M74A (probably with T25A/B) is SOTA solution for midrange. It might give comparable solutions, but probably not better.
SOTA for 10" is Purifi and Kartesian 265_vKi. Excellent performance, but both very expensive.
At least, Revelator 26W is half the price of the Purifi 10" - and I would prefer double 26W configuration compared to a single Purifi 10".
The RSS265HF-8 is also a really great selection, but needs power to get the high cone mass and stiff spider moved.
Also RSS has more subwoofer like design and Revelator 26W more woofer like with low moving mass and very low mechanical losses, and when to be used up to >500Hz I somehow would prefer the latter.
BTW: Don't underestimate the 26W Discovery! A pair per side in 60l CB "Air suspension" with bucking magnets and EQ can play crystal clear and controlled without noise or perceived distortion during stroking near x_mech. When you observe them in operation, you see no "pumping" aka dynamic coil offset, just +/-1cm "blurring" of the cone... I very like this driver, it's not a compromise solution.
SOTA for 10" is Purifi and Kartesian 265_vKi. Excellent performance, but both very expensive.
At least, Revelator 26W is half the price of the Purifi 10" - and I would prefer double 26W configuration compared to a single Purifi 10".
The RSS265HF-8 is also a really great selection, but needs power to get the high cone mass and stiff spider moved.
Also RSS has more subwoofer like design and Revelator 26W more woofer like with low moving mass and very low mechanical losses, and when to be used up to >500Hz I somehow would prefer the latter.
BTW: Don't underestimate the 26W Discovery! A pair per side in 60l CB "Air suspension" with bucking magnets and EQ can play crystal clear and controlled without noise or perceived distortion during stroking near x_mech. When you observe them in operation, you see no "pumping" aka dynamic coil offset, just +/-1cm "blurring" of the cone... I very like this driver, it's not a compromise solution.
Thanks for the suggestion, the 10RS430 looks interesting. It looks like it is designed for rear mounting, which would be challenging for me to integrate into the kind of cabinets I build. I suppose I could recess mount it from the front and make a trim ring to clean up the look... ?
Again thanks !
You can not have high efficienty and low Fs. Most of the pro drivers are out for deep bass in that size. The Daytons, SS, will have 10 to 20 hz lower résonance.
The SS cheat about their Fs data, it is above like is showing the impedance graph. Must be corrected for pré simulations. The Faital RS are someting in between.
The SS cheat about their Fs data, it is above like is showing the impedance graph. Must be corrected for pré simulations. The Faital RS are someting in between.
Thanks for pointing that out, I had not noticed that the published impedance plot has a peak at about 26 Hz, versus the spec Fs of 19 Hz.
I have never experimented with bucking magnets, nor researched them.
What effect does a bucking magnet have on performance? Thanks for any information... j.
Bucking magnets CAN SOMETIMES increase the field and lower the Q while increasing Bl. This is for like polar attachment where field increases. Sometimes the magnet does not saturate the gap and additional Ferrite or Neo magnets can help. This also can work if a driver does not meet spec.
If you reverse it, your Q goes up and Bl goes down. In smaller boxes, you can get a boost before rolloff this way.
Measure T/S before gluing it down to verify which way you prefer it In your application. Use a 5 minute 2 part epoxy and mark the magnet with a sharpie to make sure you get it right. Amazon has Ferrite magnets, and likely Neo rings too.
If you reverse it, your Q goes up and Bl goes down. In smaller boxes, you can get a boost before rolloff this way.
Measure T/S before gluing it down to verify which way you prefer it In your application. Use a 5 minute 2 part epoxy and mark the magnet with a sharpie to make sure you get it right. Amazon has Ferrite magnets, and likely Neo rings too.
Have a look at @Kwesi's thread (first post, scroll down a bit):
https://www.diyaudio.com/community/...tbcd-gb-dxt-sb-acoustics-sb15nbac30-4.391310/
@Kwesi, I wonder between the sb34RNXL (50/55L )and two 26 Discovery (60L) as the load is near will sound the best ? For sure the two 10" is good for a less intrusive enclosure. (at least pyramids enclosure are hard to do for 99.9% of the diyers)
You had the chance to listen the SB34RNXL ?
I wonder as well if a half way has any advantage over 2 // when it comes to low end Xmax and when one has an Active system ?
