3ll3d00d Posted April 13, 2016 Share Posted April 13, 2016 I offset them a bit to separate them, the measurements were taken at different times and at different levels so were not comparable in absolute spl terms at all. My point was just that the resolution available from 3ms tends to have an effect a fair way beyond 2x the resolution so IME it is worth getting outside. If you look closely then you'll see the knee is sharper on the outdoor one and the 1-2kHz (if not 1-6k) range is generally flatter/smoother. Of course it does ultimately depend on the size of the room (and the magnitude of the reflections after that 1st reflection), perhaps your room is just more amenable to measurement than mine. I've also seen the baffle impact directivity as you'd expect, in my case it was centred on 800Hz or so but obviously that depends on the size of the baffle. Quote Link to comment Share on other sites More sharing options...
SME Posted April 14, 2016 Author Share Posted April 14, 2016 OK, that makes way more sense! The impulse responses are very interesting to me. I'm not talking about the part after 2.5 ms or so, where I expect to see some reflections. The significant difference in impulse response, from roughly 0.5 to 1.0 ms is interesting to me. Sound moves roughly 1 foot in a millisecond (ms), so whatever is causing the indoor measurement to differ from the outdoor one is very near the horn or the mic. Any thoughts as to the difference here? Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted April 14, 2016 Share Posted April 14, 2016 The significant difference in impulse response, from roughly 0.5 to 1.0 ms is interesting to me. Sound moves roughly 1 foot in a millisecond (ms), so whatever is causing the indoor measurement to differ from the outdoor one is very near the horn or the mic. Any thoughts as to the difference here?It is odd and I can't explain it. The mic stand, speaker and speaker stand are the same in each measurement so the differences are all in the surrounding environment. Clearly I am missing something but I don't know what. Quote Link to comment Share on other sites More sharing options...
SME Posted April 27, 2016 Author Share Posted April 27, 2016 Just a quick update. Progress has been slow due to lots of other real life stuff getting in the way. I have been doing some more pre-design work on the subwoofer subsystem. I hope to be able to order subwoofer drivers very soon. Last weekend, I attempted to do polars of the SEOS-15 with the alternate compression driver, but I couldn't pull it off using the compass method at all. The magnet was just too strong even with the phone something like 3 feet away. Worse still, the magnetic field seemed to cause the phone's magnetic directional sensor to progressively lose accuracy as the phone was rotated. It was only after I got to about 45 degrees that I realized the compass was off. As such, I'm very suspicious that this problem may have fouled up the horizontal polars I posted for the DNA-360 + SEOS-15. As such, I am going to re-measure it using a different method that I hope will be much more precise. Hopefully I will be able to do that this weekend. Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 3, 2016 Share Posted May 3, 2016 FWIW I picked up some SEOS12 recently and measured outside with a 8.5ms window, graphs (0-90 degrees at 7.5 degree intervals) are in http://www.avsforum.com/forum/155-diy-speakers-subs/2188265-attempting-3way-seos10-4.html#post43753986 I think you can see the sharper knee I was talking about before if you compare them Quote Link to comment Share on other sites More sharing options...
SME Posted May 5, 2016 Author Share Posted May 5, 2016 (edited) Excellent, thanks for the update! Edit: It looks like your BMS driver may be exhibiting diaphragm break-up around 17 kHz just like the DNA-360 as evidenced by the considerable beaming there. I used to think this kind of feature was due to a mismatch between the CD exit and horn throat, but better measurements indicate this is break-up, at least for the DNA-360. I'm curious about how much output you can get before that break-up goes non-linear. If you are up to it, you might try doing the sine-sweep tests I did. Obviously, care should be taken when doing the higher level sweeps. I looked for evidence of power compression or excessive distortion after each sweep before going forward, just so I didn't damage anything. I finally got a bunch more measurements done including a full redo of my DNA-360 / SEOS-15 measurements. My hunch was correct in that the compass app on the phone did not give me the most accurate angles. As the horn was rotated, the error in the reported angle from the compass app accumulated. Here's before: Here's after: Both of these are normalized to 30 degrees instead of 37.5 degrees (as was in my original post) because 30 degrees looks better for the MLP now that I have good data. Note that the horn horizontal dispersion appears to be 90 degrees instead of 120 degrees, as it should be. After doing the new measurements, I discovered an unwanted filter in-line, so I'm only updating the normalized maps until I have time to apply the inverse filter to all the data I just collected. Whoops. I also did some measurements of the woofer in its box, which are posted here. While I was at it, I also did a set of polars at 10 degrees elevation. I am hoping that I can assume that the response is approximately separable, i.e.: P(theta, phi) = P(theta)*P(phi) for pressure P and azimuth and elevation angles theta and phi. In terms of SPL, that can be written as SPL(theta,phi) = SPL(theta) + SPL(phi) + C. where C is some constant. If the response is separable, then I can estimate the polar response anywhere on the unit-sphere using only the data along each axis. I'll use the elevation 10 degrees data to try to verify this assumption. If this works out, I can use the data I have to model the response at each listening location with any particular crossover as well as to calculate the power response and true directivity with that crossover. Edited May 5, 2016 by SME Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 5, 2016 Share Posted May 5, 2016 Edit: It looks like your BMS driver may be exhibiting diaphragm break-up around 17 kHz just like the DNA-360 as evidenced by the considerable beaming there. I used to think this kind of feature was due to a mismatch between the CD exit and horn throat, but better measurements indicate this is break-up, at least for the DNA-360. I'm curious about how much output you can get before that break-up goes non-linear. If you are up to it, you might try doing the sine-sweep tests I did. Obviously, care should be taken when doing the higher level sweeps. I looked for evidence of power compression or excessive distortion after each sweep before going forward, just so I didn't damage anything. that was some compression sweeps going from 80dB right? I'll give it a go next time I measure in room, I think that sweeping at that level gets me an ASBO if it do it in my garden Quote Link to comment Share on other sites More sharing options...
