Announcement

Collapse
No announcement yet.

listening impressions of some interesting drivers...

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • augerpro
    started a topic listening impressions of some interesting drivers...

    listening impressions of some interesting drivers...

    A friend of mine is outfitting a home theater and the speakers are falling to me to decide. Given the budget and goals we have settled on going with bookshelf size speakers all around, matched to 3 or 4 Peerless XLS 12" subs. So I rounded up some drivers and set to work on some designs and thought I'd share what I found. This all very subjective, so don't shoot the messenger, and keep in mind it is only my own perception and tastes expressed.

    The woofers are the Zaph ZA14 and Scanspeak Discovery SS15. Tweeters are Peerless HDS, Dayton RS28A, SEAS T25C001, and Vifa NE19VTC. Since there are some very good commercial designs in this price range I picked two: the Wharfedale Diamond 10.1 and Kef Q100. Actually the Kef was not the retail speaker, but the raw driver, more on that later. The DIY designs all used the Dayton .38cf cabinets with the same driver layout so baffle diffraction would be the same. I used the minidsp for crossover work, and I have to say this is an invaluable tool for prototyping and quickly comparing changes. The other key thing is these are comparisons between single speakers in mono, since I would need 4 of everything instead of 2 of everything to do direct A/B comparisons if I wanted to do it in stereo. I'm not sure how listening in mono may change the sound, so keep that in mind. I used a 2khz crossover for the RS28A, HDS, and T25C001, but as I've found in the past I like the HDS a bit higher, so it was at 2.3khz for the final reviews below. The NE19VTC was at 3khz. I've also been doing final voicing on my Bosebuster v2 during this time and those made it apparent the NE19 likes to be crossed a bit higher than 3khz to get the cleanest sound. The grittiness or spittiness goes away very quickly above 3khz and I'm finding 3.3khz-3.5khz is probably the sweet spot. Anyway for this comparo matched with the SS15 they were at 3khz, so there are some improvements to be made later. For the Kef I used a LR4 2.5khz crossover to match the retail speaker. Where they might differ is BSC. I basically matched all the speakers to be basically the same, about 4db of BSC. I've heard the retail Kef can sound a bit lightweight, so it is possible they use less BSC, and they may not sound exactly like my version. All speakers were level matched from 200-10000hz.

    I started with comparing the HDS/ZA14 versus the HDS/SS15, so basically just comparing the woofer. At 2khz these sound near identical, I mean it was really hard to pick out a difference. And this is even with turning off the tweeter to just listen to the woofer, a very enlightening thing to do, especially with tweeters. It really helped to narrow down which driver was misbehaving at times. I would find later the differences between these woofers become more obvious as you go up in XO frequency, with the ZA14 getting a bit forward and shouty by 3khz. But at 2-2.3khz, near identical. After a lot of switching back and forth my impression was that the ZA14 was just a touch more composed in the middle/upper midrange, and so this is the woofer I used when comparing the HDS, RS28A, and T25C001 to each other. Both ZA14 and SS15 had refined if not voluminous bass. Both stayed pretty composed when the cone really got moving, with maybe a slight nod to the ZA14 for being better behaved, but this is probably not much of a distinction since they both were well behaved right until they started bottoming at about the same point. So with the woofer decided on to the tweeters.

    HDS/ZA14 vs RS28A/ZA14
    I felt the RS28A was very clean and clear, but a bit more forward and harder or "clinical", while the HDS was more open, natural and spacious - vocals and many instruments just sound more real and present. The RS28A is very well behaved, probably more than any other tweeter I've used since the Scanpeak 9800 (which I've heard was the "inspiration" for Usher when making the RS28A style tweeter). The HDS will occasionally misbehave: there is some area on the low end of its range that will occasionally jump out at you. Not crossing too low helps a lot, but I really need to get around to using a waveguide to see if I can finally solve it, because to me the HDS is otherwise the best sounding tweeter here. Still both of these are very good tweeters, and if it sounds like I'm being a little hard on the RS28A with it being a bit harder or forward, it is only in comparison to the HDS in quick A/B switching, it is not a hard or forward tweeter in an absolute sense.