Also what Martin Logan did with two 10" push pull or push-push, I dunno, was puting them back to back, so ione forering towards the front wall with some dsp cancelation to figth the out of phase in order the phase is summed at the front. The advantage if any, at least a design choice, is the two driver are close to the floor so have some enforcment and one reflexion less.
I think hifijim has a good trilogy of drivers to choose there, not saying in a half way one can choose assymetric driver size (the mid driver with more efficienty and lesser diameter). Dunno how to marry the patern of a wide dome with a woofer though. Just guessing...
You had the chance to listen the SB34RNXL ?
I wonder as well if a half way has any advantage over 2 // when it comes to low end Xmax and when one has an Active system ?
Also what Martin Logan did with two 10" push pull or push-push, I dunno, was puting them back to back, so ione forering towards the front wall with some dsp cancelation to figth the out of phase in order the phase is summed at the front. The advantage if any, at least a design choice, is the two driver are close to the floor so have some enforcment and one reflexion less.
I think hifijim has a good trilogy of drivers to choose there, not saying in a half way one can choose assymetric driver size (the mid driver with more efficienty and lesser diameter). Dunno how to marry the patern of a wide dome with a woofer though. Just guessing...
Not sure why this is a problem: at 500Hz, a 10" or 12" driver doesn't beam at all; ditto a mid dome.
1) TSP. The 26W Disco has pretty high VAS (well, in this case I wanted low suspension stiffness that is dominated then by pretty linear air stiffness), and in only 30l for each you get quite high Qb >0,85-0,9. The 115x50x20 Y35 bucking magnet increases BL by measured whopping +10% and accordingly lowered Qes --> Qts to achieve a nice Qb of <0,75 in the given enclosure then (my usual goal for CB is 0,7 - 0,75 for active, 0,8 - 0,85 for passive speakers).
(Tested a bucking magnet on a 15M8631G00, and change of TSP was exactly 0%. So you have to check and measure on the specifiy driver if there is some potential to gain.)
2) Improving drive system magnetic stiffness. More saturation of the iron gives less modulation of the static drive field by the dynamic coil strayfield. Better counterfield stabilty --> less hysteresis distortion. So, at least in theory. In reality you measure slightly decrease of Le by just a few percent as improvement on the 26W Disco with buckings...
(Completely saturated maget drive circuits typically have really high B in the gap <2 Tesla, e.g. modern compression drivers, Bliesma and some Morel and Sica tweeters, some have low Le also without shorting rings. )
3) Very dependent on the Iron geometry, there could also be negative impact on Bl(x) linearity. I cound not find a crosscut of the 26W Disco, but got some hints: 26W might have a straight inner pole piece slightly prononged over the outer pole plate. (Alu shorting ring might be on the ground of the gap / on the pole shoe). Due to my brain simulation this is one of the more suitable geometries for field strenght increase (have some past background, did FEM simulations for electromagnetic actuators used in automotive and space applications during my bachelor and master thesis).
E.g. a 26W Revelator has SD drive, this means T-Pole with shorting ring above and below the pole (and one outside the coil alog the magnet), with a comparably thin pole plate and long coil. This in general will give a very linear drive force within the coil overhang, but it drops massively as the coil leaves the gap. A stronger saturation might force this behaviour - so bit better linearity within, but more massive drop below overhang. But perhaps magnet system is optimized by manufacturer like with the 15M I measured, and the bucking magnet will have no effect.
BTW: that a bucking magnet shows no effect does not mean the iron is fully saturated overall, but it shows that there is a dedicated part of the crosscut that is in saturation and its resistance so dominates the magnet flow. This helps with stiffness at lower frequencies, when they couple well to the core. But other parts of the crosscut might not be saturated, typically at the gap of a T-Pole when the diameter increases out from the core again. Locally the coil field can repress the field in the pole still dynamically. The both shorting rings there fight this behavior.
The 26W has no T but straight pole, so I just try to push that more in saturation. Prolonged inenr pole piece helps that Bl(x) assymetry gets not worse.
Best regards
Peter
Last edited:
Hmmm some shorthand I'm not quite getting. I think you're saying with DSP, it'll all work out? 😀
I agree that there's more than just directivity to matching drivers. A heavier subwoofer cone might not be as nimble as a light mid dome... but a powerful motor should compensate for higher mass. We'll see what Jim finds out.
Sorry never had listened to them. But would be one of my first choices when I would be searching for a reference 12".
2x Scan Disco have ~700 cm^2 SD, one SB34 has ~500 cm^2 SD. With SB34 having more linear stroke VD comes close.