SME Posted May 6, 2016 Author Share Posted May 6, 2016 Yeah, I think you'll definitely want to do them indoors and with hearing protection. I also suggest doing it with the crossover in place, if possible. You want to see how the driver handles the highest frequencies when the typical top end roll-off is compensated for, and you don't want your sweeps to be limited by excursion on the low end. Definitely do some sanity checks after each sweep to ensure there are no visible signs of power compression and that THD remains reasonable for the lower frequencies. The "80 dB" refers to the SPL at the MLP (at 9 feet or so from the speaker) that my speakers reproduce for the "-20 dBFS" sweep in those measurements, if that makes sense. My "-2 dBFS" sweep was ~98 dB at the MLP, just as a -2 dBFS treble signal in a soundtrack played at MV "0" would be 98 dB at the MLP. The horns are toed in, so I'd expect the MLP response is maybe 1.5-2 dB below the on-axis response. Today, I captured the filter response so I can correct the data I have and post non-normalized results once I get around to doing this work. The changes have no impact on the normalized plots. I also attempted to assess IMD distortion with two-tone stimuli. One of the two CDs (not the DNA-360) has a shorting ring and notably lower inductance, so I thought I could reveal an advantage for it by doing these tests. I put on hearing protection, locked the cats in the bedroom (why they think they needed to be in the room for this is another question) and ran dual-tones at levels exceeding 100 dB at the MLP. For the lower tone, I chose a frequency at the low end of the range of each driver. The DNA-360 got 800 Hz, which is slightly below its recommended crossover point. The other driver, which doesn't play as low, got tested with 1100 Hz. Casual inspection suggests that the lower inductance CD with the shorting ring does indeed have lower IMD when the higher tone is > 10 kHz or so but it's very hard to quantify the differences using the capabilities that REW provides. Even with the overall response being dominated by direct response energy, there is always the chance that one or both chosen tones will fall in a narrow peak or null, which completely throws off the numbers. I think to do this "right" requires a series of sine-sweeps simultaneous to a low frequency stimulus. The stimulus would be increased in level, but not necessarily the sine sweep as most of the IMD seems to be caused by the low frequency tone, even though it only manifests with high frequency tones playing simultaneously. I'm not sure of how the results should be analyzed either. So, this will require software tools that I don't have right now, and since I'm rather eager to settle on my choice of driver, I'm thinking I probably won't worry about it, even if it is potentially significant. I'd like to do some more simulations with the data I just got before making a final decision, but right now the DNA-360 is my favorite because of its excellent low frequency performance. The DNA-360 has a recommended XO of 950 Hz vs. 1.2 kHz for the other driver but after my distortion measurements (not posted), I'd be comfortable using the DNA-360 even lower, whereas 1.2 kHz looks a bit too low for the other CD without a very steep slope. Likewise, seeing how much the TD12M woofer starts to beam above 1.4 kHz or so suggests that the DNA-360 will work much better in a crossover with a wider transition region, which is something I'm shooting for. The SEOS-15 itself appears to hold its horizontal pattern all the way down to 600 Hz in its box, and I'd love to be able to take advantage of that control given that I am designing for multiple listeners. My polar response measurements indicate that the change of driver has a negligible impact on polar response, except in the break-up region, which is likely to get EQed way down anyway. Even then, the differences are fairly minor. Both of the drivers I'm testing break-up at around 17 kHz, and your measurements suggest that the BMS (4550?) driver has a break-up right around that point as well. I would guess the particular frequency involved has mainly to do with diaphragm size and material choice. Incidentally, both the drivers I have use polyimide diaphragms vs. polyester for the BMS 4550. In any case, the advantage to having "extended response to 20 kHz" is mostly ruined by the presence of break-up, IMO. If you toe-in to 22.5 degrees or more and EQ the response including the top end to be smooth, listeners sitting outside the MLP and directly in front of a speaker will get a peak there. Furthermore, the power response will likely be uneven and possibly accentuated in that region. IMO, not nearly enough attention is given to top end response in a typical speaker. Many people test their hearing with sine waves and conclude things like "I can't hear above 15 kHz" or some such. For starters, hearing sensitivity drops off quickly above there, so you may be able to hear it if you used a higher playback SPL. More importantly, sine waves may not be nearly as audible as broadband content there, including transients. Too much output above 15 kHz adds an ear-pinching sensation and makes the sound brittle and inexplicably loud. I've used the adjective "glaring" before. Like it's irritating to your ears even though it doesn't sound loud. Another analogy is finger nails on a chalk board, which probably does involve a lot of top octave sound in real life. Any problems in response here tend to be very dependent on source material. A lot of content has nothing above 15 