Scanspeak B1371

[Matt ITD]

Hi all, I'm thinking about tackling the scanspeak B1371 by PBN audio. Has anyone built them or heard them. There doesn't seem to be a lot of discussion or reviews on the design.

I'm looking for a high end scanspeak design. 3 or 4 way, passive crossover. Music is my thing, with about 30% movies. Room is 5 x 4 meters. Wanting an efficient design, my woodworking skills are quite good.

Any feedback on the B1371 is appreciated.

[Evil Twin]

If you are looking for feedback based on personal experience, it may be likely that there is no one who has purchased and constructed this kit on the forum.

On the other hand, reviewing the data provided by PBN, some comments do come to mind, based on technical interpretation of what they have published and experience working with the same (woofer, tweeter) or similar drivers (midrange) from ScanSpeak.

While some might pejoratively label me a "Negative Nancy", I prefer to describe myself as an informed skeptic.

So, only reviewing the data presented, as if a friend or acquaintance had come to me, saying, "what about this design? how does it look to you?"


Let us examine the provided SPL plot first...




There are three things I immediately noticed.

1. The woofer level is shown at about 93 dB- and this is for a driver operating in a cabinet which at low frequencies has to drive a larger acoustic space than the 2pi space above the front panel baffle transition frequency. But the raw driver sensitivy in an IEC infinite baffle test setup is only 90 dB.
2. The woofer response is down about 15 dB a the port tuning of 20Hz (as derived from the impedance plot)- this implies a rather ineffective Q and coupling of the port system. The response is flat to 40Hz but rolls off rapidly below that. Perhaps a less than optimal design in order to keep the enclosure size reasonable? or simply a mistake in choosing the port tuning frequency...
3. The SS 7140 appears to be drivin by a textbook flat filter, wtih no shaping of it's inherent "non flat" response. Due to time offset issues, a low order roll off is maintained down to 1500Hz, probably because of phase issues due to differences in the Z axis origin.
4. They draw a flat line for the target response, and put it on the graph, nominaly to show the target characteristic, but this is a "visual trick" to make the system response look more consistent.

The impedance characteristic is demanding, dropping to three ohms at more than one point, so amplifier choice will be important. This is somewhat typical of "classic" constant voltage crossover networks, as opposed to Linkwitz-Riley implementations or others that achieve a more uniform power transfer without unreasonable impedance characteristic.




In comparison, a floor standing three way design of similar height on this site created by Jonmarsh is relatively easy to drive...




With the impedance varying between 5 and 8 ohms.


Something that bears consideration is the interaction and placement of any system in the listening room.

Troels Graveson often includes some detailed in room measurements, especially for his larger kit systems.

These two articles might be worth reading...

http://www.troelsgravesen.dk/The-Loudspeaker.htm

http://www.troelsgravesen.dk/FUSION-3.htm

[Norm]

Ouch. Note to self - If submitting my designs to DFAL for peer review have bottle of Patrón Reposado Tequila and shot glass ready for support of ego while reading critique.

[Matt ITD]

I appreciate all perspectives and contributions. I am chasing a great scanspeak 3 or 4 way DIY design. My technical knowledge is fundamental so I do take to heart the advice of more seasoned designers/builders. Any recommendations for a better kit. Note I live in Australia (yes, its either on fire or flooding) and can purchase parts from suppliers here given our dollar is rubbish at the moment. I do love the aesthetic of the Respons Grand D (http://responsaudio.com/produkter/grand_d.html) but nothing on the internet on cabinet or crossover. All comments welcomed.

[Bear]

I appreciate all perspectives and contributions. I am chasing a great scanspeak 3 or 4 way DIY design. My technical knowledge is fundamental so I do take to heart the advice of more seasoned designers/builders. Any recommendations for a better kit. Not I live in Australia (yes, its either on fire or flooding) and can purchase parts from suppliers here given our dollar is rubbish at the moment. I do love the aesthetic of the Respons Grand D (http://responsaudio.com/produkter/grand_d.html) but nothing on the internet on cabinet or crossover. All comments welcomed.

Troels has a large number of ScanSpeak options, though you mostly need to buy them as kits from Jantzen. The SBA Satori kits often provide better value for those with currency challenges.

[Matt ITD]

Troels has a large number of ScanSpeak options, though you mostly need to buy them as kits from Jantzen. The SBA Satori kits often provide better value for those with currency challenges.

Yes, thanks Bear. I have looked at Troels kits and am aware you have to buy the kit through Jantzen, something that I don't want to do. I totally understand why he doesn't publish the values of his crossovers.