    HDS/ZA14 vs T25C001/ZA14
    These tweeters sound REALLY similar. In the end I felt the HDS was just a squidge more natural and spacious, while the T25C001 was a better behaved in that area the HDS isn't, but a bit harder and less spacious. I don't want to overdo the differences here though, they are very similar. If I can solve or just accept the occasional HDS misbehavior I'd go with it, if not the SEAS is probably the least troublesome pick while sounding much the same as the HDS in all the ways I really like about that tweeter.

    I also compared the SS15 with the above tweeters versus the ZA14 versions, but as I said there just isn't much difference at this crossover point so I won't bother with any impressions. Now that I found a good combo in the HDS/ZA14 I compared it to the Kef Q100.

    HDS/ZA14 vs Kef Q100
    This Kef is sort of a heartbreaker. It does some things SO well, and than it falters a bit. First the midrange and lower treble are really good: coherent and solid. Nothing wanders back and forth between tweeter and woofer. More forward than anything else here (due to the solid upper midrange and possibly the directivity), and most of the time this was a good thing: vocals are very clear and real, solid. Instruments in that area are similarly clear and solid. It was really very different in presentation from the other speakers in this regard. Usually a big, solid image. But...the forwardness could really get "in-your-face" at times. The tweeter is a bit rolled off in the "sparkle" area, and can get gritty when pushed. The treble is unrefined compared to most others here. Imaging is very good, but spaciousness was not that great, it could often sound small and speaker-like for lack of a better term on some music that didn't have much content in that prominent upper midrange area. Besides a bit of tweeter grittiness, it could also get congested and closed or "hooded" sounding. The HDS/ZA14 in comparison was more relaxed and refined. More spacious and more "3D" like imaging. More musical and forgiving. If it sounds like the HDS/ZA14 was hugely superior speaker, well it was, except in those areas I noted on the Kef, where it really was no comparison. I think if the Kef woofer breakup could be better controlled or brought down in level, and the tweeter brought to the level of some here, its weaknesses would melt away and you would be left with a very clear, and realistic sounding speaker with big, excellent imaging. If appears the Kef LS50 made improvements in exactly these areas, and they may very well be the speaker to have, doing some important things much better than most DIY designs.

    HDS/ZA14 vs Wharfedale 10.1
    The first thing I noticed about the 10.1 is how "immediate" and "present" it was. I really like this in a speaker. Combined with a warmer character the music is very involving and enjoyable, just pulling you in. Percussion was very clear and lifelike, the shimmer and sparkle regions are prominent and pleasing. Total lack of sibilance. Overall a very pleasant, well behaved speaker. Except...a hint of constant graininess in the treble. Nothing that ever really annoys you, just overall not as refined and clear as teh HDS, RS28A, or T25C001. Could get a bit congested and honky on more complex material or higher volumes. Sounded smaller and more speakerlike in comparison to the HDS/ZA14. The HDS/ZA14 on the other hand was simply smoother, more refined treble. More spacious and bigger image. Cleaner, calmer, and more composed everywhere. The bass stronger than the 10.1, and much more refined, the 10.1 could get a bit flabby in this area. On the other hand it lacked that immediacy and presence of the 10.1, sounding more detached and colder, maybe harder, and so for all its refinement and clarity, it lost some emotion and musicality to the 10.1

    HDS/ZA14 vs NE19VTC/SS15
    I must say the NE19/SS15 combo surprised me. But then I made this speaker because I suspect that sometimes the push to lower tweeter crossovers to better match woofer directivity results in a less pleasing speaker,and I wanted to test this out. The problem as we all know is the prominent dip/hump in the power response with a typical speaker. IMHO waveguides are the way to solve this, but the fact is this is not very realistic for most DIY'ers, or at least common yet. Sans waveguide, I'm not always convinced pushing the tweeter lower is always the best way to deal with the dip/hump. Often the speaker can become forward and harder. I suspect it may be because we still have a hump, but now we are pushing it down into the 2khz area where the ear is very sensitive. I wanted to push that hump up in frequency. There will be other tradeoffs I think, in clarity and other aspects, but I was betting the overall sound be a pleasing speaker, versus a clear but hard one. And that is pretty much what I got. They ended up having close (but not quite) the same immediacy and presence of the Wharfedale, similarly lush sounding, though I wouldn't say too warm. Just a musical, involving, pleasing speaker, that pulled you in. The HDS/ZA14 in comparison sounded more distant, and sometimes "flatter" somehow, less involving. OTOH it had a bigger and more 3D soundstage, with smoother and cleaner treble.

    continued...
    Last edited by augerpro; 12-28-2013, 04:48 PM.