kHz, but some stuff has a huge amount of energy up there. I'm fairly certain I can hear the 17 kHz peaks in my current CDs at times. With most stuff, it adds just a bit of crispness and sparkle to the sound, which while not necessarily accurate, is at least pleasing. However, I do have some stuff that's not so pleasing to listen to. Some of the worst offenders IME are a few hi-res remasters where the 15 kHz+ stuff seems to have been pumped way up, presumably to satisfy the customer who expects to hear very high frequencies in digital content mastered with a higher sampling rate. I have a remaster of "Miles Davis - Sketches of Spain" with muted trumpet that is practically un-listenable in my current setup. Even standing in the kitchen (outside direct line of sight from the speakers), the trumpets are irritating at even fairly moderate levels. Thankfully, this is the exception more than the rule. Otherwise, differences in speakers used in mastering probably account for a lot of variability here. I suspect the ears are concerned with power response at least as much as direct response in this region, so it wouldn't be surprising if content mastered on domes that beam in a relatively dry room tends to sound very bright on a system with horns and a lot of high frequency ambiance. The next thing I want to do is try to see if I can EQ down the break-up peak in the DNA-360 using a fairly high Q notch filter and if the new response is more or less forgiving with high SPL sweeps. If it works out, then I can hopefully retain some output above the peak. Otherwise, I'll probably just have to roll things off a bit more aggressively starting at 15 kHz or so, which kind of defeats the point of having a CD that can extend all the way to 20 kHz in the first place. Doing this is hardly unprecedented of course. In fact, I believe most examples I've seen of speaker response voicing involve top-end roll-off that starts well below 15 kHz and becomes quite substantial (6-9 dB or more) above that point. Maybe some day, I can try out a Be diaphragm driver I can actually afford, like a Radian, to see if it's better behaved, but knowing that the Radian's don't have the best reputation for low frequency performance kind of dampens my interest in them for this design. For now, I want to wrap up this design so I can start building a center channel speaker. Voices sound so amazingly good with these speakers, and I can't wait to be able to use one for dialog in movies. Quote Link to comment Share on other sites More sharing options...
SME Posted May 9, 2016 Author Share Posted May 9, 2016 Today, I iterated on some PEQ filters to squelch the 17 kHz break-up peaks on each CD. I ended up using 3 PEQs on each as I couldn't suppress the peaks with a single filter without doing harm to the adjacent regions, and I wanted the roll-off to be a bit lower Q to avoid ringing. My long term fix will probably use FIRs to a nice low Q target curve. Although the improvement was subtle, the sound is definitely better, most notably on transients, which sound more natural now. The on-axis IR (measured with sine-sweep) looks way better as the break-ups contributed quite a bit of ringing within the first 1 ms or so. I think my cat appreciates the change too, and knowing that her hearing is a lot better up there, I trust her judgement. The sound also seemed a bit less bright overall, so I added another 1 dB to the high frequency shelf. I'll have to listen to a variety of content before I decide whether to leave it there, but I think that makes the treble dead flat out to 15 kHz or so. The hi-res "Miles Davis - Sketches of Spain" album is still un-listenable. I also have a copy in 128 kbps MP3 format, presumably ripped from the CD, which sounds way better (in more than one way, in fact) despite the MP3 artifacts, so it must be the master that's just royally screwed up. In comparison even the woodwinds sound alien in the remastered version due to way too much top-end energy. I'm guessing they hyped it up to appease all the audiophools who obsess about inaudible frequencies to make them think they were getting something "better". Yeesh, "hi-res" sound is such a racket. I just wish I could get my money back. I did the linearity measurements on the DNA-360 again, and everything looks great up to -5 dBFS @ MLP. At -2 dBFS, there is misbehavior, but only above 18.5 kHz. A slight increase in roll-off should help with that, but I'm not sure it's important right now. The only caveat here (to repeat what I wrote in an earlier post) is that the misbehavior is non-linear, so it's possible it may manifest more readily with transients than with sine sweeps. However, I'm leaning toward the opposite being true; i.e., that transients will be less susceptible to trouble than slow sine sweeps. I won't know for sure without more sophisticated measurement tools, but at this point, I'm confident enough that I am committed to going with the DNA-360 drivers. They are superb drivers, especially with regard to low frequency performance. My measurements suggest they should work great with a gentle crossover (2nd order electrical < 1000 Hz) and will let me take full advantage of the SEOS-15 horn. I'm looking to orient them at 7.5 degrees vertical from the MLP as that seems to be a sweet spot, outside the on-axis null > 13 kHz (excepting the break-up peak), where the directivity in the top octave is most smooth. It's also a good compromise for good response for the other seated positions and for standing listeners. Quote Link to comment Share on other sites More sharing options...
