Hello everyone!
I'm planning on building a pair of rear surround speakers for my system. They will be used mainly for music, but movies too. I have quite a few 5.1 albums which I'd love to listen.
I have a pair of Dayton Audio RS100-4 (4", long-throw, aluminium cone, 4 ohm fullrange speakers) laying around that I would like to use for this job.
Usually, I'd use them alone by themselves without any tweeters, but since I'd be listening mostly off axis I'd love them to have a good off axis response. Therefore I think I need to add a tweeter.
I've been thinking on how to do this and every approach I can come off with has its own set of drawbacks and I can't decide. I'll list them here.
1. Just to add the tweeter, let the RS100 run full range and filter the tweeter at about 3.5kHz (30 degree off axis FR plot here, green line it's for the 8ohm variant but the 4 ohm version should be pretty similar).
I think this is the simplest approach but if I aim for a flat off-axis response, the on-axis will look like trash because it will be both tilted upwards and I will also have to deal with comb filtering because the tweeter and the woofer will play the same frequencies, and due to the wave travel paths not being equal based on how I move it might look a little be better or worse. Impedance also might become an issue. I have no idea how bad this might be.
2. Treat the RS100 as a regular midwoofer and use a regular 2-way LP/HP 1st or 2nd order crossover and cross lower (I'm thinking in the 2.5kHz/3kHz ballpark). In theory this should be the approach the will provide the best on-axis and off-axis, closely matching the directivity between the two speakers. However, the damping factor will be pretty low. The RS100 already have a high Qts, which is already somewhat compromised by the 10m (32ft) of wire that I will have to run from my receiver to the speakers. I would prefer to avoid adding another choke point in the form of an inductor.
2b. What if I did this but instead of using an electrical crossover on the RS100 I used a "mechanical" one (eg. a piece of cloth in front of the woofer)? Would this work? Has anyone had any experience with this?
3. Get a very small tweeter (I'm thinking SB Acoustics SB14ST-C000-4), remove the phase plug of the woofer, and make it into a coaxial. Then I'd just filter the tweeter and let the RS100 run full-range. This way I'd have to deal only with the rising response on-axis and not with comb filtering, because the wave paths of the woofer and the tweeter would be equal.
4. Sell the RS100 and get something else :'
I can use DSP on the entire channel, I'm using a PC as a source that is running Equalizer APO so I have access to many filters. The speakers will be hanged on to a wall, at the two opposite corners of the room, near the ceiling. I don't think I have to deal with baffle step because of such placement.
I'm planning on building a pair of rear surround speakers for my system. They will be used mainly for music, but movies too. I have quite a few 5.1 albums which I'd love to listen.
I have a pair of Dayton Audio RS100-4 (4", long-throw, aluminium cone, 4 ohm fullrange speakers) laying around that I would like to use for this job.
Usually, I'd use them alone by themselves without any tweeters, but since I'd be listening mostly off axis I'd love them to have a good off axis response. Therefore I think I need to add a tweeter.
I've been thinking on how to do this and every approach I can come off with has its own set of drawbacks and I can't decide. I'll list them here.
1. Just to add the tweeter, let the RS100 run full range and filter the tweeter at about 3.5kHz (30 degree off axis FR plot here, green line it's for the 8ohm variant but the 4 ohm version should be pretty similar).
I think this is the simplest approach but if I aim for a flat off-axis response, the on-axis will look like trash because it will be both tilted upwards and I will also have to deal with comb filtering because the tweeter and the woofer will play the same frequencies, and due to the wave travel paths not being equal based on how I move it might look a little be better or worse. Impedance also might become an issue. I have no idea how bad this might be.
2. Treat the RS100 as a regular midwoofer and use a regular 2-way LP/HP 1st or 2nd order crossover and cross lower (I'm thinking in the 2.5kHz/3kHz ballpark). In theory this should be the approach the will provide the best on-axis and off-axis, closely matching the directivity between the two speakers. However, the damping factor will be pretty low. The RS100 already have a high Qts, which is already somewhat compromised by the 10m (32ft) of wire that I will have to run from my receiver to the speakers. I would prefer to avoid adding another choke point in the form of an inductor.
