Announcement

Collapse
No announcement yet.

DIY LCR passive sound bar

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

  • #46
    Another choice then:

    I've re-worked BOBinGA's CX120-8 coaxial xo using his original on baffle measurements to raise the xo frequency so the tweeter isn't stressed out so badly.

    Here's what the FR looks like with his original 'flat' xo using XSim and setting the relative acoustic centers/delay to 0. It should and does match his original summed FR almost perfectly:

    Click image for larger version

Name:	Coaxial original Summed FR.jpg
Views:	302
Size:	121.5 KB
ID:	1485467

    Now the key here is that I take this original summed FR and use it as the target response for the new xo. It wasn't actually quite as hard as I thought it might be:

    Click image for larger version

Name:	Coaxial Summed FR.jpg
Views:	269
Size:	117.3 KB
ID:	1485468

    The original summed response/target response is in yellow. The reverse null is in grey. The xo frequency is now up at about 4800Hz with the tweeter Fs at 2800Hz, down about 30dB from the fundamental (instead of the original 3dB) and the summed FR matches the original almost perfectly.

    And the impedance response:

    Click image for larger version

Name:	Coaxial Summed Z.jpg
Views:	275
Size:	121.5 KB
ID:	1485469

    And the new xo:

    Click image for larger version

Name:	Coaxial xo schematic.jpg
Views:	257
Size:	43.7 KB
ID:	1485470

    The downside would be just over a doubling in xo parts required. The tweeter needed to go to 2nd order electrical and I needed to add in a tanking filter aimed at the tweeter Fs to bring it sufficiently down in SPL. But instead of a 3 component resonance compensating notch filter to do the latter, I stole a little trick from Jeff B and made double duty out of L3 by just adding the cap, C4, onto it with a resistor value basically of 0.

    For the woofer, the only way to get more SPL out of it higher up in frequency was to lower the value of L1 but then that forces you to use a notch filter to tame the peak that now exists centered at about 1500Hz. But it should be noted that besides lowering many of the original component values (so therefore just a little less expensive), most of the new values are also on the small side and so also inexpensive. Anyways, I would expect it to sound pretty much exactly the same as the original but better with less strain/distortion/harshness out of the tweeter, especially when pushed hard.

    So maybe this keeps it in the running, if high SPL isn't too much of a priority. Although for this soundbar application, I'd probably re-work the xo again to reduce the baffle step compensation and increase sensitivity somewhat.

    Comment


    • eavo44
      eavo44 commented
      Editing a comment
      Mobius…thanks for investigating this. I would have commented earlier but I had to watch a few tutorials on how to read these charts. I have a better understanding now but still a bit lost. Is the yellow line in the second chart representing the cx-120 with the new xo right? The response does seem quite similar to the original xo reading so that’s where I get a bit confused. Both seem to be fairly flat which is a good thing…just learned that…lol. Need a little more experience interpreting these charts I suppose. I think the Bobinga design will be the safest design at the moment for the bar. My preferred setup would be a tm, mtm, tm with a 5-5.5” driver for a little more output while keeping the height of the bar at 7” max. A few good options at the moment though.

    • mobius
      mobius commented
      Editing a comment
      eavo44, yes, yellow is BOBinGA's original response, which we know sounds good. But it looks to me like it is asking the tweeter to do too much work. So what I'm trying to do here is to design a new xo version that gives the same summed response as the original (yellow) because we know that it sounds fairly good, but does it while asking less from the tweeter (red) and more from the woofer (blue). So same overall summed response more or less, but the xo frequency has moved up and the output of the tweeter has been considerably reduced around its resonance frequency of about 2800Hz.

      To put it another way, you could say that the overall quantity of the speaker output (the summed FR) has stayed just about the same, but the overall quality should be improved, hopefully with less distortion, resonance and/or compression at certain frequencies.