  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Originally posted by johnastockman View Post
    Nice! The write-up is excellent and the info and graphs are very thorough.
    John A.
    Thanks. I'm hoping to get some off axis plots up soon. I'm really impressed by these little Founteks, you would assume they would be best with non-stressing jazz or something, but they are really in their comfort zone with rock and roll. Amazing how much those little cones can move with little change in the character of the sound.

    Leave a comment:


  • johnastockman
    replied
    Re: listening impressions of some interesting drivers...

    Nice! The write-up is excellent and the info and graphs are very thorough. I haven't seen many set-up/integration instructions, but yours are great. Love the effort and end-result.


    John A.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    FINALLY got my BoseBuster sequel done and written up!! I've shortened the name to BBV2 so I don't get any hassle from Bose. Next I'll be finishing the Scan Disco TM and MTM, the other flavors after. My Soma Sonus site is now hosting my projects. Here is copy of that page:



    BBV2

    A small format sub/satellite system


    The goal for this project was a small, high quality speaker that was shallow enough to be mounted on wall next to the thin LCD TV’s that are now common. The Fountek FR88EX was perfect for this project, with its low distortion and ability to fit in a truly tiny box (you can get the cones really dancing with little change in character). After several tweeters, I finally settled on the SB Acoustics SB26STCN-C000-4. This tweeter not only measures well, but is subjectively a very clear and open sounding tweeter. Crossover point is 2.8khz @ LR4. Here is the on-axis frequency response:


    With reverse polarity to check phase integration:


    The frequency response measurements consist of a far field measurement with a 6.5ms gate (no smoothing!!), merged to a near field measurement at 300hz, resulting in a quasi-anechoic response. The low frequency ripple is the noise floor of my shop.

    Impedance:



    Here is the crossover schematic:


    Inductors are 18g on the woofer and 20g on the tweeter. The small resistance value shown with each inductor is the resistance of the inductor itself, not a separate resistor.

    Baffle layout:


    The following is how I initially integrated the sub and satellites. This is intended to be a quick way to get the system up and running. Obviously there is always a lot more one can do to get the smoothest bass and mid-bass response, and this is only one simple method.

    First, I use two subs, one roughly at the foot of each main channel. In this case I tried REW as the measurement software since it’s free and easy to use. I use a mic from Cross Spectrum Labs. I didn’t bother with all the calibration of the measurement setup since we just want to see real-time in-room response for integration. Given that, it is pretty straightforward to setup: just open the RTA window, change the scale to go up to 2khz, 1/6 octave smoothing, select FFT Length 32768, Blackman-Harris 7 window, 8 averages, and that’s it. You might have to open the preferences and set the soundcard output if you get no sound, or the input if the mic doesn’t read anything. I aimed for a 175hz crossover, I’d like to experiment with 200-250hz in the future.

    1) First set the levels of each channel (without subs on), I use a Radio Shack meter on A weighted Fast

    2) in the AVR set the sub to “LFE+main” (this is on a Marantz, other brands should be similar)

    3) set the sub crossover to 250hz

    4) set the mains to small and crossover of 80hz

    5) starting with just the left channel outputting pink noise (sub off), adjust the crossover until your -6dB point is at roughly 175hz:


    My bass doesn’t appear to drop off quick, but that is due to the noise floor of my apartment being fairly high. Raising the signal will correct this, but I didn’t want to get my neighbors bent out shape!

    6) Now running only the left sub, adjust its level to match the main. I like a little boost in my home theater bass though

    7) Adjust the sub crossover until its -6dB is also 175hz:


    8) Now turn on both the main and sub, and check integration:


    9) Looks good so I moved on to the right channel

    I was lucky and integration was very good right away. All I really have left is to fine tune the sub response. I’ll probably have to move it, requiring me to go through the integration again. Anyone who measures their setup knows how badly the front wall (often in combination with the floor and side wall) can really mess up the response in the 100-500hz area. One of the nice things about a sub/sat system crossed around 200hz is that you now have drivers with different crossover slopes, and in two different positions in the room, to smooth out the response. At this point I would switch to 1/12 or 1/24 octave smoothing to make sure I wasn’t smoothing over the ugly parts.