2b. What if I did this but instead of using an electrical crossover on the RS100 I used a "mechanical" one (eg. a piece of cloth in front of the woofer)? Would this work? Has anyone had any experience with this?
3. Get a very small tweeter (I'm thinking SB Acoustics SB14ST-C000-4), remove the phase plug of the woofer, and make it into a coaxial. Then I'd just filter the tweeter and let the RS100 run full-range. This way I'd have to deal only with the rising response on-axis and not with comb filtering, because the wave paths of the woofer and the tweeter would be equal.
4. Sell the RS100 and get something else :'
I can use DSP on the entire channel, I'm using a PC as a source that is running Equalizer APO so I have access to many filters. The speakers will be hanged on to a wall, at the two opposite corners of the room, near the ceiling. I don't think I have to deal with baffle step because of such placement.
While you could probably get reasonable results a number of ways, the better way would be to treat the fullrange as if it were a regular woofer. The dispersion of a fullrange typically narrows at the same frequency as a woofer of the same size, right around the frequency where breakup becomes significant.
1. Hmm, FVc = ~34400/pi/2.5 = ~4380 Hz end of pistonic, then assuming the dustcap/dC is ~2.5 cm = 34400/2/2.5 = ~6880 Hz - up are its breakup modes, leaving the ~4380-6880 Hz BW its TL modes radiating off the Vc to launch off the diaphragm at whatever the diaphragm's SoS is.
Using the pioneer's way of choosing XO points that ~matches up polar responses we get the 1st point @ sqrt(84.8*4380) = ~609 Hz, 2nd = sqrt(609*4380) = 1634 Hz, 3rd = 2675 Hz, etc., so the latter a good plan for a typical tweeter, but use one of these new mini FR drivers and the 609 Hz is my preferred way to go as it puts the ctc <1/4 WL and may make 1st order an option.
2b. Yes, the pioneer's were all about getting as much efficiency as practical, so 'acoustic solutions for acoustic problems' was their mantra and still alive and well among the 'FR' driver aficionados, though not to the extent it once was when we were mostly stuck with traditional cheap systems/ceiling drivers that required mass quantities of tweaking.
That said, you're really only needing to damp the dC, TL modes, so some pioneer's suspended a dense material to just cover only as much of the diaphragm required or removed dC and bonded a felt ring of the required diameter to it, sometimes doping to 'taste'.
3. I guess if the components are available to meet the needs of the app................
4. Dunno, I only know what its published and/or measured specs 'tell' me, but based on your app (far away), two drivers or larger coax seems a better choice overall unless desired average SPL is relatively low.
Using the pioneer's way of choosing XO points that ~matches up polar responses we get the 1st point @ sqrt(84.8*4380) = ~609 Hz, 2nd = sqrt(609*4380) = 1634 Hz, 3rd = 2675 Hz, etc., so the latter a good plan for a typical tweeter, but use one of these new mini FR drivers and the 609 Hz is my preferred way to go as it puts the ctc <1/4 WL and may make 1st order an option.
2b. Yes, the pioneer's were all about getting as much efficiency as practical, so 'acoustic solutions for acoustic problems' was their mantra and still alive and well among the 'FR' driver aficionados, though not to the extent it once was when we were mostly stuck with traditional cheap systems/ceiling drivers that required mass quantities of tweaking.
That said, you're really only needing to damp the dC, TL modes, so some pioneer's suspended a dense material to just cover only as much of the diaphragm required or removed dC and bonded a felt ring of the required diameter to it, sometimes doping to 'taste'.
3. I guess if the components are available to meet the needs of the app................
4. Dunno, I only know what its published and/or measured specs 'tell' me, but based on your app (far away), two drivers or larger coax seems a better choice overall unless desired average SPL is relatively low.
actually, now that I think about it, for rear, you may not want to add a tweet/supertweet.