  • #47
    What is the Fs magnitude of said tweeter? It might not be an issue if well damped and enough attenuated.
    Wolf
    "Wolf, you shall now be known as "King of the Zip ties." -Pete00t
    "Wolf and speakers equivalent to Picasso and 'Blue'" -dantheman
    "He is a true ambassador for this forum and speaker DIY in general." -Ed Froste
    "We're all in this together, so keep your stick on the ice!" - Red Green aka Steve Smith

    *InDIYana event website*

    Photobucket pages:
    https://app.photobucket.com/u/wolf_teeth_speaker

    My blog/writeups/thoughts here at PE:
    http://techtalk.parts-express.com/blog.php?u=4102

    Comment


    • #48
      Wolf, Fs for the tweeter is 2745Hz and in the original xo, pictured above, it is only down about 3dB from the fundamental (1st order filter only). Again, minimum recommended xo frequency by Dayton is 4500Hz.

      Comment


      • Wolf
        Wolf commented
        Editing a comment
        How high is the Fs magnitude? You didn't answer my question. A max of how many ohms at Fs? If it's a 4 ohm tweeter, and has a max at Fs of 8 ohms, then it likely won't matter much and likely still sound okay. The one octave rule is just a ROT, not gospel.

    • #49
      Here's a family of designs with an 8 ohm woofer. Note the first design, the TM, is designed for optional sideways use and has a reduced BSC option. Ported or sealed.

      http://zaphaudio.com/ZA5/

      Click image for larger version  Name:	za5.2.jpg Views:	0 Size:	26.6 KB ID:	1485525
      Last edited by djg; 04-21-2022, 10:40 PM.

      Comment


      • #50
        Some nice 8 ohm designs.

        https://www.divine-audio.com/ochocinco/

        Comment


        • #51
          Another 8 ohm TM, with the Dayton RS150.

          https://www.rob-elder.com/LouC/speak...la/CKilla.html

          Comment


          • eavo44
            eavo44 commented
            Editing a comment
            Thanks for these links djg. I looked through them and like the compact box on these CKillas. Wish the CKillas had an mtm design to go with it. The Zachaudio design also looks good cause I could create the tm, mtm, tm layout that seems promising…but I’m not a fan of the silver mids though. I appreciate your input…great to see what all the options are.

        • #52
          And here's a PE boombox using the CX120-8 with another XO design. (in the manual pdf).

          https://www.parts-express.com/Execut...r-Kit-300-7164

          Comment


          • #53
            The C-Killa's need about 8-8.5L net per driver for ported. They'll hit about 103dB at 1m with 40W of power with a 60Hz HP xo to your subs on your AVR. F3 is about 63Hz. So 3 of the TM's ported will need 24L net. Sealed the RS150-8 needs about 4L and will get you an F3 of about 110Hz. If you wanted to get rid of the baffle step compensation and gain about 6dB in sensitivity, it'll be close due to the sensitivity of the tweeter but looks like it may work.

            The obvious problem with an RS150-8 MTM CC (besides needing a xo designed) is again that 2 x 8ohm drivers in parallel will get you a 4ohm speaker. However, the RS150-4 looks to have an almost identical FR, so using 2 of those in series could work. Sealed needs again about 4L and gets an F3 of about 110Hz. Ported needs about 6-7L to get an F3 of about 66Hz. Again besides the xo design the problem this time is that the sensitivities of the RS150-4 and the ND28F-6 don't work out well without baffle step compensation which isn't really needed for an on-wall application but now sort of becomes necessary here.

            Anyways, if you did go TM 8ohm L and R ported (16L) and MTM CC 2 x 4ohm ported (12L), that's about 30L net for the whole thing. Going again to a sealed MTM CC with the RS150-4, the whole thing will need about 24L again.

            One little problem is that the C-Killa's were designed using an older version of the RS150-8 which has a slightly different FR than the current/newer one so maybe the xo would benefit from a little tweaking. Not 100% sure on that one. But I've actually designed a 3-way W/MTM/W CC using the RS180-8, the RS150-8 and the Peerless DX25TG59-04 for someone to match an existing RS180-8 MTM design originally by Jeff B with pretty good success, so I know that driver, the RS150-8 is workable and has got at least one very happy listener..

            Hopefully not too technical but if any thing isn't clear just ask away.

            Comment


            • #54
              These 8 ohm TM links were provided as a followup to OP's call to Meniscus where he was advised 3 5" TMs in a soundbar would give good performance. The Zaph and cjd/k Ochocinco links have sealed 5' TMs also. In fact if you read the Zaph literature, he states the TM design works sideways and and he provides 3 different XOs with varying BSC.

