Low cost high value 4 way...

[tktran]

After reviewing some measurements from the HifiCompass.com, it seems that it's steadily becoming the most extensive collection of drivers on the Internet.

I'm registered user, and with that you can click and compared up to 3 drivers at once in your browser. To that allows you to compare, for instance, a 19mm dome tweeter to 25mm tweeter to a 34mm tweeter. Or three 29mm tweeter. Or an 8" , 10 and 12" woofer.

Or, what I've been doing to selecting my T, M and W for my 3 way.

It's be interesting how this this data could be easily imported to a baffle/cabinet modelling program.

For instance, my ultra cheap 3 way would be
Scan-Speak 26W/8534G, Scan-Speak 10F4424G00 and DX20BF00-04. Starting crossover points 250-400Hz and 3-4KHz...

Or maybe triple 8" woofers, WWW, LM, M, T

[tcpip]

For instance, my ultra cheap 3 way would be
Scan-Speak 26W/8534G, Scan-Speak 10F4424G00 and DX20BF00-04...

Can't help remarking: I never thought I'd see "ultra-cheap" and Scan-speak in the same description.

Sent from my HTC U Ultra using Tapatalk

[5th element]

Can't help remarking: I never thought I'd see "ultra-cheap" and Scan-speak in the same description.

Sent from my HTC U Ultra using Tapatalk

Yes that's my problem too.

Ultra cheap three way for me would be Reckhorn D165 on bass. Vifa TC9 on midrange and a DX20BF00 for the tweeter.

It isn't sensitive but it's compact, low distortion and goes deep. Sounds perfect for most applications.

4 way? I don't see the point. You only need to go this route if you're after meeting some specific goal, usually higher sensitivity with lots of low bass. It's more of a high efficiency three way with an active subwoofer element to it.

[tcpip]

Ultra cheap three way for me would be Reckhorn D165 on bass. Vifa TC9 on midrange and a DX20BF00 for the tweeter.

I'm willing to go upto Dayton RS drivers and a suitable tweeter. If I use the RS100 as a midrange, I can use the XT25SC90-04 Vifa ring radiator too, above 2.5KHz or higher. That's a cheap tweeter, but nice if crossed over high, I'm told?

I just checked the Reckhorn. That's one ultra-cheap animal! EUR 30 for a neat little deep-bass woofer. It'll beat the Dayton RS stack I proposed easily. An Aurasound 3-incher can be an alternative to the TC9... Interesting.

4 way? I don't see the point. You only need to go this route if you're after meeting some specific goal, usually higher sensitivity with lots of low bass. It's more of a high efficiency three way with an active subwoofer element to it.

I've thought of one more case where I may consider a 4-way: first order crossovers. Won't those shallow slopes force one to use narrow passbands, thus increasing the number of ways in most cases? For adequate power handling, the tweeter will need to come in only above, say, 3K or 3.5K to prevent it blowing up at high SPL. And so on... (Of course, it's possible you may find just the right mix of very wide-range drivers and 3-way may be adequate...)

My impression of the first-order challenge is that you need very well-behaved drivers which are smooth and linear a decade on either side of the pass-band, and since such drivers are hard to find, one lands up adding more ways to keep each driver within its well-behaved region.

[Zvu]

Are there any available measurements of Reckhorn D165 ?

[tktran]

Yes, it would be interesting to make 3 way or 4 way with $60 or $80 worth of drivers.
I'd added the DX20FB00-04 to my shopping cart...
Which TC9 are we talking about?

Can't help remarking: I never thought I'd see "ultra-cheap" and Scan-speak in the same description.

Currently I'm Australia. Scan-Speak is well represented here, with the 26W8534G costing about US$100 and low cost road freight for a boatload of drivers.

I think the Dynavox LY302F is also objectively very good for ~$20, but like the Reckhorn D-165 is probably unobtanium for me.
I haven't seen a good 10" woofer for $20...

I think the minimum for the bottom of a 3 way should be a 10", or least a pair of 8s" to make the whole exercise worth it.

I'm moving to Canada in a couple of months... I wonder if the situation is much different there for DIY audio nuts...

[5th element]

I'm willing to go upto Dayton RS drivers and a suitable tweeter. If I use the RS100 as a midrange, I can use the XT25SC90-04 Vifa ring radiator too, above 2.5KHz or higher. That's a cheap tweeter, but nice if crossed over high, I'm told?

