I have been considering a few options for a PC desktop speaker. I like small FR drivers for such an application, but was looking to mitigate a couple of possible issues.
As to issue 2, what if we simply rolled-off one driver around the frequency where the rise in response happens? By 8-10kHz, a single driver's beam should already be narrowing, even for a 3", so that it does not make the array moot for the desktop bounce issue described above.
Finally for issue 3, the need for BSC depends on taste and bass alignment. For several reasons, I would not go as far as adding 2 additional drivers to make a 4-driver 1.5-way. What if we only added one extra driver? It would provide partial BSC, hopefully in a manner complimentary to the bass alignment and speaker location.
The driver playing fullrange (A) and the one rolled-off in the 8kHz or so region (B) would naturally be arrayed vertically, most likely as close as possible to each other, or spaced so as to adjust the vertical pattern. The BSC driver could be placed anywhere on the enclosure really. I would possibly make a triangle, with the BSC driver on side, mid-height, of the two "main" drivers, just because it would look cool. 🙂
I see this being useful with drivers such as the Tymphany TC9/TG9, which do seem to have a rise up high and are affordable enough to buy 6 of easily. I have never worked with them or measured them myself though.
The one awkward thing might be wiring and impedance. It would also depend on the used driver's own impedance, 4-8-16 ohm. One could wire in series with bypass capacitors on two drivers, or in parallel with series inductors on two drivers. I'm wishfully thinking first-order electrical roll-off should be enough for both "steps", but this would be a case-by-case ultimately IMO, particularly for the HF step. The impedance response of a complete speaker will likely range from 4 to 12+ ohm nominally, Le notwithstanding. SS amplifiers should not mind if rated for the lowest impedance range.
Does this make sense and could it be worthwhile? Is there anything obvious I overlooked? Is anybody readying a straight-jacket?
- Desktop bounce; I don't know how much of an issue this actually is, but if using smaller drivers, 4" and below, their greater midrange dispersion may comb-filter with the reflection to an extent where addressing it could be worthwhile.
- A "hot" top-end; often found on such drivers, where the upper kHz range sees a rise in response up to the driver's HF roll-off. This can help balance the in-room power-response if listened to with some distance and off-axis, but for nearfield on-axis use, it can be a bit much in some cases.
- Baffle-step; nothing new, but a typical 1.5-way likely offers too much compensation at 6dB, especially for desktop use with a wall behind the speakers.
As to issue 2, what if we simply rolled-off one driver around the frequency where the rise in response happens? By 8-10kHz, a single driver's beam should already be narrowing, even for a 3", so that it does not make the array moot for the desktop bounce issue described above.
Finally for issue 3, the need for BSC depends on taste and bass alignment. For several reasons, I would not go as far as adding 2 additional drivers to make a 4-driver 1.5-way. What if we only added one extra driver? It would provide partial BSC, hopefully in a manner complimentary to the bass alignment and speaker location.
The driver playing fullrange (A) and the one rolled-off in the 8kHz or so region (B) would naturally be arrayed vertically, most likely as close as possible to each other, or spaced so as to adjust the vertical pattern. The BSC driver could be placed anywhere on the enclosure really. I would possibly make a triangle, with the BSC driver on side, mid-height, of the two "main" drivers, just because it would look cool. 🙂
I see this being useful with drivers such as the Tymphany TC9/TG9, which do seem to have a rise up high and are affordable enough to buy 6 of easily. I have never worked with them or measured them myself though.
The one awkward thing might be wiring and impedance. It would also depend on the used driver's own impedance, 4-8-16 ohm. One could wire in series with bypass capacitors on two drivers, or in parallel with series inductors on two drivers. I'm wishfully thinking first-order electrical roll-off should be enough for both "steps", but this would be a case-by-case ultimately IMO, particularly for the HF step. The impedance response of a complete speaker will likely range from 4 to 12+ ohm nominally, Le notwithstanding. SS amplifiers should not mind if rated for the lowest impedance range.