Perhaps I am being to narrow in my design criteria (Scanspeak 3 or 4-way passive floor standing, musically). A little background, my father purchased a pair of Tannoy monitor golds 15" back in the 70's and still has them to this day. He swears by them as being a true to life reproduction. I have listened to them often and while 'there is no replacement for displacement', I think I can do better at reproducing sound. Scanspeak, as I understand are first class drivers as they are used by many loudspeaker designers for high-end systems. The Scanspeak B1371 appealed to me for a number of reasons including, Full range, Scanspeak drivers, not a difficult cabinet to build, produced by a reputable company.

Perhaps I might ask, what is your holy grail DIY build and why?

[Carl V]

Rick Craig of SELAH audio has done designs like this.
3-way Scan speak drivers...as well as Satori, etc.,

[Steve Manning]

Matt, I built this from Troels site a while back, nice sounding speaker. http://www.troelsgravesen.dk/Jensen.htm. This is prior to him hiding his crossover values. Here is the link to my version of it. http://www.htguide.com/forum/showthr...ht=jensen+1071

Steve

[Evil Twin]

Ouch. Note to self - If submitting my designs to DFAL for peer review have bottle of Patrón Reposado Tequila and shot glass ready for support of ego while reading critique.

I salute your choice of beverages.

Let us put things in perspective, and consider that we should hold a commercial firm like PBN to a higher standard than an amateur DIY hobbyist like myself...

[Matt ITD]

Matt, I built this from Troels site a while back, nice sounding speaker. http://www.troelsgravesen.dk/Jensen.htm. This is prior to him hiding his crossover values. Here is the link to my version of it. http://www.htguide.com/forum/showthr...ht=jensen+1071

Steve

Thanks Steve, I appreciate your thread. I sympathise with the ups and downs of DIY. Sadly the crossover values are no longer published from Troels.

[Evil Twin]

Some general thoughts...

some considerations I use...

I find it is key to select a midrange driver without cone or surround resonances in it's operating range.


Let's compare the Revelator midrange...







With one of the better recent Accuton ceramic models, the C90-6-724.







Or for that matter, one of the SB Acoustics Satori paper cone midranges.




Examine the impedance curves of the midrange drivers in the 500-2,000 Hz area- the bobbles you see are the EMF from mechanical resonances in the cone and surround which create irregularities in the simple SPL frequency response as well as energy storage that shows up on a cumulative spectral decay plot.

It takes considerable effort to get a pistonic cone behavior in this region, and is quite difficult to do with paper- but not impossible- witness the new PuriFi 6.5" midwoofer.

Drivers such as the Accuton ceramic cone midrange may also achieve freedom from resonances in this area. This is a region where the human ear is quite sensitive, and where much of music lies.

This is why in the first design I published on this forum, targeting a low cost and high performance, I chose the Dayton RS180 aluminum cone drivers, because there other issues at higher frequencies could be dealt with, but the behavior in the critical midrange region was quite good, though the nonlinear characteristic of the motor could be improved on.


There are many companies offering kits and you could do much worse than to at least read through Troels's offerings, to gain some background knowledge, and perhaps to review the "proven designs" listed on this forum.

I would not claim there is any "holy grail" to be found... but you might be surprised by the range of projects and results available. I suspect you've hardly scratched the surface here.

[Evil Twin]

This thread may be useful...

http://www.htguide.com/forum/showthread.php?28655-A-Guide-to-HTguide-com-Completed-Speaker-Designs

[Matt ITD]

This thread may be useful...



http://www.htguide.com/forum/showthr...peaker-Designs

Thank you. I have looked at that thread many times trying to balance taste, budget, skill level etc. I'm keen to move up to that next price bracket of DIY sub 4k.

[Steve Manning]

Thanks Steve, I appreciate your thread. I sympathise with the ups and downs of DIY. Sadly the crossover values are no longer published from Troels.

Matt go to the bottom of the page for the link I sent on Troels site. There is a picture of the "final crossover" that shows the values. If that doesn't work and your interested in the build, PM me, I have the values.

[Bear]

Yes, thanks Bear. I have looked at Troels kits and am aware you have to buy the kit through Jantzen, something that I don't want to do. I totally understand why he doesn't publish the values of his crossovers.

Perhaps I am being to narrow in my design criteria (Scanspeak 3 or 4-way passive floor standing, musically). A little background, my father purchased a pair of Tannoy monitor golds 15" back in the 70's and still has them to this day. He swears by them as being a true to life reproduction. I have listened to them often and while 'there is no replacement for displacement', I think I can do better at reproducing sound. Scanspeak, as I understand are first class drivers as they are used by many loudspeaker designers for high-end systems. The Scanspeak B1371 appealed to me for a number of reasons including, Full range, Scanspeak drivers, not a difficult cabinet to build, produced by a reputable company.