Oh you can use any drivers you want and the ones you proposed were good choices I just thought that ultra cheap it was not so chimed in with my ultra cheap idea

The small tymphany looks amazing, actually, at any price. Providing you don't need it to go low. The DX20BF00-04 that is. Really what more could you ask? No 4th or 5th order to speak of, even at elevated drive levels, ridiculously low levels of 3rd and what you'd expect for 2nd. It's literally faultless/scanspeak beater for $8 @ madisound's current sale and with a small face plate. 3kHz xover good enough? What are you waiting for! I'm not in America either but you get the point.

I just checked the Reckhorn. That's one ultra-cheap animal! EUR 30 for a neat little deep-bass woofer. It'll beat the Dayton RS stack I proposed easily. An Aurasound 3-incher can be an alternative to the TC9... Interesting.

Yeah it looks like ridiculously good value and it has a cast basket too!

I've thought of one more case where I may consider a 4-way: first order crossovers. Won't those shallow slopes force one to use narrow passbands, thus increasing the number of ways in most cases? For adequate power handling, the tweeter will need to come in only above, say, 3K or 3.5K to prevent it blowing up at high SPL. And so on... (Of course, it's possible you may find just the right mix of very wide-range drivers and 3-way may be adequate...)

My impression of the first-order challenge is that you need very well-behaved drivers which are smooth and linear a decade on either side of the pass-band, and since such drivers are hard to find, one lands up adding more ways to keep each driver within its well-behaved region.

Ah yes first order speakers...if one wishes to attempt the truly foolish (at least in my opinion) then going four way is probably necessary in getting as close to that goal as possible. I guess first order designs skip my mind as such a bad idea that they don't even register when I'm brainstorming :S

If I wanted to do one though the scan 10F would certainly be involved, plus one of scanspeaks neo, small dome tweeters, the chambered ones that can do a LR4 at 1.5kHz. Put them as close together as possible and cross at 2.5-3Khz. Should get about 2.5 octaves of 1st order roll off between them. Cant ask for more than that really. Then 1kHz cross to something like a satori mw16p, or 13p if you wanted better off axis. They can cross low so pop anything with bass below it. I know a scanspeak 32W/4878T00 crossed at 200-250Hz. There we go world class attempt at a first order hifi speaker with awesome bass extension. Could be a three way if you were happy with the satori as your bass drivers too. I would not be so 4 way it is.

Are there any available measurements of Reckhorn D165 ?

It seems odd that there are not. Considering its price you'd figure every one and their aunt would have one. Still I don't expect there to be copper in the gap. If the rest of the motor is well engineered (and I'd sure hope it would be coming from Germany!) I'd expect it to be good up to 300Hz or so.

[tcpip]

... Ah yes first order speakers...if one wishes to attempt the truly foolish (at least in my opinion)...

I too do not understand the fascination with first order. I've heard the arguments in their favour, but they put unrealistic demands on drivers, can damage your tweeter, and all for a final system which has good reason to sound worse than a conventional slope. I feel I'd start with 2nd order at the least for any conceivable situation.

I guess there are certain things I just have given up trying to understand. Full range drivers is one. I love FR drivers, but in at least a 2-way design. (I'm designing one, with the FF225WK and the AC G2Si.) Another thing I don't understand is this FR + super-tweeter config with the FR running full range and a single cap on the tweeter: most of the guys who build them don't bother to take a single measurement to see how they pair up. A friend who's a fan of vintage drivers has a 2-way he likes a lot, and it has 15" alnico drivers and small horn tweeters. A quick measurement of the rig showed that the two drivers overlap from 500Hz to 5KHz, both running flattish SPL curves in this massive overlapping range. This is what its crossover delivers! The crossover applies only to the tweeter; apparently the Prescribed Way to set up this rig is to run the 15" driver full-range. What makes a person in the 21st century choose such a system?

I used to believe until recently that the subjectivist audiophile fanatics with their cryo-treated cables and vibration dampers on amplifiers are the guys who believe in voodoo; we DIY guys are engineers. I now realise that there's as much voodoo worship in the DIY world too. We are just the weird engg minority caught in the middle. If the FR cult is not a religious belief, I don't know what is. They talk of "colouration" introduced by a crossover, and refuse to see the obvious, glaring colouration of the raw SPL of their untamed drivers with 6dB peaks and troughs. Tom Christiansen released the design of his Modulus-86 amp with sub-PPM distortion through the audio range, and people kept badgering him for months asking "How does it sound? How does it sound?"