    If one did not have good integration yet, you can adjust the phase on the sub, or adjust the delay setting in the AVR, or move the sub around, to get the proper phase match. Sometimes I find the response improvement to be ambiguous as I change settings. When that happens I like to run the main in reverse polarity, switch to 1/24 octave smoothing, and adjust delay/phase and sub placement to get the deepest null at the crossover. When you find it, you then switch the polarity back to normal and you should have excellent phase integration.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Was hoping to finish measurements on the 3-way (which I've dubbed Karnak's in homage to the incredible Egyptian temple) today, but looks like I'll need another day. I am extremely happy so far though. A LOT of baffle diffraction modeling, not just for the smoothest response, but a response that will work well with the drivers' response, and the intended crossover points. And a number of prototypes to verify the model. Not doing this well in the past has caused surprises after the speaker is built, and ultimately caused me to drop the project. Here are the results (6.5ms gate, no smoothing):

    TangBand 75-1558SE:


    BG Neo8:


    Transducer Lab N26C:

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Had to take a couple weeks off from these projects to get the house ready to sell. Today I was able to finish off the construction of the last speakers, all ready for measurements this week. The ones below use the Peerless Exclusive 8", but I have an Aurum Cantus AC200/50C2C and SB Acoustics SB23NRX on the way so I can make versions with those for friends.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Originally posted by jsr View Post
    Nice builds Brandon. The SB MFC and SB26SCTN are both drivers I've been interested in. The MFC version seems mow behaved at higher frequencies than the NRX according to SB's plots, but you say the NRX is better behaved. That's interesting.
    The BoseBuster looks neat.
    I had to reread my comments because overall I think the poly is noticeably better behaved than the paper. The high frequency hash just doesn't show up like it does with the paper (caveat: when crossed at 2.8khz, crossing lower the difference may not be as noticeable). What I did notice though, was that on a couple songs where there was vocal high note, the paper stayed a little smoother. But I may have overstated that being the norm, overall the poly is better behaved just due to the breakup not being as noticeable (probably as non-linear distortion).

    Leave a comment:


  • jsr
    replied
    Re: listening impressions of some interesting drivers...

    Nice builds Brandon. The SB MFC and SB26SCTN are both drivers I've been interested in. The MFC version seems mow behaved at higher frequencies than the NRX according to SB's plots, but you say the NRX is better behaved. That's interesting.
    The BoseBuster looks neat.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    So after 5 months I have distilled my experiences in this thread down to some finished designs that I just finished constructing. First is super cheap and easy Silverflute W14 and SB26STCN MTM, using the 4 ohm drivers in series for an easy load:

    Click image for larger version

Name:	IMAG0212.jpg
Views:	1
Size:	398.7 KB
ID:	1158136

    For my buddies home theater, SB15MFC woofers and SB26STCN in MTM or TM:

    Click image for larger version

Name:	IMAG0211.jpg
Views:	1
Size:	361.3 KB
ID:	1158137

    For my own home theater I'll be using the ScanSpeak Disco 15cm and Scanspeak HDS in TM, here is the MTM:

    Click image for larger version

Name:	IMAG0213.jpg
Views:	1
Size:	459.0 KB
ID:	1158138

    I'll also be doing versions of the SB15MFC with Scanspeak HDS tweeter, no pics yet though.

    New reference 3-ways for my music room. Peerless Exclusive 8" woofer (I'll also be doing a version with the SB 8"), Transducer Labs tweeter, and a mid to be selected after auditioning: either the TangBand 3" dome or BG Neo8:

    Click image for larger version

Name:	IMAG0214.jpg
Views:	1
Size:	344.3 KB
ID:	1158139

    And pending a writeup my BoseBuster version 2 are done finally:

    Click image for larger version

Name:	IMAG0215.jpg
Views:	1
Size:	330.6 KB
ID:	1158140

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    The W20, probably the 8 ohm version, though I don't remember for sure. I went with passive radiators for a tiny box (12" cube). I'll be posting that info with the BoseBuster update within the month.