2x 10" will have wider horizontal but more narrow vertical directivity than a single 12".
But with 2x 10" you are flexible to put a coil in series to the lower woofer, fading it out to have a more point source at higher frequencies. So you are close to the midrange at 500Hz xover, but still near the floor at low frequencies for good coupling. Or, you put each one 10" above and below the mid-high unit and have acoustic center between, and excite the room differently if this gives advantage.
Beside the driver quality, I see advantages and more flexibility in the speaker architecture with 2x10" than with a single 12" (I would also prefer 2x10" aesthetically...).
I'm excited about following your project. I got the M74A on hand as well as four 26w/8867. I have subs so no need to go down low. Haven't decided on the two 8867 or a TD15M per channel. I use the TD15M now from 100-700hz and I could not be happier.
For your project I would go with the 8867, excellent drivers.
For your project I would go with the 8867, excellent drivers.
@hifijim
I was not aware that you will be using FA253. For that assumption, I made a gross error in my previous SPL calculations for the woofer choices in post #52
For 250+250+100W (4 ohms) we're talking about a voltage swing of ~31V + 31V + 20V before the onset of hard clipping.
With reference to 4 ohms, that is 21, 21 and 17dB gain. For a moment let's assume the Hypex can more or less maintain these voltage swings up to 8 ohms or down to 2. Please refer to FA253 documentation to check the the exact voltage.
With 21dBW of gain with that amplifier on the midrange, the M74A at 95 dB/2.83V reach 116dB without the amplifier ever clipping, which allows all the transients to remain clean. I believe this is one of the reasons the old ATC 3" super dome and Volt VM752- at 94dB/2.83V is was well liked- 3" cones can match it in dispersion, but I've not seen one that can match it in sensitivity or power handling.
Moving onto the HF unit. 17dBW will allow at 93dB/W tweeter to reach ~110dB without clipping.
Suppose you want to reach THX / Dolby Atmos theatre standards of 85dB at listening position with 20dB of headroom. At typical 10 ft listening distance, that 105dB means 115dB @1m. So not quite. The are a few tweeters/CD that can reach 115dB/m. On the other hand, there is a lot of data to show that there is less musical content in the top 2 octaves, so 110dB is enough.
Depending on where you crossover, woofers need to play up to 500Hz-600Hz cleanly. To hit 110dB, with 21dBW gain from the F253, your pair of woofers will need to be approximately 89dB/W AFTER baffle step compensation to hit 110dB. So now we're looking at 93-94 dB/W in 2pi.
For the bass frequencies, the SPL limitations are often displacement limited instead of power handling. The RSS265HF will have a 2dB advantage below 50Hz, but both 26W8867 and it will reach 110dB in the midrange in 2pi.
Below is an example of the pair of 26W vs RSS265HF in 2 cu ft sealed cabinets. All SPL responses are 2pi.
- bottom 2 traces is the 2.83V sensitivity of a pair of RSS265HF-8 vs 26W/8867. The 26W/8867 clearly have an advantage here. In an active system it's less important because you can adjust your gains independently, but it's not completely unimportant...
- top 2 traces are the maximum SPL as determined by manufacturer's rated x-max and maximum continuous power ratings*
The RSS265HF-08 clearly have an advantage in the bottom octave as well as lower midrange.
- Middle 2 traces is most important to note, because it it determined by the F253:
The pair of RSS265HF-8 will reach 105dB @30Hz, limited by the amplifier at 31V. As you can see the purple solid line is lower than the dashed. This is where a Linkwitz Transform or some PEQs can shape the low end, and you can get the solid purple line to "hug" the dashed line (not shown)
The pair of 26W/8867 can't play as loud in the at 103dB@30Hz. This is reached at 19V. But the 26W/8867 has about 1dB more output from 100Hz up.
with another 4dB to spare (31V vs 19v)
The sound of solid state amplifiers clipping is not nice, like class A/B and Class D. You might not hear it in the bass, but in the lower mids, where your woofers are going to play to need to play to reach the 3" midrange (up to 2x Fc ?500-600Hz), but in the midrange this might be important. This is why available power and sensitivity can matter.
20 years ago Bob Cordell did a demo with his compact 3 way speaker. One of the tracks "Ghetto of my Mind", by Rickie Lee Jones, has a wide 24dB dynamic range. The lesson was that to reach live-life SPLs, one needs to ensure that the amplifier doesn't clip. IIRC, he had a 400W/channel amps at this demonstration.