SME Posted May 11, 2016 Author Share Posted May 11, 2016 Been listening a good bit after the changes to suppress the break-up peaks, and there is definite improvement even if it's fairly subtle. For lack of better description, it sounds more refined. As already mentioned, the difference is most apparent on transients. I went ahead and placed an order for more SEOS-15 horns and DNA-360 drivers so I can finish these speakers and have parts I need for the center and surrounds. I did some more research on enclosure damping. When it comes to speaker enclosure damping, it's very hard to find good information. It's even harder to find products suited to the purpose, IMO. I can probably find at least 20 different products of various kinds which claim to damp vibrations in speaker enclosures but have no engineering data to support their claims. I can also find products intended to damp sheet metal; however, these products are not likely to work well with speaker enclosures which have a lot more flexural rigidity to begin with. Based on my current theoretical understanding of the engineering problem, the best candidate material I'm aware of right now to glue between sheets of 1/2" BB plywood is 1/8" thick 70 durometer Sorbothane sheet, which can be bought at a reasonable price through McMaster Carr. They also offer the stuff in thinner sheets, which I believe would damp better, but is not cost effective in the sizes available. What'd be really nice is to be able to buy 40 durometer in 0.5 mm thick sheets in a size of 12x24" or bigger. Maybe I can find something better if I look hard enough. With that said, the enclosures I built out of braced 0.5" plywood are a lot more dead than the Hsu speakers made with 3/4" MDF and no bracing. However, there's still some vibration that finds its way into the floor. I can isolate the speakers as I did with the Hsu's, but that only solves part of the problem as the enclosures themselves may still radiate unwanted sound. However, I can't readily quantify the problem with the tools I have, and there is the risk that any attempt to improve things with damping could make things worse by adding more mass than stiffness. Subjectively, I don't notice any ill-effect on the sound (vs. the Hsus which made a nasty buzzing sound when trying to reproduce enough sound at around 260 Hz or so), so I think I'll continue to ignore the issue for now. I can always revisit it later. Depending on what I decide, I may start working on a center channel enclosure tomorrow and maybe also do some experiments with angled cuts and joints, which I'll need for the new horn boxes to angle the front face down and for the surrounds. Quote Link to comment Share on other sites More sharing options...
Ricci Posted May 11, 2016 Share Posted May 11, 2016 Enclosure bracing and damping is a deep topic. On the surface it's easy. Make the walls 6" thick concrete with re-bar cross bracing every 8", no parallel walls and make the whole thing weigh 3000lbs. Add a liberal amount of poly fill strategically. Of course none of that is really practical. The trick is getting to a good compromise point of complexity/weight/cost that starts to run up against diminishing returns if pushed any further. Subwoofers are generally easy. Brace the hell out of it and push the resonances up above band. The bigger issue with them is vibration of the entire cabinet mass. Dual opposed takes care of that of course. If that isn't an option it can become an issue with powerful subs though. I've seen well braced Othorn systems bouncing off of and vibrating across the floor. Midrange is much tougher IMO. The response issues and resonances are in band and more difficult to push out of band or squelch satisfactorily. Quote Link to comment Share on other sites More sharing options...
SME Posted May 11, 2016 Author Share Posted May 11, 2016 Great comments! The issue with the Othorns is interesting. Do you think that the Mms of the woofer impacts things a lot? Not that it's the only factor, but I can imagine it being pretty significant. Quote Link to comment Share on other sites More sharing options...
Ricci Posted May 14, 2016 Share Posted May 14, 2016 Yes the moving mass of the drivers does matter. A lot of the modern big bass drivers are 400 to 600g of mass in the moving assembly. Once you start to think about the acceleration and velocity forces involved with high output from them and coupling those rigidly to a cabinet it's no surprise. A 15 with 200g mass isn't going to vibrate the entire enclosure as much as one with 400g. This can have an effect on tactile response at the LP as well. The heavier vibrations from the more massive driver can be directly transmitted through the floor to the furniture and listener. I believe this may be part of why many like downfiring drivers. The RF 19 causes a lot of reactionary force with 1Kg of mass. Good luck running it solo in a cab and not vibrating the whole assembly. It may as well be an oversized tactile transducer.I was running the first one free air sitting vertically on my floor with a channel from an SP2-12000 on it. I decided to see how much excursion it would produce before sounding stressed. The driver vibrated so hard that its 70lb mass was literally jumping off of the floor and started jack hammering itself. This is on carpet at about 2.5" peak to peak at 25Hz. Quote Link to comment Share on other sites More sharing options...
