover for my three-way speakers, I found a surprising peak in response at 230 Hz. I measured the impedance of my Audax HM210Z0 bass drivers in the cabinet, and found a peak at this frequency, which I ascribed to a longitudinal resonance in the box, whose internal height of 72 cm would match this pretty much perfectly.
Aha, I thought, I can tackle this resonance by putting in a couple of extra horizontal shelves in the cabinet to break up this mode. When I measured the impedance again, I found that the peak at 230 Hz had vanished, but to my puzzlement a new one had appeared at 188 Hz. Surely a cavity with its largest dimension 72 cm cannot support a mode of this frequency (half-wavelength 90cm)? Could it be some kind of interaction with either the driver suspension or the port?
Here is a plot of the old and new impedance curves, with the Audax data as a reference:
The second peak at 425 Hz is relatively unaffected by the extra shelves, and may well be a front-back resonance (the half-wavelength is 40 cm, which is consistent with that).
Alex
Aha, I thought, I can tackle this resonance by putting in a couple of extra horizontal shelves in the cabinet to break up this mode. When I measured the impedance again, I found that the peak at 230 Hz had vanished, but to my puzzlement a new one had appeared at 188 Hz. Surely a cavity with its largest dimension 72 cm cannot support a mode of this frequency (half-wavelength 90cm)? Could it be some kind of interaction with either the driver suspension or the port?
Here is a plot of the old and new impedance curves, with the Audax data as a reference:
An externally hosted image should be here but it was not working when we last tested it.
The second peak at 425 Hz is relatively unaffected by the extra shelves, and may well be a front-back resonance (the half-wavelength is 40 cm, which is consistent with that).
Alex
I can certainly try that, but I always thought that the accepted wisdom was to avoid damping a ported system?
Alex
200-400Hz is such an important region, I'd rather compromise in favour of this region. The damping will have a greater effect up here than down at the tuning frequency so I'd try to find just the right amount to use. I wouldn't be afraid to stuff a vented box if I had a good enough reason like this.
An old rule of thumb, useful in the days before widespread availability of decent test equipment and software to avoid misalignment of system response. Now you can see what you're doing, not a great issue.
Vented boxes should not be fully stuffed but they will suffer from internal resonances and need at least a decent amount of wall lining. Try a thickness of approximately 1/5 each dimension on each wall (leaving 3/5 clear in the middle). Note that Dacron is useless for damping standing waves! Fiberglass, wool or felt are probably the best.
Are you sure these are length related standing waves? There is also the possibility that cabinet wall resonances are giving you problems. Its a little difficult to tell one from the other but give the cabinet a slow sweep while lightly touching the various cabinet surfaces. See if you feel a tingling vibration at the resonance frequencies (also whether the woofer chassis has taken off (vibrates) at the same frequencies. This would suggest more that it is a mechanical resonance.
If it is an acoustical resonance then the internal damping should cure it. If mechanical, then see the recent thread on best cabinet materials.
Regards,
David S.
Are you sure these are length related standing waves? There is also the possibility that cabinet wall resonances are giving you problems. Its a little difficult to tell one from the other but give the cabinet a slow sweep while lightly touching the various cabinet surfaces. See if you feel a tingling vibration at the resonance frequencies (also whether the woofer chassis has taken off (vibrates) at the same frequencies. This would suggest more that it is a mechanical resonance.
If it is an acoustical resonance then the internal damping should cure it. If mechanical, then see the recent thread on best cabinet materials.
Regards,
David S.
Thanks for all the replies - I suspect that, as Stuart suggested, I have increased the effective length of the enclosure with my shelves without damping or suppressing this mode at all. I will try putting some generous handfuls of long-haired wool at the top and bottom of the cabinets at the weekend and see what happens to the impedance peaks. I agree that this amount of damping is unlikely to have much effect on the reflex system tuning, which is between 35 and 40 Hz.
David - My feeling is that the significant drop in frequency on installing the shelves implies that this is a cavity resonance rather than a wall resonance: the shelves have little bracing effect on the cabinet walls, and in the case of the upper shelf simply block the holes in an existing shelf. I will try the finger test, all the same - perhaps the 425 Hz peak is a wall resonance.
The cabinets are 18mm MDF, lined with bitumen sheet and lead to give a constrained layer, with carpet underlay fixed on top of that. I'm not sure how prone you would expect them to be to wall resonances, but the cavity damping is almost negligible.
Alex
David - My feeling is that the significant drop in frequency on installing the shelves implies that this is a cavity resonance rather than a wall resonance: the shelves have little bracing effect on the cabinet walls, and in the case of the upper shelf simply block the holes in an existing shelf. I will try the finger test, all the same - perhaps the 425 Hz peak is a wall resonance.
The cabinets are 18mm MDF, lined with bitumen sheet and lead to give a constrained layer, with carpet underlay fixed on top of that. I'm not sure how prone you would expect them to be to wall resonances, but the cavity damping is almost negligible.
Alex
Well, that sounds like an extremly well damped cabinet, from the mechanical point of view, so yes, probably an acoustical resonance.
I'm not sure I've seen bracing stretch the effective path length and push a standing wave resonance down, but there is a first time for everything.
Note that acoustical damping for length resonances becomes more effective when off the surface. (Moves it away from a velocity minimum towards a velocity maximum.) You might try damping a few inches off the walls or even a layer bisecting the cabinet.
David S.
This photo shows the layout of the front of the box:
An externally hosted image should be here but it was not working when we last tested it.
Note that the top enclosure is separate and isolated from the bottom one with carpet underlay (not shown in this photo).
Internal dimensions are approximately 72 cm (H) x 38 cm (D) x 22 cm (W).
I do have a medical stethoscope - could be useful.
Alex
Yes, that is a tallish box with the driver pretty close to the end. Both factors make the length resonance (height dimension) hard to tame. Last time I had one with that problem I found that a partition halfway up was the only way to really cure the resonance.
Whatever you try you will see the effectiveness in the impedance curve. If you have a mic then you can drop it in the port and see what is really going on. (Very educational way to follow your cabinet stuffing experiments.)
David S.
Whatever you try you will see the effectiveness in the impedance curve. If you have a mic then you can drop it in the port and see what is really going on. (Very educational way to follow your cabinet stuffing experiments.)
David S.
As it happens, my original design aim was for a cabinet with sides in the ratio of the golden mean to avoid precisely this, but as you can see from the front things didn't work out quite that way! Along the way I decided to put more priority on keeping a narrow baffle, so I sacrificed what might have been a more ideal ratio of dimensions. All the same, the cabinets ended up quite deep: more than half the height, anyway. In the flesh they certainly don't look like pipes.
A partition halfway up seems like a very promising approach.
Alex
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.