              All good designers with good components.

              If you can't stomach the silver woofer, buy some PE black speaker grill foam.

              I built cjd's RS150 MTMs. They were designed for the RS150S-8. I scrounged some used ones, and later found that cjd stated the design could use the newer RS150-8 as is.

              And we have another CX120-8 XO to improve on the horrible one I built.

              I don't really care what the OP builds. I mostly post for my own amusement. I do enjoy showing new builders designs beyond the PE kits.

              If I wanted a somewhat bulky L/C/R soundbar to fit in 6' width, and I had a AVR that can't handle 4 ohm loads, I'd build 3 of one of the 8 ohm 5" TM designs I linked.

              If I wanted a compact L/C/R 6' soundbar to look good on the wall under a flatscreen, I'd do the CX120-8. I think the CX120 concept is a modest concept to go with a modest AVR in a modest install.

              Comment


              • #55
                Wolf commented, 04-22-2022, 09:32 PM

                "How high is the Fs magnitude? You didn't answer my question. A max of how many ohms at Fs? If it's a 4 ohm tweeter, and has a max at Fs of 8 ohms, then it likely won't matter much and likely still sound okay. The one octave rule is just a ROT, not gospel."

                Wolf, if you had asked me what the Z magnitude was at Fs, I would have understood your question. As written, it wasn't clear to me what you were asking.

                Agreed, the strict dogma of RoT's can be limiting and in fact I find it much more useful to express this particular RoT not in terms of the frequency magnitude (ie. number of octaves) between the tweeter Fs and the Fxo, but instead in terms of the amount of SPL attenuation below the fundamental at Fs (or above) that is needed. This is based primarily on the levels of harmonic distortion, CSD behavior, mechanical excursion and power limits of the driver near Fs and within useable frequency range. Setting that kind of target will aid me in choosing the xo filter slopes, the Fxo and sometimes deciding if some kind of impedance flattening notch will be helpful, the latter of which will certainly be impacted by the magnitude of the impedance peak at Fs.

                Is that last thing what you were trying to get at? Here the low impedance peak at Fs of CX120-8 tweeter isn't creating any kind of interaction with the xo filter and therefore also not creating any kind pf problematic increase in SPL in the area around the Z peak. If this was what you were referring to above as unlikely to matter much, I certainly agree. Or were you referring to something else I'm not familiar with?

                I did go looking and found some 3rd party measurements of the CX120-8: http://www.dibirama.altervista.org/h...0-40-wmax.html

                For LOUT, just eyeballing the FR's, they look a little closer to BOBinGA's on-baffle measurements than Dayton's spec sheet. And looking at the tweeter harmonic distortion and CSD behavior which both start rising below about 5kHz, it would appear to me that Dayton got it about right when they recommended a minimum Fxo of 4500Hz.

                Just my 2 cents anyways.

                Comment


                • djg
                  djg commented
                  Editing a comment
                  Still willing to send you a pair of CX120-8s to see what you can come up with. No charge.

                • Wolf
                  Wolf commented
                  Editing a comment
                  I see the HD is 2% even at 4.5kHz under typical levels, so that number could be improved. I however see the Fs is only about 7 ohms in magnitude. I think a simple parallel resistor wil knock that down and allow it to be useful lower than 4.5k as the resonance will then be treated. I'd really like to see the HD of Bob's build, and of Bill's Kowaxials for further information. I did not want to see you base the build on the ROT, and void possible options that should have not been omitted. It's not SOTA, but not awful either.

                • mobius
                  mobius commented
                  Editing a comment
                  Thanks again, djg but I just can't fit it in right now. I've got 2 of my own projects that I'm trying finish currently, a design for an on-wall atmos speaker using the RS180-8 in the queue for someone else and 2 more projects again that I want to experiment with after that.

                  Wolf, looking again at dibirama's data, I'm convinced that a Fxo between 4-5KHz is the way to go with that driver.

              • #56
                Here's a Tyger23 design I just remembered. 8 ohm, 4". Small.

                https://techtalk.parts-express.com/f...p-8-and-nd16fa

                Comment


                • oldloder
                  oldloder commented
                  Editing a comment
                  Djg ... thanks for posting this... I love the rs100's but haven't used the paper version and would really like to hear these.