SME Posted May 14, 2016 Author Share Posted May 14, 2016 I'll just say that I've been doing a lot of listening since correcting the break-up peaks and adding 1 dB to the top, and the differences are way more than subtle. These drivers sounded very good before, but the sound quality improvement is like having upgraded the drivers again. Clearly, the excess energy from the break-up peak was masking perception of much of the rest of the top octave sound, and this happened even at off-axis seats (including the MLP) where the break-up wasn't visible. (It was visible in spectrograms as a frequency region of higher decay time.) Never in my life have I heard so much detail in music. Violins, pianos, flutes, synths, hats, and cymbals are all stunningly life-like. If someone's interested, I could post the PEQs I used for the DNA-360, but I would caution that diaphragm break-up is one of those things that might vary subtly between production samples. So if you want to tame the break-ups on your DNA-360, your best bet is probably to do your own measurements and corrections. Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 15, 2016 Share Posted May 15, 2016 you may have seen them but there are loads of lengthy threads on cabinet construction techniques on diyaudio, so many that it's hard to tell the wood from the trees if you're familiar with the mechanics of it. I don't know about you but it takes me so long (in elapsed time) to build one cabinet let alone a load of different ones that experimenting with different construction techniques just isn't really feasible I did pick up one of the accelerometers listed in http://www.libinst.com/accel.htm thoughto see if I could measure what was going on before I build the rest of my cabinets. I need to try to get that hooked up to see if I can measure something reliably (seems tricky in itself frankly but got to be worth a try). It would be interesting to see the before and after measurements btw just to see the impact on the measurement to compare against your subjective assessment. I find objective data is essential to make sense of something I can't hear myself. Quote Link to comment Share on other sites More sharing options...
SME Posted May 16, 2016 Author Share Posted May 16, 2016 you may have seen them but there are loads of lengthy threads on cabinet construction techniques on diyaudio, so many that it's hard to tell the wood from the trees if you're familiar with the mechanics of it. I don't know about you but it takes me so long (in elapsed time) to build one cabinet let alone a load of different ones that experimenting with different construction techniques just isn't really feasible I did pick up one of the accelerometers listed in http://www.libinst.com/accel.htm thoughto see if I could measure what was going on before I build the rest of my cabinets. I need to try to get that hooked up to see if I can measure something reliably (seems tricky in itself frankly but got to be worth a try). Rather than having to build an entire speaker for each test, I'm thinking I could test the suitability of damping materials by gluing them between plywood sheets cut to be of similar dimensions to the spaces between braces inside my current cabinets. Something like 6x9" pieces of something. Then I could attach an exciter at one end and an accelerometer at another. I would expect a good damper to reduce both the peakiness and the overall response level of the exciter for a given drive level. I can also do the test on single plywood sheets alone or on a pair glued together without damper. Quote Link to comment Share on other sites More sharing options...
SME Posted May 17, 2016 Author Share Posted May 17, 2016 It would be interesting to see the before and after measurements btw just to see the impact on the measurement to compare against your subjective assessment. I find objective data is essential to make sense of something I can't hear myself. Sorry, I know I'm bad about posting pictures a lot of times. It takes a while to prepare them, and I'm never certain if anyone is even interested. Is there something in particular you'd like to see data on that I haven't posted yet? For example, do you want to see before/after of the break-up peak suppression I did with PEQ? Or was this statement meant with regard to data on enclosure construction/damping methods, when I get around to doing that assessment? Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 18, 2016 Share Posted May 18, 2016 It was the suppression of the peak I was interested in. On cld, as far as I can see there are 2 practical problems. Firstly the availability of materials and hence the amount of trial and error required to find something that works well. You probably have more chance taking a short cut here as many posts on diyaudio reference material that seems US specific. Secondly going from a test piece to using it in an enclosure seems tricky. Some commonly available materials (e.g. various types of roofing felt) make it hard to cut the panel afterwards so actual box construction gets harder and more time consuming. Nothing insurmountable for sure but certainly seems like a lot of work! Any thoughts on how you will measure this? Seems like common choices are using a guitar pickup hot glued to the cabinet or an accelerometer. Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 18, 2016 Share Posted May 18, 2016 FWIW I've been doing some more reading on CLD over the last few days and it seems https://www.amazon.co.uk/Sika-Sikaflex-Polyurythane-Adhesive-Sealant/dp/B003UTA2BKseems a decent option. Objective data is somewhat thin on the ground but here's an example of such a construction anyway - http://www.diyaudio.com/forums/multi-way/218437-3-way-active-time-aligned-constrained-layer-construction.html#post3139996 Quote Link to comment Share on other sites More sharing options...