    Leave a comment:


  • nikkoluvr
    replied
    Re: listening impressions of some interesting drivers...

    Hey Brandon, what Silver Flute driver are you using for the 8" sub?
    thanks,
    nikkoluvr

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    happy happy joy joy

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Transducer Labs tweeters are on the way! Will be interesting to see how they compare to the AC AST2560 and Scan HDS (just replaced my Peerless HDS with the Scans on the off chance there has been a modification improvement). Hope they are a real step up and end up in my 3-way using the TangBand 3" dome mid.

    And the final designs that resulted from this thread are being fabbed as we speak, here are about half the baffles:



    Several flavors depending on what my friends decide to do, Silverflute woofer & SB26STCN in MTM, SBA poly cone & SB26STCN in TM and MTM, Scan Disco & HDS in TM and MTM, and possibly SBA poly cone & HDS in TM and/or MTM.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    Had a chance to play with the Aurum Cantus AST2560 air motion tweeter. I took the felt off the back, built small enclosure behind maybe 2" deep, and filled it with that denim stuffing from HD. Here is the response on 8" x 14" baffle from the Dayton .38cf cabinet that I've been using:



    The dip at 3khz and peak at 4-5khz is baffle diffraction and will be easily corrected on a baffle for a final design. I've only been able to compare to the Peerless HDS, both crossing 2.5khz LR4 and matched to the Scan Disco woofer I've been using (and growing very fond of). These are very well behaved tweeters. No weird resonances, no sibilance or splashiness, just very calm. Decently open sounding. I think it gives up a bit of musicality *at times* to the HDS, but only in the areas where the HDS excels past everything I've heard anyway: bells, plucked strings, *some* female vocals and piano. But everywhere the HDS falls short, the AST2560 is just playing along calmly. Integrates well with the woofer and doesn't bring attention to itself. I would say better behaved than maybe even the RS28A, but more open and dry sounding.

    Played with the Silverflute 14cm to pair with the SB26STCN and I don't think there is a better sounding woofer for <$50. Really good, just open, smooth, and with good bass. Yeah high frequency non-linear distortion ultimately limits it, but not easily, and not as much as you'd think versus much higher priced woofers like the SB 15NRX. Definitely going to have a cheap design coming with this woofer for friends dipping their toes into good audio.

    Using the SB SB26STCN tweeter with the Fountek FR88 in my Bosebuster sequel is a done deal. Love it. You would think these would only be good at easy listening music while hard rock or metal would bring them to their knees, but playing some Tool or NIN really brings these little buggers alive!! I have them crossed to Silverflute 8" subs that impress the hell out of me for 12" cubes. I'll finish the documentation and publish that if anyone is interested.

    I tried the poly cone SB15MFC for the heck of it to see how it compares to the paper version I already reviewed above. I like it better than the paper. It is a bit more liquid and fuller sounding, and before you write that off as detail killing warmth, I think it actually has a slightly more detailed midrange than the paper. OTOH the paper is just a bit better behaved on the high notes, but the presentation is just flatter and less real. Now these are aren't huge differences, I'm playing music for hours while switching back and forth via minidsp, so it is easier to the tell the difference. I had pretty much decided to go with the Scan Disco over the paper SB15NRX for musicality, but with this poly SB having some of those same strengths - maybe 1/3 of the way from the paper NRX to the Scan Disco- it makes the choice much harder since with the SB I'll also gain stronger and more defined bass than the Disco. More investigation needed there.

    Only had about 30 minutes to listen to the B&C 8CXN51 coaxial driver. I love these pro drivers, just great percussion and liveliness. The tweeter seems a lot like other B&C polyamide CD's, which is to say at or near the top of the heap. But still not as well behaved or smooth as a really good hifi dome. I think the sweet spot might be 8" pro woofers matched with good hifi domes on waveguides. But still haven't spent enough time to see where to cross, or where to smooth the response so this is still very much a work in progress and I have high hopes.
    Last edited by augerpro; 04-12-2014, 03:56 AM.

    Leave a comment:


  • augerpro
    replied
    Re: listening impressions of some interesting drivers...

    I'm not real big fan of ring radiators

    Leave a comment:

Working...
X