                  OP ... it seemed you were getting so many ideas I didn't float one option I'd thought of which was cjd's Pecorinos. Reading Tyger's comments on how surprisingly low they go and how well they fill a larger room confirms what my experience was with Pecorinos in a larger box ported box similar to the Canzonettas. One of the nice thing about the Pecorinos is there is an on-wall version with no/less bsc and one with full bsc. My sister is auditioning a couple of sets of speakers including them for a soundbar and if I had to pick for her they'd likely be my choice.


                  http://www.divine-audio.com/pecorino/

              • #57
                TLDR, but I remember this. look at the tiny WTMW!

                http://www.htguide.com/forum/showthr...37c5ca7c051dbf

                Comment


                • #58
                  I'm not sure where the OP is in his decision making and I don't want to throw a wrench into the works, but I was having a hard time believing that that HK receiver would only accept 8ohm speakers so I asked them directly and here is their answer:

                  HK: "Our HK AVR 2600 is rated at 8 ohms. You're unlikely to hurt the speakers, using an 8-ohm receiver with 4-ohm speakers. This just means that for a given voltage supplied by the receiver, the receiver is rated to send less current than it will have to power the 4-ohm speakers. 4-ohm speakers can place higher demands on 6- or 8-ohm rated receivers and amps' power reserves, but even those receivers/amps won't run into trouble with 4-ohm speakers, as long as you're not playing music or movies at a very loud volume."

                  https://www.harmankardon.com/AVR+2600_.html


                  So basically, that amp will be fine like most every other solid state amp if the current draw from whatever speakers it's playing doesn't exceed the amp's maximum capacity. From HK's spec sheet, that receiver's maximum instantaneous current capability is 60 amps which is actually really high. Continuous amps available is probably less than that but I suspect is still fairly high. Anyone with more insight on this will hopefully chime in.

                  It's may be worth doing the math to get a feel for what's happening here with the current. I think it's Ohm's Law that states the relationship between power (P) in watts, current (I) in amps and resistance (R) in ohms, so that:

                  P = R*I^2
                  or rearranged,
                  I = SQRT(P/R).

                  You've got 65W per channel on that AVR, so if the speakers are 8ohm nominal, which can typically have a minimum of 6ohm, the current draw at 65W will be SQRT(65/6), which is 3 amps. If you feed that 65W into a 4ohm speaker instead which can typically have a minimum of 3ohm, you get SQRT (65/3), which is 4.65 amps. So when the max current draw is up at 60 amps, I don't think 3 - 4.65 amps are going to present any particular problems even if you run the amp full out at 65W, which you probably should never do in reality.

                  Now let's look at the role that speaker sensitivity can play here. Very simply, the higher the sensitivity of the speakers, the less power you will need to produce any given SPL and therefore the less current as well. So let's say for example that you have an 8ohm speaker with a sensitivity of 84dB 1W/1m. To reach up to 99dB, you'll need 32W of power which will draw about 2.3 amps from the receiver. On the other hand, if you've got a 4ohm speaker with a sensitivity of 90dB 1W/1m, to play up to 99dB you will only need 8W which will now draw about 1.6 amps. So in this example, the 4ohm speaker will actually be drawing less current and actually be easier on your amp than the 8ohm one.

                  Something else to note. When you combine 2 drivers in series, the sensitivity remains the same as with a single driver. But when you combine 2 drivers in parallel, you gain 6dB in sensitivity. So taking 2 x 8ohm drivers and combining them in parallel can often be a better solution than 2 x 4ohm drivers combined in series, especially when you don't have a lot of power to start off with, which is sort of the case here with your HK 65W AVR. This is what I more or less based the above example on.

                  The upshot for me, especially when you consider that with an on-wall soundbar you won't be losing the usual 4 or 5db to baffle step loss, is that I'd be looking at running something like 2 of the Dayton RS150T-8 with the truncated frames in parallel in an MTM resulting in a sensitivity of about 90dB 2.83V/1m (when you take into account the driver's real usable FR not just what the spec sheet sensitivity is listed at) and running them all sealed. The RS150-8's want about 4L each when sealed so you'll need about 24L net for 6 of them and they'll will give you an F3 of about 104Hz. Or I think one of the SB 15's would work just about the same if they can come in on budget. With 2 mono subs right up front under your L&R's, any frequencies above about 80Hz that you might be able to localize and that the sealed soundbar speakers aren't going to be producing are just going to be centered right under the TV which I think will be perfectly acceptable. A much better situation than say if the subs were on a side or rear wall for eg.