"Back in my day" before ac-3, or dolby digital, we had prologic.
If I remember, the rears rolled off above 7khz I think, but rears were used more for enveloping fill, think city background noise or underwater, sound that is hard to localize.
But no, the industry wanted 20-20khz stereo sound, or 5 channels plus subwoofer.
So, if you want a smoother, fall into the scene surround versus the sharpest crinkliest surround that easy to localize, forget the add on tweet for the surrounds......
"Back in my day" before ac-3, or dolby digital, we had prologic.
If I remember, the rears rolled off above 7khz I think, but rears were used more for enveloping fill, think city background noise or underwater, sound that is hard to localize.
But no, the industry wanted 20-20khz stereo sound, or 5 channels plus subwoofer.
So, if you want a smoother, fall into the scene surround versus the sharpest crinkliest surround that easy to localize, forget the add on tweet for the surrounds......
Thanks for the many answers.
I want to keep my rears as full range as possible, they will be used mainly for music because I have a few 5.1 albums and I count on adding more to my collection. The RS100 provide decent low-end and power handling for a 4 inch speaker in a very small enclosure. I don't need very high SPL, I guess around 90dB is fine, the room is about 7m*6m, or 23ft*20ft.
I think I'll start by the simplest approach, that is #1 as suggested by GM, and then move to the more complex ones.
If it works I may use the same approach for the center channel, maybe with its bigger brother RS135-8.
But I have some questions:
Why shouldn't I cross right at 4380Hz for example, at the end of the pistonic range, or at 3kHz which is the -3dB point at 30 degrees according to the published (and independently measured) response?
I want to keep my rears as full range as possible, they will be used mainly for music because I have a few 5.1 albums and I count on adding more to my collection. The RS100 provide decent low-end and power handling for a 4 inch speaker in a very small enclosure. I don't need very high SPL, I guess around 90dB is fine, the room is about 7m*6m, or 23ft*20ft.
I think I'll start by the simplest approach, that is #1 as suggested by GM, and then move to the more complex ones.
If it works I may use the same approach for the center channel, maybe with its bigger brother RS135-8.
But I have some questions:
This is very nice to hear, but I am struggling to understand where the formulas for chosing the XO point come from - and why for example 2675Hz should be a "better" frequency for XO then, say, 2400Hz, or if we don't even know what the polar response of the woofer looks like, or how the cone material behaves rigidity-wise.
Why shouldn't I cross right at 4380Hz for example, at the end of the pistonic range, or at 3kHz which is the -3dB point at 30 degrees according to the published (and independently measured) response?
Do you think attaching some blue tack to rear of the cone might help? Or can I just use an old sock "wrapped" on the speaker grill?
intelligibility...........
Not getting into time/phase alignment (which is big for intelligibility to me).........
To me, the upper harmonics of s's and t's help intelligibility, those fall about 8khz I think.
Otherwise voice would sound fine on a woofer with a flat response that rolls off at 5khz.
Am I wrong on this ?
When I cut the whizzer off my pioneer b20 maybe 20 years ago, the sound was pretty muffled............
Not getting into time/phase alignment (which is big for intelligibility to me).........
To me, the upper harmonics of s's and t's help intelligibility, those fall about 8khz I think.
Otherwise voice would sound fine on a woofer with a flat response that rolls off at 5khz.
Am I wrong on this ?
When I cut the whizzer off my pioneer b20 maybe 20 years ago, the sound was pretty muffled............
Last edited:
It is worth realizing >
Back in the day of 'Prologic', home movies were stereo video tapes needing stereo VCR's required for surround sound.
The rear channel (mono) was simply derived using a L/R differential amp. The maximum separation - front to rear was only 6dB.
The deriving of a front-center channel was very well done >Hats off to Dolby for that.
However, the simple method of deriving rear signal showed-up massive amounts of high frequency distortion. (very sibilant stuff)
So the high frequency filtering (7KHz) was not to improve localization, but to reduce high frequency distortion.