Where does this all sit, as a comparison?
+ A lot of bass (too much!)
The debate between passive, active, sealed or bass reflex will go on. But I hope people can see the pros and cons of each.
At the end of the day, it still depends on what size /form factors of speakers you like (tolerate), you music you listen to (who needs dynamics when listening to modern pop with DR <10dB), how you you listen...
I was not aware that you will be using FA253. For that assumption, I made a gross error in my previous SPL calculations for the woofer choices in post #52
For 250+250+100W (4 ohms) we're talking about a voltage swing of ~31V + 31V + 20V before the onset of hard clipping.
With reference to 4 ohms, that is 21, 21 and 17dB gain. For a moment let's assume the Hypex can more or less maintain these voltage swings up to 8 ohms or down to 2. Please refer to FA253 documentation to check the the exact voltage.
With 21dBW of gain with that amplifier on the midrange, the M74A at 95 dB/2.83V reach 116dB without the amplifier ever clipping, which allows all the transients to remain clean. I believe this is one of the reasons the old ATC 3" super dome and Volt VM752- at 94dB/2.83V is was well liked- 3" cones can match it in dispersion, but I've not seen one that can match it in sensitivity or power handling.
Moving onto the HF unit. 17dBW will allow at 93dB/W tweeter to reach ~110dB without clipping.
Suppose you want to reach THX / Dolby Atmos theatre standards of 85dB at listening position with 20dB of headroom. At typical 10 ft listening distance, that 105dB means 115dB @1m. So not quite. The are a few tweeters/CD that can reach 115dB/m. On the other hand, there is a lot of data to show that there is less musical content in the top 2 octaves, so 110dB is enough.
Depending on where you crossover, woofers need to play up to 500Hz-600Hz cleanly. To hit 110dB, with 21dBW gain from the F253, your pair of woofers will need to be approximately 89dB/W AFTER baffle step compensation to hit 110dB. So now we're looking at 93-94 dB/W in 2pi.
For the bass frequencies, the SPL limitations are often displacement limited instead of power handling. The RSS265HF will have a 2dB advantage below 50Hz, but both 26W8867 and it will reach 110dB in the midrange in 2pi.
Below is an example of the pair of 26W vs RSS265HF in 2 cu ft sealed cabinets. All SPL responses are 2pi.
- bottom 2 traces is the 2.83V sensitivity of a pair of RSS265HF-8 vs 26W/8867. The 26W/8867 clearly have an advantage here. In an active system it's less important because you can adjust your gains independently, but it's not completely unimportant...
- top 2 traces are the maximum SPL as determined by manufacturer's rated x-max and maximum continuous power ratings*
The RSS265HF-08 clearly have an advantage in the bottom octave as well as lower midrange.
- Middle 2 traces is most important to note, because it it determined by the F253:
The pair of RSS265HF-8 will reach 105dB @30Hz, limited by the amplifier at 31V. As you can see the purple solid line is lower than the dashed. This is where a Linkwitz Transform or some PEQs can shape the low end, and you can get the solid purple line to "hug" the dashed line (not shown)
The pair of 26W/8867 can't play as loud in the at 103dB@30Hz. This is reached at 19V. But the 26W/8867 has about 1dB more output from 100Hz up.
with another 4dB to spare (31V vs 19v)
The sound of solid state amplifiers clipping is not nice, like class A/B and Class D. You might not hear it in the bass, but in the lower mids, where your woofers are going to play to need to play to reach the 3" midrange (up to 2x Fc ?500-600Hz), but in the midrange this might be important. This is why available power and sensitivity can matter.
20 years ago Bob Cordell did a demo with his compact 3 way speaker. One of the tracks "Ghetto of my Mind", by Rickie Lee Jones, has a wide 24dB dynamic range. The lesson was that to reach live-life SPLs, one needs to ensure that the amplifier doesn't clip. IIRC, he had a 400W/channel amps at this demonstration.
*Power ratings can vary between manufacturers
Dayton Audio RSS265HF-8
?How is this determined?
IEC 268-5? and 18.4 Rated noise test, or 18.2 long term power handling test?
eg. Scan-Speak 26W/8867T
Reference: Scan-Speak Technical Note 2
or some other standard? eg. AES-1984 standard 2 hour noise test, AES max power handling eg. SB Audience, ?DIN ie. 1KHz to 10 minutes THD <1%, eg.Morel
*x-max can vary between manufacturers eg. determined by voice coil geometry, Dumax or Klippel
Dayton Audio RSS265HF-8
?How is this determined?