                  Also on the odd chance that the OP owns the house and is willing/able to do wall repairs, if you really needed/wanted larger boxes, your front wall setup might actually allow you to make cutouts behind the soundbar and in between the studs to sort of make the soundbar a hybrid in-wall/on-wall speaker combo. If you get lucky enough not to have HVAC or electrical or plumbing between the studs in that front wall that is.

                  Just a few more thoughts anyways.

                  Comment


                  • eavo44
                    eavo44 commented
                    Editing a comment
                    I had looked into this a bit myself but didn’t reach out to HK…thanks for doing that. So it sounds like the receiver may not have an issue powering a 4ohm speaker? When I was looking into this there where some mixed reviews, but most of the info stated that it’s probably not a good idea? Sounded like there was not much of an issue with speaker damage from the pairing but most likely amp damage or sending the receiver into a safety mode where it would shut itself down. If I could run 4ohm speakers in the bar… that would open up some speaker options. Those Dayton rs 150T-8 are nice looking speakers. I haven’t made a decision yet on which speakers to use but I’m leaning towards the bobinga coaxial at the moment. If I can find a solution with a slightly larger woofer with more output then I may go that route. When I first started this project I knew the bare min about speakers but I’ve been trying to learn enough to design and build a quality speaker bar. You guys have helped me out a lot…big thanks! I like the current direction of the bar so far…I just need to explore a few more design options. I do own the home and there is a good amount of wiring in that wall…I don’t think I want to build into the wall just yet. But it could be an option.

                  • eavo44
                    eavo44 commented
                    Editing a comment
                    Hey mobius…is there a speaker design using those Dayton rs150t-8’s with a matched tweeter and xo?…or would using those speakers require a custom xo and tweeter selection?

                • #59
                  This past week was a bit crazy with work and some drywall repair that I needed to address so sorry for not replying earlier. I have read the recent posts and it’s given me a lot to think about. I’m continuing to do more research and am gaining a better understanding of comb filtering and some of the undesirable affects of having a tweeter and woofer placed in a horizontal orientation. It seems like this will always produce a lower quality of sound but it can be mitigated through a well designed xo. I’m still willing to take a small hit on quality if I can get the look of a sleek bar. As a designer, I do want to spend a little time playing with the form of 3 standard bookshelf boxes and see if I can make it feel more like a horizontal element. I want to give this a try because it seems this would be the highest promise of a quality sounding bar. I’ve tried a few things with no success…so I’ll most likely end up developing the bar concept. With the bar, I feel that a max height of 7” is desired. A 6” height would be even better. This allows a 5” woofer to be used for a little more output. A 4” woofer setup would also be a nice solution I’m sure. All the links have been fantastic so thanks to everyone for your input. I placed a call to ask a few questions about the Zaphaudio ZA5’s. Main question was can I reduce the baffle height to 7” and they said it wouldn’t be an issue. A tm,mtm,tm design using these components could be a nice design with good output. If there was a similar design to these with black speakers that would be a great fit. As you mentioned djg, I could use a speaker cover to hide the silver speakers. The Bobinga design is still a top contender for it’s simplicity. If I had to make a pick today I feel like I would lean towards the Bobinga. The pecorino, canzonetta, and ckillas all seem like good options also. It seemed like there may be a mtm design for the pecorino out there but I couldn’t locate one. I’ll work on a few of these options in 3d and post them in the next few days.

                  Comment


                  • #60
                    You know you could forget the soundbar and mount three substantial TM L/C/R speakers and two subs all in the credenza, tilting the TMs back so as to aim at the main listening position. I have low mounted HT speakers that sound fine to me, but then the horrible CX120-8s sound fine to me Just a thought.

                    Comment


                    • eavo44
                      eavo44 commented
                      Editing a comment
                      I think that could make a cool design also…it would give you a lot more room to work with.
                  Working...
                  X