Hi all,
I keep making this more complicated than it probably needs to be. So, tossing in the towel for now and just going to ask for guidance.
I'm trying to keep it simple and calculate changes in Q from measurements.
An example would be:
I design a sealed enclosure. The Q is around 0.750 for example. Impedance measurement is around 35hz and is around 25ohm, again, just as examples. I then stuff the box with polyfil, let's assume around 1lb per ft^3 as conventional wisdom typically puts it. Then I measure and impedance sweep after stuffing, and notice the impedance shifts its central peak frequency and of course diminishes the impedance value itself. Can I calculate a new Q from this? And could I use this to help guide how much stuffing for that particular system to use? Ie, do I need to add more to re-measure and see Q shift in a direction with intension?
(mainly trying to understand how the amount of stuffing and a measured change in impedance sweep information can be used to calculate changes in Q and be able to compare to simulation, both just for the sake of knowing how and why, and for the practical results. I'm not trying to achieve a specific Q with stuffing, just curious what it is. My stuffing is solely to smooth out enclosure standing waves and box issues.)
Very best,
I keep making this more complicated than it probably needs to be. So, tossing in the towel for now and just going to ask for guidance.
I'm trying to keep it simple and calculate changes in Q from measurements.
An example would be:
I design a sealed enclosure. The Q is around 0.750 for example. Impedance measurement is around 35hz and is around 25ohm, again, just as examples. I then stuff the box with polyfil, let's assume around 1lb per ft^3 as conventional wisdom typically puts it. Then I measure and impedance sweep after stuffing, and notice the impedance shifts its central peak frequency and of course diminishes the impedance value itself. Can I calculate a new Q from this? And could I use this to help guide how much stuffing for that particular system to use? Ie, do I need to add more to re-measure and see Q shift in a direction with intension?
(mainly trying to understand how the amount of stuffing and a measured change in impedance sweep information can be used to calculate changes in Q and be able to compare to simulation, both just for the sake of knowing how and why, and for the practical results. I'm not trying to achieve a specific Q with stuffing, just curious what it is. My stuffing is solely to smooth out enclosure standing waves and box issues.)
Very best,
Last edited:
Thanks,
Qt there, is that Qts?
So indirectly, the impedance measurement would give me the Fc and then use the driver's native Fs and Qts and from these derive the new Qtc?
Very best,
Yes, when that's all you have, otherwise its Qts' = Qts + any added series resistance, though personally just used Qes unless a high output impedance amp was used:
(Qts'): (Qts) + any added series resistance (Rs)
(Qts'): (Qts) + any added series resistance (Rs)
Q can be measured directly. It's a more involved process and can give greater precision for when that's needed. Just seeing whether the resonance is shifting and by how much (F and Q are linked), this method can be revealing.
Thanks,
Hrm, Ok, let's see if any of this makes sense then.
Here's the unstuffed impedance measurement:
Here's the same thing, but with stuffing (about 1lb per ft^3) in the sealed enclosure.
The impedance spike shifted left from about 48hz to closer to 42hz and the impedance itself went from close to 43ohm down to 23ohm or so.
Simulated values:
The enclosure's net internal volume is around 7.870 ft^3.
Qtc before stuffing was around 0.834.
Fsc: 44.69hz
Driver T&S for discussion:
Please note, I'm using four of these drivers in the above configuration (wired series-parallel, final 4ohm load).
Re 3.6ohm
Le 2.57mH
Sd 551m^2
BL 14.08
Vas 76.7L
CMS 177.0um/N
MMS 179.0g
Fs 28hz
Qms 8.85
Qes 0.57
Qts 0.54
Xmax 12.1mm (one way)
Maybe this is complicating it? I'm just trying to understand a little more about how this worked out. Perhaps the 4 drivers wired together in series-parallel is throwing off my numbers expected?
When I do the calc based on the modeled numbers, they don't land on what I simulated.
And I'm not sure how to incorporate the post-stuffing impedance information to see how it changed in the calculations.
Very best,
Hrm, Ok, let's see if any of this makes sense then.
Here's the unstuffed impedance measurement:
Here's the same thing, but with stuffing (about 1lb per ft^3) in the sealed enclosure.
The impedance spike shifted left from about 48hz to closer to 42hz and the impedance itself went from close to 43ohm down to 23ohm or so.
Simulated values:
The enclosure's net internal volume is around 7.870 ft^3.
Qtc before stuffing was around 0.834.
Fsc: 44.69hz
Driver T&S for discussion:
Please note, I'm using four of these drivers in the above configuration (wired series-parallel, final 4ohm load).