SME Posted May 18, 2016 Author Share Posted May 18, 2016 (edited) It was the suppression of the peak I was interested in. I did on-axis measurements to verify the filters were doing what I wanted, but it looks like I didn't save those measurements. If you'd like, I can re-do them and post them here. Do you want FR, IR, or both? Firstly the availability of materials and hence the amount of trial and error required to find something that works well. You probably have more chance taking a short cut here as many posts on diyaudio reference material that seems US specific. Secondly going from a test piece to using it in an enclosure seems tricky. Some commonly available materials (e.g. various types of roofing felt) make it hard to cut the panel afterwards so actual box construction gets harder and more time consuming. The hard part about material availability is that very few available materials have engineering data published for their visco-elastic properties. Most such materials are designed for damping sheet metal, which I believe tends to have a lot less flexural rigidity. There's no shortage of product sold that claims to be useful to "damp speaker enclosures", but you know, without data, such claims must be taken with a grain of salt. My idea is to use the test pieces as models for how the material will ideally behave as part of the speaker construction. Then, once a suitable CLD composite is found the damper and constraining layer can be glued directly to a fully-constructed speaker. That's the big reason I built these with 1/2" ply instead of 3/4" ply, so I could add CLD later if I wanted to. FWIW I've been doing some more reading on CLD over the last few days and it seems https://www.amazon.co.uk/Sika-Sikaflex-Polyurythane-Adhesive-Sealant/dp/B003UTA2BKseems a decent option. Objective data is somewhat thin on the ground but here's an example of such a construction anyway - http://www.diyaudio.com/forums/multi-way/218437-3-way-active-time-aligned-constrained-layer-construction.html#post3139996 Lots of materials are potentially suitable for use as dampers, but again, the trouble is getting data on the viscoelastic properties. I am using this book as a reference: Handbook of Viscoelastic Vibration Damping, written by David I.G. Jones, published by Wiley. The book is written for mechanical engineers and is quite expensive, so I don't recommend buying it unless you are comfortable with a text that's heavy on math and physics. For a good CLD system, you first need to choose a material with good internal damping properties at the temperature and frequency of interest. From those materials, one needs to choose one that has a favorable elastic modulus and use the right thickness of that material between constraining layers for best results. As far as thickness and elastic modulus are concerned there is a kind of sweet spot, dependent on the thickness and elastic moduli of the constraining layers. A relatively simple analysis of CLD systems leads to the so-called Ross, Kerwin, and Ungar (RKU) equations. These equations can provide some guidance as to how well a CLD system will damp relative to the damping of the damping layer alone. But to use the RKU equations, one must have at least some rough figures for the elastic and/or shear modulus and thickness of the constraining layers and for the damping layer. In the latter case, the moduli will actually be frequency dependent and will be complex-valued (i.e., have both magnitude and phase) to express the tendency of the material to both store and return energy elastically and to dissipate it. This is in close analogy to electrical circuits in which reactance stores energy and resistance dissipates it. Thus far, I have not found any forum threads or speaker design pages that attempt to use RKU equations to optimize the CLD system before materials are selected, leading to a better chance of getting good results with less effort. The thing is, it's not enough to choose a material with good damping properties. The thickness and flexural rigidity of the material must also be matched well to the rest of the enclosure or damping may be poor. The consequence of getting this wrong would be an unnecessarily heavy enclosure that may actually perform worse than one without the damping layer. Edit: It occurs to me that with the construction method I propose, the outside constraining layer won't be as stiff as the inside unless it's bolted through to the inner plywood where the braces are attached. That may not make a huge difference but it's something I'll have to keep in mind. Edited May 18, 2016 by SME Quote Link to comment Share on other sites More sharing options...
SME Posted May 19, 2016 Author Share Posted May 19, 2016 Woohoo! The two additional DNA-360s I ordered have shipped and should arrive on Saturday. I can finally stop listening with mismatched drivers. And, I can try experimenting with a lower crossover. My long term goal is to have a fairly broad region of overlap between the drivers, but this may require more sophisticated filters than I'm using currently. Either way, I'm going to try to get the center channel build going ASAP. That'll be a huge win even with the crossovers being relatively unoptimized. As it is, I can't praise the sound of these speakers enough. I know I need to get out into the world to see some of what others (particularly on this forum) have done, but right now, these are the best speakers I've ever heard. Nothing else I've heard comes even close. The woofers are particularly stunning. The horns and CDs do not disappoint either, but coming from older speakers with controlled directivity horns, I had a pretty good idea of what to expect. While there are different strokes for different folks, I do have to wonder why horns don't already rule the world of speakers. Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 19, 2016 Share Posted May 19, 2016 I have not come across such equations either. I don't think this is surprising though, for example you don't see much discussion of exactly how a motor is constructed/designed either. Ultimately the universe of materials that are readily available, reasonably priced, easy enough to work with without specialised equipment and suitable for speaker construction seems quite small and the construction method seems as important anyway. I believe the point is made that a panel is a much less complex system compared to a box.There are some interesting comments from Earl Geddes on diya on the subject, he advocates cld on the baffle and rear panel as well as the braces which are attached to implement a skyhook damping scheme (details in http://www.diyaudio.com/forums/multi-way/276721-best-cabinet-material.html albeit that thread goes all over the shop). There are some other adhesives mentioned in http://www.diyaudio.com/forums/multi-way/284239-seos-dayton-2-way-3.html#post4718307that might be useful too. There is an interesting article here on building a pre amp for an accelerometer at http://www.audioxpress.com/article/Test-your-speakers-performance-with-this-do-it-yourself-measurement-systembtw (you can buy one from that guy in Greece if you like).Be interesting to see what you come up with anyway Quote Link to comment Share on other sites More sharing options...