With todays Digital Surround, it is the recording engineers that determine what highs do or don't come from the rear.
So, for proper surround today you definitely need rear speakers with proper high frequency response, especially with 5.1 music
Ditto
No, ~ 4-10 kHz is where sibilance and sharp sounds in general resides: https://alexiy.nl/eq_chart/
You're welcome!
'Sound is round', expanding exponentially plus point source drivers are presumed to be pistonic, ergo Fs to its Vc diameter is normally its BW, though if if the dC is larger, then need to use it as its upper BW limit. As such, octave spreads math is required to find a driver's equal acoustic power BW mean, i.e where a mechanical XO would be located if a biflex driver, then the next mean point would be for a triflex such as this one which is at the dC since it's a one piece molded diaphragm.
The higher the XO point, the less optimal blending of drivers both on and off axis, 'forcing' a higher XO slope, the less efficient the design/system is.
The way drivers I'm familiar with were/are normally 'laid up', no, though if it has a 'Flexiflyer' (flimsy, no damping coating on the rear) construction, then yes.
No clue about nowadays, but grill screen was factored in for the final voicing, so you can experiment to find the best overall trade-off with open cell foam usually preferred.
Octave math:
Fl = Fh/2^n
n = ln(Fh/Fl)/ln(2)
where:
Fh = upper frequency
Fl = lower frequency, or the XO point in this case
n = octave spread
ln(2) = 0.6931
Whatever works!
Wouldn't work for me, but I use(d) large horns designed to only cover the desired listening area and when tinkering with FR drivers and multiways wasn't much interested in extreme highs since there weren't any recordings to need it till right near the end of my active DIYaudio 'career' nor for a variety of had the keen hearing, relying on my female's much more acute hearing to guide my designs, HF tweaking, etc..
Ah, ty mister audio, did not know that.
I know a 6khz peak (resonance) did not offend.
W4-1320sf tang band 4" bamboo.
https://rutcho.com/speaker_drivers/tang_band_w4_1320_sb/tang_band_w4_1320_sb.html
https://www.rjbaudio.com/Bandit/bandit.html
"but when I took measurements I realized that there was a peak at 6kHz that needed to be tamed"
It did not make s's and t's overwhelming run without a filter (to me).
It seemed to add some clang to a cymbal hit.
I know a 6khz peak (resonance) did not offend.
W4-1320sf tang band 4" bamboo.
https://rutcho.com/speaker_drivers/tang_band_w4_1320_sb/tang_band_w4_1320_sb.html
https://www.rjbaudio.com/Bandit/bandit.html
"but when I took measurements I realized that there was a peak at 6kHz that needed to be tamed"
It did not make s's and t's overwhelming run without a filter (to me).
It seemed to add some clang to a cymbal hit.
Quote:
Usually, I'd use them alone by themselves without any tweeters, but since I'd be listening mostly off axis I'd love them to have a good off axis response.
Manger used once for his Manger fullrange transducer a star like diffusor.
It can widen up dispersion without adding a delay like some other diffusing elements.
Here you see it for a tweeter but you can build it out of craft paper more deep and bigger
Usually, I'd use them alone by themselves without any tweeters, but since I'd be listening mostly off axis I'd love them to have a good off axis response.
Manger used once for his Manger fullrange transducer a star like diffusor.
It can widen up dispersion without adding a delay like some other diffusing elements.
Here you see it for a tweeter but you can build it out of craft paper more deep and bigger
Last edited:
This one adds delay but it's worth a try
https://m.indiamart.com/proddetail/hvac-round-air-diffuser-19127827248.html
https://m.indiamart.com/proddetail/hvac-round-air-diffuser-19127827248.html
This one looks good, too.
https://m.indiamart.com/proddetail/aluminum-ceiling-round-air-diffuser-27048839533.html
https://m.indiamart.com/proddetail/aluminum-ceiling-round-air-diffuser-27048839533.html