Re 3.6ohm
Le 2.57mH
Sd 551m^2
BL 14.08
Vas 76.7L
CMS 177.0um/N
MMS 179.0g
Fs 28hz
Qms 8.85
Qes 0.57
Qts 0.54
Xmax 12.1mm (one way)
Maybe this is complicating it? I'm just trying to understand a little more about how this worked out. Perhaps the 4 drivers wired together in series-parallel is throwing off my numbers expected?
When I do the calc based on the modeled numbers, they don't land on what I simulated.
And I'm not sure how to incorporate the post-stuffing impedance information to see how it changed in the calculations.
Very best,
Yea, the Fsc 44.69hz was the simulation value.
The 41hz and 48hz were my measurements.
Just trying to see what else I can figure and calculate to know more about how things shifted, and how it compares to a model, etc. Mainly trying to see what changed with the stuffing that I can use to then compare to modeling or explain a change.
Very best,
Published is of course different from measured.
Here's my DATSV3 on one of the drivers. Again, the above measurements were 4 of these together in series-parallel.
Here's just one (free air, no enclosure):
Suggests:
Fs 35.87hz
Qts 0.6843
Qes 0.748
Qms 8.039
Vas 2.252 ft^3
So then, Fc=Qtc/Qt*Fs: 44.69hz = 0.834 / (0.6843 * 35.87). Or if we assume Qtc is not correct, we re-calc for it:
44.69hz = X / (24.5)
X = 1094
I'm confused. LOL
Very best,
Last edited:
I'm building a new box currently for some new drivers, so I will try to do a better job of doing measurements before & after and more care getting some information so I can compare and try to understand the changes after I stuff this one. I'm not trying to get to a specific Q, I don't really care too much about that. Just stuffing to eliminate resonance and standing waves, etc. But out of interest, I like to learn with each iteration, and seeing things move on that impedance measurement of course are interesting. So, trying to get an idea of the calculations and where my measurements fit in so I can understand it and maybe try to simulate it too, to show understanding, if possible.
Very best,
Very best,
Thanks, that's interesting.
Going back to a comment earlier, directly measuring a system's Q. I'm curious if I have what's needed to do that, and not just derive. Or if its just much easier to measure the driver free air T&S and impedance sweep of the driver in the enclosure and from that derive Q?
Very best,
Just seemed like a good tool to have for all this.
I think I'm complicating it because of the 4x drivers in a single box. Compared to having a measurement of a single driver in a box. I imagine the wiring of 4 of these drivers in series-parallel is changing some things.
On my next go around I'll try to do single driver metrics and group metrics to see the differences there. Maybe derive results that better match.
Currently with my above measurements, I can't get anything to really agree.
Very best,
@MalVeauX Have you tried a nearfield measurement of the sound pressure response of one of your woofer's in your enclosure? This will of course include all of the effects of the stuffing, as well as the effects of the mutual radiation impedances of the drivers. I suggest you lay the enclosure on its back, with all four drivers facing upwards. That way they will all be loaded by their nearby environment as equally as possible. They should all produce quite similar curves if each driver's free air resonance frequency is the same.
If each driver's nearfield SPL response has a peak relative to the response at higher frequencies (say > 5fc) , then you know that the Qtc of the overall system is a little high.
Small (1972) has calculated some results that will help you to get an idea of where your stuffing-filled-system sits in terms of Qtc:
Qtc = 0.50: SPL @ fc = -6 dB
Qtc = 0.71: SPL @ fc = -3 dB
Qtc = 1.00: SPL @ fc = 0 dB
Qtc = 1.40: SPL @ fc = +3 dB
Qtc = 2.00: SPL @ fc = +6 dB
Reference
R.H. Small (1972), "Closed-Box Loudspeaker Systems Part I: Analysis," J. Audio Eng. Soc., Vol. 20, pp. 798–808 (December).
If each driver's nearfield SPL response has a peak relative to the response at higher frequencies (say > 5fc) , then you know that the Qtc of the overall system is a little high.
Small (1972) has calculated some results that will help you to get an idea of where your stuffing-filled-system sits in terms of Qtc:
Qtc = 0.50: SPL @ fc = -6 dB
Qtc = 0.71: SPL @ fc = -3 dB
Qtc = 1.00: SPL @ fc = 0 dB
Qtc = 1.40: SPL @ fc = +3 dB
Qtc = 2.00: SPL @ fc = +6 dB
Reference
R.H. Small (1972), "Closed-Box Loudspeaker Systems Part I: Analysis," J. Audio Eng. Soc., Vol. 20, pp. 798–808 (December).
Last edited:
Hi,
I did some pre-stuffing ground plane measurements of 4 of the drivers in a sealed 7.87 ft^3 enclosure at 1 meter & 2 meter:
Very best,