SME Posted May 20, 2016 Author Share Posted May 20, 2016 You are absolutely correct that a panel is a much less complex system than a box. However, boxes do have panels that can bend very similarly to panels in isolation. The RKU equations only actually treat panels in free space, but approximations are available for estimating performance of panels with different boundary conditions such as those that are essentially clamped at their ends to the braces. The thing about a box though is that it can exhibit more complex motions involving the whole structure. However, my (admittedly uneducated) guess is that a good bracing system will suppress these higher order motions very well, and that bending of the unbraced parts of the outer panels will dominate the vibration of the box. Although only 1/2" thick, I have a pair of shelf-style braces to tie together the front, rear, and side panels as well as a pair of braces running top to bottom along the rear and across the top and bottom and connecting to the doubled front panel. Most of the joints use loose hardwood tenons, which may help a lot, or not. I'm sure they would make it much harder to crush one of the boxes, but do resist vibratory motion along the joints better than glue alone? I have no idea. That thread was a rough read with so much stuff that made me shake my head. It's a tough subject for sure. I found Geddes's description of his skyhook damping scheme to be pretty vague, but it looks like there is a short blurb on Skyhook Theory on wikipedia: https://en.wikipedia.org/wiki/Skyhook_theory I don't expect I'll ever get that fancy. Geddes himself confessed that it's very hard to assess how audible something like cabinet vibration is. I agree, at least, in so far as my speakers are concerned. If I sine sweep them, I can feel them vibrate throughout the bass and mid-range by touching cabinets, and I can feel that vibration in my feet through the floor too. However, they sound quite clean during that sweep whether I'm standing right next to them or at a distance. I also don't notice any obvious vibrational response peaks by touch, which gives some assurance that the cabinet damping is at least decent to begin with. In my listening so far, I have not heard anything that I could readily ascribe to cabinet vibration resonance. That doesn't mean I won't eventually notice something. For example, audible issues that are there might be currently masked by decay problems in the room. For contrast, the Hsu HC-1 and HB-1 speakers these are replacing have cabinet resonance problems that are definitely audible. The worst area is around 260-270 Hz with both models. After becoming acquainted, I can clearly hear it in sine sweeps. I can also feel the peak in the vibration response there. Unfortunately, the peak in vibration also tends to trigger a nasty buzzing sound, possibly due to the woofer baskets vibrating independently from the pieces of plastic used as decorative trim in front of the woofers. Anything with a strong enough fundamental in that region could trigger it. If you can't hear anything wrong, how do you know that there isn't something that you (or someone else) will hear later? My approach is to worry more about stuff I know I can hear and come back to it later if need be. Even though that means I won't be doing a proper finish for these any time soon. Actually I think they turned out pretty nice with just sanding, which was a nice benefit to using the BB Plywood. Quote Link to comment Share on other sites More sharing options...
3ll3d00d Posted May 20, 2016 Share Posted May 20, 2016 completely agree that the proof is in the pudding, if there are no audible artefacts during (normal or spirited) playback sessions then all is good as far as I'm concerned (and if not, will cross that bridge when I get there!). I actually have a spare TD12 though so I keep that and use it for cabinet build experimentation purposes (though I certainly have more ideas about what I could do than time to do it so that one might be on the back burner for some time). Quote Link to comment Share on other sites More sharing options...
SME Posted May 25, 2016 Author Share Posted May 25, 2016 A few points to cover here: Study of vibration isolation While listening to some music (I forget what), I was following a bass or guitar or something that was playing with the fundamentals hitting pretty strong. I noticed a bunch of notes in the upper bass where the imaging collapsed and the sound got muddy and weird. I decided to try doing higher level sine sweeps to find problem areas for the speakers. I tried putting my hands on the speakers, and I did notice some peaky areas, but nothing dramatic. Instead, my attention was grabbed by a buzzing sound at around 165 Hz. From the listening position, the buzzing was very clearly heard. However, I also noticed a big response dip there. (It's visible in measurements taken at my seat.) I approached the speaker and tried to find the buzzing, but it didn't seem to be the speaker. It was also very loud near the speaker with > 80 dB @ MLP sine waves. I never figured out what was buzzing, but I think it was my racks. I couldn't confirm it though, possibly because the tone was just too damn loud in the vicinity of the racks. At the MLP, however, the buzzing was easy to hear because of the response dip there. In any case, I decided to try mechanically isolating the speakers from the tables they are on. I used the same Sorbothane squares I used under each corner to isolate my previous speakers from the tables. I decided that the act of installing these would make for a great experiment to see how much sound radiating from surfaces *other than* the speakers contributed to the sound at the MLP. These are un-smoothed measurements of the raw woofer response at the MLP, before and after, without touching the mic and tripod: The caveat is that the speakers are about 1/4" higher than they were before the sorbothane. We can see here that the changes are negligibly small. Note that you have to ignore anything in the very narrow nulls because those are ultra sensitive to changes in the room environment. Unfortunately, I stilled heard the buzz during the sine sweep. Did the isolation work? I played a sine at 165 Hz and approached the speaker. It felt like