Does this make sense and could it be worthwhile? Is there anything obvious I overlooked? Is anybody readying a straight-jacket?
On a few speakers I've made I've use to identical full range drivers wired in series with a small ( 1.5 uf ) cap across the lower one, you could play around with the value of the cap to get your preferred balance of dispersion/beaming.
With 3 drivers a 4Ω driver is suitable for series wiring with a big cap shunting the 2 that only do LF. XO should be set such that the C-C is less than a quarter-wavelength. You will not get an BSC, but it is unlikely needed.
I am doing one with 3x A6.2p.
dave
I am doing one with 3x A6.2p.
dave
Yeah, I can see I glossed over too quickly with the series wiring scheme I proposed, it would not work as I intended. If going with a TC9 or TG9, the high Q would indeed lessen the need for BSC.
If the driver is available in 16ohm, parallel could be viable.
I was also considering mixing impedance with a 3FE22, offered in 4-8-16 ohm, but their T/S parameters differ quite a bit; I don't want to analyze this on top of the rest, for now.
If the driver is available in 16ohm, parallel could be viable.
I was also considering mixing impedance with a 3FE22, offered in 4-8-16 ohm, but their T/S parameters differ quite a bit; I don't want to analyze this on top of the rest, for now.
two drivers, close listening, there will be horrible combing.
When we are talking reducing vertical pattern in a line array, there are more than just 2 drivers, and we are sitting quite farther back.
If you really want 2 drivers, I'd go with one front firing and one up firing.
When we are talking reducing vertical pattern in a line array, there are more than just 2 drivers, and we are sitting quite farther back.
If you really want 2 drivers, I'd go with one front firing and one up firing.
I never messed with arrays of any kind, so I could not tell from experience how bad combing would be. I imagine that using smaller drivers such as 3.5" units would not be so bad, if you can aim the beam at ear height. Perhaps it would indeed be worse than the desktop bounce?
A top-firing driver could deal with the hot top-end issue OTOH, good call. Would that require more space around the speaker, such as no back-wall?
A top-firing driver could deal with the hot top-end issue OTOH, good call. Would that require more space around the speaker, such as no back-wall?
Many of the usual suspects - on the budget end of the spectrum - seem to posses such a response. The TG9FD10-04 for example, appears to behave quite well until that hump around 10kHz. I've listened to worse mind you, look at my avatar (FF125K in case I ever change it...) and while TG9 is mild in that regard, it could still bear addressing IMO.
Even having to buy 4 to 8 units for messing about with mini-arrays (or multi-pole configurations) is still relatively cheap fun!
Even having to buy 4 to 8 units for messing about with mini-arrays (or multi-pole configurations) is still relatively cheap fun!
Danny Richie GR research is now offering a single driver 3" wide-band design, Little Giant Killer 2.0
On my traffic light speakers I used a total of four identical drivers, three in the front and one in the rear; they were wired in series/parallel all full range at first, but they sounded terrible, down to beaming more than combing I think; I later re-wired them so the middle and rear were in series with a cap across the rear and the top and bottom in series, with a choke to bring them in around ( below ) 1,000 Hz. They seemed to work quite well like this, however they developed an intermittent buzz, ether from a loose connection or driver bolt. All the drivers share a common enclosure, vented to the hollow stand.
I wanted cheap test speakers for amps and ended up buying 6x used Visaton FRS8M.
Decided to put them in a box for fun and cross over passively about 1st order ("d'appolito" arrangement).
Had to roll the middle "tweeter" off too as the FRS8M are "screamers" otherwise.
Other than very limited LF-extension quite a nice speaker that probably can be made to work even better than my humble attempt.
Can imagine how good this could be with Mark Audio´s or similar.
On the question of desk top bounce, there are a couple of things you could try. I find having a small gap (<1") between the speaker and the desk top frees up the sound a bit. This can simply be done by tilting the speaker back. Having a horizontal strip of carpet or felt on the baffle under the driver cuts down HF off the desk surface. These ideas cost nothing to try, even time wise.