IEC 268-5? and 18.4 Rated noise test, or 18.2 long term power handling test?
eg. Scan-Speak 26W/8867T
Reference: Scan-Speak Technical Note 2
or some other standard? eg. AES-1984 standard 2 hour noise test, AES max power handling eg. SB Audience, ?DIN ie. 1KHz to 10 minutes THD <1%, eg.Morel
*x-max can vary between manufacturers eg. determined by voice coil geometry, Dumax or Klippel
Where does this all sit, as a comparison?
+ A lot of bass (too much!)
- Needs twice as large cabinets (5 cu ft externally),
- Needs amps stable to 3 ohms.
- Only needs ONE good 12" woofer (in a right sized BR box) Great value!
- Can also have mighty dynamics with lower excursion!
- Only needs one 8 ohm stable amp, and less than 50W! (17V in Zmin 5.3 ohms= 50W)
- Someone has already done the clever work. Bravo @mbrennwa --> The diyAudio fan favourite OSMC
The debate between passive, active, sealed or bass reflex will go on. But I hope people can see the pros and cons of each.
At the end of the day, it still depends on what size /form factors of speakers you like (tolerate), you music you listen to (who needs dynamics when listening to modern pop with DR <10dB), how you you listen...
Last edited:
Well the Faital 12pr320 is not exactly a bass driver. Fs : 42 hz. To lowish the bass you have to use not the most musical BR alignment. You have anyway a lot of it coming from the vent + driver "overlap" !
I don't know if one has benchmarked such (non demodulated woofer) to a more hifi one like the SB34 RNXL.
Very interresting post tktran, thanks.
I don't know if one has benchmarked such (non demodulated woofer) to a more hifi one like the SB34 RNXL.
Very interresting post tktran, thanks.
Sensitivity and Xmax ratings of drivers vary a lot per brand. Also different motors have different tolerance beyond Xmax to Xmech. 20-30% distortion below 100Hz is difficult to hear in music until clipping or some noise overcomes audible. So it's very difficult to simulate real differences between drivers and enclosure types. Many amplifiers have soft clipping feature, sometimes adjustable like Hypex FA boards https://www.hypex.nl/products/amplifier-families/fusion-amplifier-family/fusionamp-fa253
In most European living rooms with semi-solid heavy walls modes at that spl make listening with average spl 95-100dB (with 20dB peak reserve) impossible and even dangerous. Still I undestand the concern of amplifier clipping or woofer hitting Xmech which however seems to happen mostly with reflex tuned systems...
A small error, actually, and a harmless one. I understood your point. 👍
My analysis closely matches yours. The RSS265HF-8 and FaitalPro 10RS430 are both excursion limited at 30 Hz, i.e. they soak up the full 250W of the hypex amp and remain within Xmax. They both perform about the same here, hitting about 104 dB at 30 Hz. The 26W/8867 is excursion limited at 30 Hz, and hits 103 dB at Xmax at 150W.
The FaitalPro 10RS430 is a promising driver, thank for making me aware @markbakk. It has very high excursion capability (matching the RSS265HF), but a much higher sensitivity (similar to the 26W/8867). The downside is that the rolloff begins at 200 Hz. This means it requires more EQ in the 40 - 100 Hz range than does the 26W woofer.
Of course all of these drivers will play higher SPL at higher frequencies, but that is mostly irrelevant. With my current system, the only driver that has ever come close to running out of SPL capability is the SB34NRX woofer, in the 30 Hz range. So as long as this new system is equal to or better than the current woofer at 30 Hz, I am not concerned with capability at higher frequency ranges. I am sure it will be enough.
Just want to point out that "cheep" 8,5 inch Seas H1208-8 L22RN4X/P is a really great driver.
Had 2 ported in 78 liter per side, and bass was amazing depth & clean.
https://www.seas.no/index.php?optio...n4xp&catid=44:utv-prestige-woofers&Itemid=238
Here simulated in Winisd 2 closed in 32 liter and 2 posted in 78 liter tuned to 30 hz.
regards John
Had 2 ported in 78 liter per side, and bass was amazing depth & clean.
https://www.seas.no/index.php?optio...n4xp&catid=44:utv-prestige-woofers&Itemid=238
Here simulated in Winisd 2 closed in 32 liter and 2 posted in 78 liter tuned to 30 hz.
regards John
