Note:
This post's content is focused on the Scanspeak 10f. Specifically, the 10F/8414G-10. If you try this with a different version of the 10f or another driver in general, YMWV (your mileage will vary).
While walking around in Home Depot today I came across this electrical box used for in-ceiling wiring. I thought it might be a quick/dirty way to do an install with my 10f/8414g-10's I recently purchased. Went back out to the car to grab the speakers (luckily, I had them on hand) and came back in to test fit. Fits quite well. These are NOT sealed, and the internal volume is stated to be 18 in^3. To put that in perspective, that's about 0.03 Liters. So, I knew two things: 1) the internal volume is not ideal and 2) the fact they aren't sealed may pose a problem. But, for $2, I figure it's worth a shot.
First, here are some various pictures ...
Testing/Results:
Once I got these home I did three tests:
The results break down like this...
Green - Qts, since it's in free-air = 0.749, Fs = 133 hz
Purple - Qtc = 0.744, Fc = 125 hz
Yellow - Qtc = 0.711, Fc = 123 hz
Analysis:
First off, let's be real: this is not an 'ideal' solution. But, from the results, I found it was actually a better option for those who wish to give it a go than I thought it would be.
Since I've noted this isn't a truly sealed enclosure, you may wonder what the result would be if I had sealed the enclosure entirely. I could easily have sealed up the enclosure with some silicone but I saw no reason to do so. For two reasons:
What the results show is simply these little enclosures can be used at a relatively minor detriment to the response. The overall impedance curve retains pretty much the same shape throughout. Qtc/Qts and Fs change but not by a whole lot. The only troubling aspect is the impedance bump at ~700hz which is an enclosure resonance. This particular type of resonance manifests typically as a dip in response at this frequency; if it's a straight reflection then you can expect a 6dB drop with the same Q of the blip (ie; if the blip itself has a Q of 3 then the dip would have the same Q. The higher the 'Q', the more narrow the dip. More reading here). You can see this bump is lessened when stuffing is used inside the enclosure so the bump is spread out over a few frequencies as opposed to mainly around a small band (this may or may not be a good thing depending on how you view it).
Overall this is a decent solution for those who don't have the means (tools or time) to cut their own rings or build their own pods. Size wise I can't say these really save you space. The depth is about 68mm (give or take), which is about 2.70 inches. The OD is a couple millimeters larger than the OD of the raw driver itself.
And finally, THIS TEST WAS DONE WITH THIS SPECIFIC DRIVER. THESE SAME RESULTS CAN NOT BE EXPECTED WITH DIFFERENT SPEAKERS. But, for a couple bucks, it may be worth a shot. There, of course, are other sizes so head to HD with your speaker and see what you can come up with.
The real issues you need to keep in mind when trying something like this (and in using any enclosure, really) are covered pretty well above but just to make sure they are very clear here they are again: 1) enclosure resonance (Qtc/Fs) with respect to the high-pass crossover point and 2) internal enclosure-based resonances (such as the 700hz blip in impedance) that will cause a null in the response at that point.
This post's content is focused on the Scanspeak 10f. Specifically, the 10F/8414G-10. If you try this with a different version of the 10f or another driver in general, YMWV (your mileage will vary).
While walking around in Home Depot today I came across this electrical box used for in-ceiling wiring. I thought it might be a quick/dirty way to do an install with my 10f/8414g-10's I recently purchased. Went back out to the car to grab the speakers (luckily, I had them on hand) and came back in to test fit. Fits quite well. These are NOT sealed, and the internal volume is stated to be 18 in^3. To put that in perspective, that's about 0.03 Liters. So, I knew two things: 1) the internal volume is not ideal and 2) the fact they aren't sealed may pose a problem. But, for $2, I figure it's worth a shot.
First, here are some various pictures ...
Testing/Results:
Once I got these home I did three tests:
- Green - Driver Free-Air
- Purple - Driver in 'enclosure'* with no stuffing
- Yellow - Driver in 'enclosure'* with stuffing
* The term 'enclosure' is only used because I can't think of a better word and I don't want to type "pseudo-enclosure" multiple times. Again, this thing is not sealed, so it's more like a leaky enclosure. More on this in the Analysis section.
The results break down like this...
Green - Qts, since it's in free-air = 0.749, Fs = 133 hz
Purple - Qtc = 0.744, Fc = 125 hz
Yellow - Qtc = 0.711, Fc = 123 hz
Analysis:
First off, let's be real: this is not an 'ideal' solution. But, from the results, I found it was actually a better option for those who wish to give it a go than I thought it would be.
Since I've noted this isn't a truly sealed enclosure, you may wonder what the result would be if I had sealed the enclosure entirely. I could easily have sealed up the enclosure with some silicone but I saw no reason to do so. For two reasons:
- Sealing this entirely would have only increased the resonance (Qtc). Given I was already above 0.711 Qtc I saw no need to seal up the little dude entirely. The holes are so small there is no real concern with any rear-wave reflection coming back in to the enclosure. It's pretty much the same scenario you get with an Aperiodic Membrane type enclosure; the holes are so small relative to the amount of stuffing you may use that you essentially are 'tuning' the internal volume to produce the desired effect. With AP you typically expect to lose output but I see no evidence of that here (the impedance sweeps are the same resistance level throughout as the free-air results).
- If you're using these to produce bass frequencies, which is when sealing an enclosure really matters, you're using these wrong. Plain and simple. These drivers should not be crossed below 300hz with a 2nd order slope to limit excursion (yes, you can cross them at 100hz but I can also punch my boss in the face... doesn't mean it's a good idea). So the concern about internal pressure used as the back spring creating 'bass' is negated by the appropriately set high-pass filter.
What the results show is simply these little enclosures can be used at a relatively minor detriment to the response. The overall impedance curve retains pretty much the same shape throughout. Qtc/Qts and Fs change but not by a whole lot. The only troubling aspect is the impedance bump at ~700hz which is an enclosure resonance. This particular type of resonance manifests typically as a dip in response at this frequency; if it's a straight reflection then you can expect a 6dB drop with the same Q of the blip (ie; if the blip itself has a Q of 3 then the dip would have the same Q. The higher the 'Q', the more narrow the dip. More reading here). You can see this bump is lessened when stuffing is used inside the enclosure so the bump is spread out over a few frequencies as opposed to mainly around a small band (this may or may not be a good thing depending on how you view it).
Overall this is a decent solution for those who don't have the means (tools or time) to cut their own rings or build their own pods. Size wise I can't say these really save you space. The depth is about 68mm (give or take), which is about 2.70 inches. The OD is a couple millimeters larger than the OD of the raw driver itself.
And finally, THIS TEST WAS DONE WITH THIS SPECIFIC DRIVER. THESE SAME RESULTS CAN NOT BE EXPECTED WITH DIFFERENT SPEAKERS. But, for a couple bucks, it may be worth a shot. There, of course, are other sizes so head to HD with your speaker and see what you can come up with.
The real issues you need to keep in mind when trying something like this (and in using any enclosure, really) are covered pretty well above but just to make sure they are very clear here they are again: 1) enclosure resonance (Qtc/Fs) with respect to the high-pass crossover point and 2) internal enclosure-based resonances (such as the 700hz blip in impedance) that will cause a null in the response at that point.