I'm building an MTM with 6th order slopes, and I'm anticipating nay-sayers jumping at my design. Let's see how the speakers sound. I'm doing my 6th order take on the classic Dayton RS180 + Seas TDFC combo. Crossover designed, 1.5KHz Fc, waiting for parts now. Let's see if they are as good as Jon's Cauer elliptic xo.

Sorry about the pre-breakfast OT rant.

[Juhazi]

Even a 3-way sets high demand for response linearity with shallow slopes. Response even 2 octaves past the xo should then be smooth without low end roll-off or upper end resonances. Only then phase changes will match each other also for woofer and tweeter, below the mid. http://www.musicanddesign.com/Duelund_and_Beyond.html




Acoustic LR or Bessel 2nd order is considered best realistic approach, used by eg. Jon Marsh and Troels Gravesen in many desings - and me! It is rather easy to achieve with dsp-xo, but passive leads to uber-complicated jungle of L-pads etc. easily.

https://photos.app.goo.gl/dPN9P4v5NCSwX5T46

[tcpip]

Even a 3-way sets high demand for response linearity with shallow slopes.

In fact, a 3-way will be more difficult to do than a 4-way with super shallow slopes.

Acoustic LR or Bessel 2nd order is considered best realistic approach, used by eg. Jon Marsh and Troels Gravesen in many desings - and me!

Hasn't Jon done a 1st order design, the Kurosawa Koncept?

When someone like him does a 1st order, you can of course see how he's taking care to get it right. In the Kurosawa Koncept, I believe he embarked on the project because a very specific set of Accuton drivers became available.

[roadrune]

Troels have a 3-way 1st order http://www.troelsgravesen.dk/SBA-761.htm

[cochinada]

Ah yes first order speakers...if one wishes to attempt the truly foolish (at least in my opinion) then going four way is probably necessary in getting as close to that goal as possible. I guess first order designs skip my mind as such a bad idea that they don't even register when I'm brainstorming :S

Vandersteen speakers ring a bell?

[Juhazi]

Troels have a 3-way 1st order http://www.troelsgravesen.dk/SBA-761.htm

Troels claims 1st order mid-tweeter xo for two of his latest designs.

He writes:
"The MW16 is one of the few drivers allowing a 1st order crossover, thus just a single coil to this driver. Overall the crossover topology is much inspired by my two most recent constructions, thus no high-pass filter for the MW16 and having this in an smaller aperiodic cabinet allows quite some power handling.
The tweeter works 1st order down to around 2 kHz from where it gradually declines 2nd to 4th order."

Tweeter and midwoofers of SBA761 are connected in opposite polarity, which is typical for acoustic 2nd order slopes. He doesn't show individual measured slopes, or step response.

His Ellipticor-3 has same quote of 1st order, but similarly opposite polarity for mid and tweeter http://www.troelsgravesen.dk/Ellipticor-3.htm. It shows 1st order near xo then gradually steeper.


"There are very few true 1st order filter speakers in this world. Some may claim they do 1. order filters, but it comes usually comes with a grain of salt. The filter may be 1st order on the electrical side, but even so, the roll-off can be 2nd, 3rd or even 4th order from very simple filters - due to the drivers' inherent roll-off. And it all depends, because how far beyond the point of crossover must a driver go before we can call it a true 1st order filter? There are no rules for this, not even any general practice. Vandersteen is probably the company that does best in this area. My best bid on this issue is 2 octaves above and below point of crossover. Based on measurements in Stereophile, Vandersteen doesn't seem to always meet that target. Here the 18WE driver follows 1st order roll-off up to around 6 kHz, then a dip around 8-9 kHz and it stops around 15 kHz but never exceeds target profile. The tweeter is pretty much 1st order all the way from 15 kHz down to 800 Hz, where is starts rolling off 2nd order. This is pretty darn good!"

[JonMarsh]

Hmmm, four way speakers with first order acoustical slopes in the critical transition region (to -12 or -15 db)... that rings a bell. Probably what Vandersteen is often doing, and what I did in the late 70's with a gentleman by the name of Bill Kennedy that passed away in 1980 from an Aortic aneurysm.


Ah yes, those were the days. No PC software. No PC's, for that matter. We did have a B&K 4144 with an HP mic preamp (which Tom Waale has had now for years), and a 1/3 octave LED display audio spectrum analyzer with pink noise generator, but the main development tool was my Tek scope and looking at the impulse reponse of the system on the listening axis. A lot of tapped inductors... a lot of switched capacitor test boxes, so we could dial through a value range, the way you would scroll up and down on a PC software tool for a cap value.

One has to use the 1/3 octave analyzer sparingly, because it had no time domain rejection, i.e., no room acoustic rejection. The impulse response was pretty accurate tuning method for on axis, and when you could get a nice clean 1 or 2msec pulse response, well then, you knew you were doing things right from below 100 Hz to above 10kHz. In those days, that took a four way- an MB 3/4" dome tweeter (polycarbonate, or some other plastic with good self dating), Audax dome midrange, a fairly high end Audax 7" woofer, and dual 10" woofers in a sealed system- I don't remember the brand or model.

BTW, I chuckled at the idea of low cost four way system- I can readily imagine finding a set of drivers that might seemingly fulfill that idea, but once you build a good 4 way crossover, even 1st order implemented properly (with all the impedance control and equalization you'll need, low cost doesn't strike me as being part of the equation.

But as Matt points out, cost is all relative.

[Scottg]

-"mostly" 1st order *series* has some advantages, often including resulting cost.


Of course with cost it's down to crossover component cost vs. driver cost. If you select really inexpensive crossover components then there might not be much advantage (if any). And this is not taking into account cabinet cost/complexity - with for a manufacturer would be significant, quite likely more so than their bulk-industry purchases on crossover components (..but which the typical DIY'er doesn't really account for).



As far as objective performance (withOUT respect to a lower order crossovers):

There can be an advantage with lowering non-linear distortion and IM distortion. (..which is why I found 5th's questioning particularly surprising given his ..um, "predilection" for lower non-linear distortion: use drivers where they objectively perform best!)

[5th element]

You do have a point there Scott but do subs qualify as 4 way? In a lot of commercial systems we see a 4 way having the bass driver only going as high as around 100Hz. This is more a 3 way with an integrated subwoofer in my opinion but if you want to reach 20Hz it's necessary.

My system is a 3 way as the mains have 4 10" drivers of their own, when correctly configured you get ample bass extension and they'll play loud too. The way I use them though is high passed, albeit low, then blended to multiple subs that go up to 150Hz. Now it'll play even louder and go lower. 3 way plus subs or really a 4 way?

Looking at the first 10 pages of stereophiles floorstanders shows the only true 4 ways really as being the ones that use 2" domes or similar. These could be done as three ways with 8" bass drivers but as the dome needs an 800Hz xover the designers have used smaller lower mids and then gone with a much bigger bass unit for real bass extension. Crossing at ~200Hz to the bass. Mainly vivid speakers.

Ironically the vandersteens cross low ~100Hz to their lower bass drivers.

The JBL 250ti looks ill conceived and the genesis 5 way! Is termed incorrectly. A line of drivers tweeter, two 3" dome mids and two bass units. It's not a 5 way it's just a three way with .5 units. The top 3" dome crosses to the tweeter and goes down to ~250Hz with the lower 3" dome going just as low but rolling off around 800Hz to fill in for bafflestep losses on the first mid. The two bass drivers essentially do a similar thing with the lower woofer rolling off around 70Hz. This isn't typical 0.5 way as the 70Hz roll off isn't going to fill in losses on the first driver properly but is probably more of an attempt at improving the vertical off axis response. Ie this way you only have one lower midrange source to interact with the lower range of the domes rather than two if both bass drivers were in parallel.

It seems like a well thought out speaker or one that at least been thought out in a logical way, it's just a shame it measures poorly, I'm sure a different xover would solve that though and maybe a proper tweeter that can cross lower. That's if the planar tweet is setting the upper xover lower limit.

[Scottg]

Sub's qualify!

.but..

Very few use an effective low-freq. high-pass for their loudspeakers. :cry:

Though of course this is good in the context of swamping room-modes, but not so much as far as non-linear distortion.

[Scottg]

My system is a 3 way as the mains have 4 10" drivers of their own, when correctly configured you get ample bass extension and they'll play loud too. The way I use them though is high passed, albeit low, then blended to multiple subs that go up to 150Hz. Now it'll play even louder and go lower. 3 way plus subs or really a 4 way?

-Low-freq. high-pass: definitely a 4-way! (..at least in overall-design, if perhaps not a physically integrated loudspeaker.)


..and frankly a lot of driver's "peter-out" as far as good non-linear performance in that range between 50-100 Hz. Cutting-down-on excursion/amplitude anywhere near in-box resonance is going to be a "boost" to non-linear performance. And of course bandwidth limiting in general will massively reduce IM distortion.