the isolation hadn't worked at all. The floor was vibrating like crazy! So I played tones at a few other frequencies, and noticed that the isolation seemed to be doing just fine for those. Indeed, while this is a subjective impression, I would say the isolation killed at least 20 dB from the vibration level at *most* frequencies. But 165 Hz was still trouble. Then I noticed that the sound level was actually even higher behind the speaker that wasn't playing, and the vibration there was no better. Then I put it all together. The problem is entirely acoustic. The dip in the response around 165 Hz at the listening position is most likely due to a series of primary and secondary reflections involving the floor and ceiling, and these reflections are creating a deep dip that's broad in both frequency response and space (as confirmed by moving around the space). In any case if you ever observe this kind of thing in a room, you can be assured that there will be hot spots. And indeed, the bass was really damn loud in the front of the room. So I think the vibration I was feeling in the floor and the buzzing of (presumably) the racks is simply due to very high SPL. So the take-away here is that I need a lot of bass absorption on my ceiling. The rack itself has metal panels with fairly high Q resonances. If it's a problem, I can try sticking some damping sheet on them. Thankfully, there are readily available products designed for damping sheet metal. New DNA-360 for matched drivers in horns In other news, I got two more DNA-360s this weekend. I replaced the alternate driver with the DNA-360 and proceeded to copy the EQ settings from the other driver. Unfortunately, that didn't sound very good, so I measured at the MLP and was dismayed to see what appeared to be a 1 dB drop in sensitivity with more loss in the top octave. I iteratively EQed until the new driver was closer to matching the other at the MLP. It still didn't sound very good. There was almost no fine-grained detail, and the treble sounded somehow fragile. I did not expect to have "more trouble" integrating a driver of the same kind with another. On a whim, I decreased the very top end of the new driver by 1 dB or so, and the improvement was marked, even though this made its response 1 dB less at the MLP. To make sense of this and to verify that the break-up PEQs were working well (can't really see this by measuring at the MLP), I decided to try measuring on-axis. This gave me a response that was actually much closer to the other driver. In fact, I'm fairly certain that there's just enough deviation in horizontal angle between the two speakers to cause this inconsistency. It doesn't help that the speakers have a lot of toe-in, so the MLP sits where the horizontal pattern starts to drop-off for most of the frequency range. (This is by design to optimize multi-seat performance.) Eventually I'll get those angles adjusted more precisely. In any case, I have things sounding good again, except the top octave, which is still better than before I fixed the break-up resonances. I'm not going to fret about it now as I expect it will be hard to perfect this response until I can get measurements from more listening positions. Alternatively, I might be able to work out a decent EQ using simulations from the polar data I have. One thing that is abundantly clear to me is that, at least in my relatively diffuse space, energy response trumps on-axis response at the highest frequencies. The 17 kHz break-up resonances that were 6 dB higher on-axis than at the MLP were just as harmful to the sound at the MLP as in the far seats. With two DNA-360s I opted to drop the crossover to 950 Hz. The DNA-360s still have 2nd order slopes put on them. I did a round of linearity tests on the new driver to look for driver-to-driver inconsistency in break-up behavior. For the most part, it appeared to behave identically. I was pleased to see distortion was still very low for the "-2 dBFS" sweep, representing close to WCS for movie playback, at least. This thing is almost a mid-range driver. The lower crossover with mild slope helped the in-room response a lot in the 600-900 Hz range where there are some room acoustic problems. That provided a nice improvement in the sound. I should post some MLP response measurements, but I keep making minor tweaks and then forgetting to re-measure. I'll get something up sooner or later. FWIW when using the oh-so-stereotypical 1/6-octave smoothing, I'm about +/-1 dB from 5-13 kHz. +/- 2 dB from 2-5 kHz, and +/- 3 dB from 250-2000 Hz. That nasty dip at 165 Hz is the worst thing going on at this point. What's real awesome, of course, is that the tonal balance is very consistent throughout the room and the house. I can hear the power response hole at at the crossover, but it's not a priority for me to investigate a solution to that right now. DNA-360 passive circuit I also designed what I hope will be a correctly functioning passive circuit for the DNA-360 to protect it and hopefully kill most of the hiss. I expect the parts on Thursday, and will likely play with that over the weekend. If it works I can post the design. I've never designed a circuit before, so I could have fouled it up completely. I wrote some code to do the calculations I thought I needed done. It let me estimate things like the on-axis frequency response (using measured data as input), the impedance response (also using measured data), and power dissipation in each component as a function of frequency. I was surprised by how quickly the script went together, so I'll soon if it is doing that calculating thing right. Center Channel woofer cabinet Some sawdust was made tonight. Some more will be made tomorrow. I'll try to put up some pictures. The design is identical to the left and right. I haven't finalized the new horn box design with slanted faces. I need to write the code to simulate responses at my listening positions and measure the precise locations of the seats within the room to use as input data. Then, using the polar response data, I can simulate different speaker placements and angles at the seats, and optimize the placement and orientation of the horns accordingly. Am I forgetting something? Yeah, I've been busy. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.