Perhaps I might ask, what is your holy grail DIY build and why?

As others have said, there is no Holy Grail to be found. However, you can embark on a quest to find it. Depending upon your narrative preference, that either ends with most of the knights being killed off, or the local constabulary arresting the cast and crew. Hard to say which is more likely.

There are a few designs on Troels' site which are more recent and have enough details to build sans kit. The SBA 741 build has reasonable drivers and form factor for my taste (not a HUGE box):

http://www.troelsgravesen.dk/SBA-741.htm

I'm curious to see a Textreme version of the Satori driver, but a full set of replacements is likely years away. The MW16TX drivers are first and they are taking a while to come to market. The others should come more quickly, but that may still be 1 - 3 years later.

Satori drivers are a full step down in price from typical ScanSpeak pricing, though their performance is generally at-or-better than mainstream Classic or Revelator drivers. That's free money for the currency challenged. In several ways, the SBA aluminum/aluminum ceramic drivers represent an even better price/performance ratio (I haven't warmed-up to the CAC visual yet...). My next build is going to either use MW19P woofers or the SB17NBAC woofers in a "simple" TM. I'm to the point where I'm arrogant enough to start embarking out on my own simply because no one seems to have built the speaker version that I want. The MW19P currently has the nod because it can handle a slightly lower crossover to subwoofers, which means more transient power handling than the SB17NBAC. But the HD performance on the SB17NBAC is hard to beat at any price point, let alone <$100/driver.

One final piece: I really like the sculpture/artistic side of good woodworking/cabinetry. Some of the larger towers are simply gorgeous. For my actual needs in a smaller, acoustically-challenged room, I keep coming back to the TM and MTM form factors since most large three-ways still require a subwoofer for the bottom-end. Ultimately, I'd really like to do a three-way with something like the Wavecor SW223 or SEAS L26ROY, but such projects are several years out due to a lack of appropriate workspace (I'm in an apartment now, and will need to acquire some maker space slots to get the TM done).

[dwk]

Yes, welcome to the inherent problem in DIY - how do you decide what to build when you can't listen to or evaluate more than a tiny fraction of what is available?

Path 1 is what Jon and ET pursue - buy some of everything and do extensive testing (generally with a particular design goal in mind) before moving forward. Not really open to most folks, though - expensive, time consuming, and requires fairly deep expertise to evaluate and interpret the results.

Path 2 is to get in as far as you can, and take a leap of faith. Pretty intimidating when you're talking about something like your $3.5k budget, to be sure.

I'm in a somewhat simliar position to you. I'm hoping to build something inspired by the Sonus Faber Elipsa (or, in DIY terms something like Troels' Poor Man's Strad), so a 3-way with an 8-10" woofer and a typical mid/tweet arrangement. I have the benefit of having a MiniDSP SHD to handle the crossover between W and M/T, and so can start with a 2-way and add the woofer if needed. That's not really how true 3-ways are designed though, so I'm not sure whether it's actually how I'll proceed. My current concept is basically Troels SB941 using the Satori drivers: WO24, MR13p and the BE Tweeter, but

a) I'm not thrilled about ordering through Jantzen
b) I'm not thrilled about the stepped baffle
c) Since I'm targeting a wider baffle, it's a fair bit to spend on specific xover parts to then go and rework them

So, I'm not quite sure whether I'm going to suck it up like Bear and start considering designing my own xover, or whether I'll select another M/T as the starting point. My hang-up is just that there are so many interesting options to explore - a wide dispersion design like the DA25Tx makes possible? A waveguide design following Revel and others (SB26CDC in one of augerpro's waveguides). Is BE worth it on the tweeter? If so, is the Satori preferable to the SB29BAC or even the various Scan Speak units? Is waiting for the Textreme Satori drivers the best plan? Stated another way, what I really WANT to do is Path 1 but I simply don't have the money, time or experience to really do it. So I'm forced into Path 2 but am standing on the precipice pushing off the leap for as long as I can.

Someone else mentioned Rick at Selah Audio, and I'd also suggest looking at what he offers. He seems to know what he's doing, and his kits have a remarkably small premium over the parts cost. His Illuminismo looks like a great design at a $3k price point if you can accept the size - it's an MTMWW design using the same Satori drivers as Troels 941. I think that if he offered a simple TMW version that I might actually just go that route. the only real downside to Selah designs is that he keeps his crossover designs proprietary - you just get an assembled xover, not a schematic. Not ideal if you're a hands-on tinkerer that really wants to understand things and/or tweak.

[Bear]

So, I'm not quite sure whether I'm going to suck it up like Bear and start considering designing my own xover, or whether I'll select another M/T as the starting point. My hang-up is just that there are so many interesting options to explore - a wide dispersion design like the DA25Tx makes possible? A waveguide design following Revel and others (SB26CDC in one of augerpro's waveguides). Is BE worth it on the tweeter? If so, is the Satori preferable to the SB29BAC or even the various Scan Speak units? Is waiting for the Textreme Satori drivers the best plan? Stated another way, what I really WANT to do is Path 1 but I simply don't have the money, time or experience to really do it. So I'm forced into Path 2 but am standing on the precipice pushing off the leap for as long as I can.

About 10 years ago, Zaph ran his first contest for people to design a TM using measurements he provided for the ZA14 and the SB29RDNC tweeter. I pulled those files a couple of years ago and started playing with them in PCD. Having access to common measurement files and being able to see the various winning designs helped a lot. After that, pulling the various FRD and ZMA files from Parts Express helps to start seeing whether driver pairings work. Of course, those are large baffle files, but you do get to see what an utter PITA it can be to get a nicely curved midrange frequency response without any kinks or issues. Plus, that pesky phase relationship... Did I say arrogant? I think Hubris is the more encompassing phrase.

For Brandon's/Augerpro's waveguides, I have a distinct preference for 25um Resin prints over nylon/SLS or PETG. The Nylon/SLS is much better than the PETG at the same resolution (100um), but the resin is more of what you want for a "finished good". I haven't attacked either the SLS or PETG prints with sandpaper, so the ridges may ultimately smooth away without too much work or compromising the print. For the SB26ADC, that phase shield can be a challenge to get right (my SLS print has the phase shield bowed-out). The 0.75" deep version is too big for a resin print according to 3dhubs (who I've been using), but it would fit if the print axes are swapped around a touch. That's something to pursue next. Otherwise, I'm going with some CNC machined waveguides that we part of the DIYAudio group buy for ScanSpeak tweeters. I happen to have a pair of AirCircs sitting in my long-term inventory -- for about the last 10 years.

The 0.625" deep waveguide for the SB26xxx is just about perfectly matched for the SB15NBAC. The 0.75" deep ones are a good match for the SB17NBAC. The MW19P, on the other hand, needs close to 1.25" (31mm). If I could get resin prints for the 0.75" deep waveguide, then I'd just go with the SB17NBAC for my fronts and pocket the "savings".

[ergo]

I've done the printing with PETG and it does clean up very ok with a bit of sanding. And this sample was just a bit of quick sanding. My friend did some more pre painting but I don't have a picture of that. For test I deemed this sufficient.
Prusa allows to do variable thickness printing, so for WG that helps as the curve becomes pretty smooth toward outer rim

[Bear]

Looks good, Ergo! The SB26 waveguides have a phase shield which complicates sanding (thin/small parts sensitive to deflection/breakage), but for a waveguide that doesn't have that, this looks like a viable approach. I've also read reports of people using primer, as well, to fill-in the valleys between the ridges. Sanding and more sanding is the common theme, though.

[Evil Twin]

Looks good, Ergo! The SB26 waveguides have a phase shield which complicates sanding (thin/small parts sensitive to deflection/breakage), but for a waveguide that doesn't have that, this looks like a viable approach. I've also read reports of people using primer, as well, to fill-in the valleys between the ridges. Sanding and more sanding is the common theme, though.

Just what a wires, sparks, and light sabers kind of guy loves to hear... more sanding!

[Evil Twin]

One more thought to share about multi-way systems... crossover design, and the impact on group delay and impulse response.


I am going to reverse the order here, by presenting the punchline first, then a brief explanation.



First, from a recent Stereophile review, the step response of the Magico M2, a large floor standing three way speaker from Magico of prodigious construction quality...






Next, the step response of the Vimberg Mino three way floor standing loudspeaker (the "budget" division of Tidal Audio, the speaker builder in Germany), as measured by the European magazine Stereo Times:





Both are three ways, but their relative time coherence and group delay are quite different. To a knowledgeable practitioner, this step response tells the story of the crossover topology and implementation.

The interesting thing to me about the Mino step response is how very similar it is to the current step response simulation in LspCAD of the Ardent D concept development, the floor standing three way currently under construction and documented in another current thread. So similar, in fact, that I am willing to say the crossover topologies are identical- which implies that the Mino is using something very much like a Duelund three way crossover. The Duelund concept is know by each filter element being derived from a single unified equation which only totals one 360 degree phase rotation over the complete range (like a typical two way system) and when properly implemented, the drivers are always in the same relative phase to each other...


Here is a partial screen shot from my MathCAD working sheet:




And here are graph plots of the target functions and target phase response from an Excel worksheet: