Announcement

Collapse
No announcement yet.

Need help with DIY Hivi Center Speakers

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

  • #31
    Originally posted by Wolf View Post
    At least use the RT1.3WE- it's a lot closer than the RT1C-A is, even though not the same.
    I was pretty certain the Swan S8 **IS** a HiVi Research/Swan product.

    What you are not understanding is that the xover for one driver will have to change to another driver to adapt to optimal. You cannot just swap networks willy-nilly and get optimal results.

    Later,
    Wolf
    The RT1.3 is the same as the RT1.3WE but in a different housing. I just learnt that human vocal range goes from 80Hz bass voice to 1,100hz soprano so that is going to come from the mid-bass driver. It is better I use the F6 instead of L6-R4 so that it with be timbre match with S8.

    I do understand that XO will have to change if the driver change. Btw, can we tweak XO by looking at specification alone? Or it has to be done by measurement?




    Comment


    • #32
      Actually, no- the RT1.3 is the former version (rectangular cast silver face) of the RT1.3WE (round cast black sculpted face), and neither is the RT1.3 (B?) variant stamped steel one used in the DIY3.1 kit.
      You'll find that the drivers used in the kits provided are not always the same as the ones available to you and me via vendors like PE. You'll also find that the specs and FR will vary from unit to unit due to housing and shape alone. Tweak? It's more of a 'we need to start from scratch' kind of thing. Placement, baffle adaptation, proximity to adjacent drivers, box and tuning all play into the result. If you count how many things you want to change from the DIY3.1 kit or the S8; layout, shape, drivers- you are really changing too many things to just translate the xover even if the drivers in fact are the same.

      Tweaking til you are blue in the face is likely what will occur, but there is no guarantee it will be remotely close given how much/many change/changes you want to make.

      Simulation of what you have in the box you build is the best bet to get the results you are trying to achieve. Being that the drivers might be different production versions can throw this off quite a bit. Manufacturer data is only good if the drivers are as is shown. This means the specs could be not what they seem, and we would have a garbage-in/garbage-out scenario. The only fool-proof method are measurements if there is no reliable data to use from the get go.

      Best regards,
      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:
      http://photobucket.com/Wolf-Speakers_and_more

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

      Comment


      • #33
        I did some play around with VituixCAD and came out with this results.
        Click image for larger version

Name:	ACtC-3dzNxJtdhd5vwJh-VY8gTTMNiBG6o4hdQPXU8nhI6bmlL3T0FqK6gX7BxFAPOW-MfVMXhi7DoNvI1HCMybDcFk-RKFoaLroJ3ClBiYxf9ccGwPqOUtfpA37O8PrLkCLQmHhMBsVu-OoUE6SHPeNcQV7Fg=w952-h576-no?authuser=0.jpg
Views:	44
Size:	117.3 KB
ID:	1440948
        Click image for larger version

Name:	ACtC-3e-VosKDs7WBP6K9OK6lv7PzKk9xt5JwVbTr3O0GDi36oq09LAFf6dbCKc5ut-ALxqsrsv7LGQajyeJWGNfI5q89r-j4o-IF6lnMbd17hEZb5xJKHVqTLCnlDp89nrAGIEZcyX-roqLuUAO8L0z2jLg-w=w507-h355-no?authuser=0.jpg
Views:	45
Size:	33.9 KB
ID:	1440947
        Crossover is taken from Hivi Diy3.1 perfectionist mode. The tweeter Crossover remains the same but the mid-range XO is tweaked to accommodate the RT1C-A probably due to the higher sensitivity of the RT1C-A. Slight changes to the woofer crossover. I'm not sure if I did every thing right. Below is a attached frd and zma files. Btw, thanks to Wolf as I manage to get some frd and zma files from file sharing.
        https://app.box.com/s/nsiq3w3c5b1aa4ymm2qjpg3nuhkm315a




        Comment


        • #34
          Are you of the opinion that matching the L & R speakers matters?

          What I might would do is to have the tweeter mounted on it's own "mini baffle" sitting on top of the shelf, and just in front of the tv screen, or flush with it, but leaving a small gap possibly. Then, below that, a box to house a woofer, or two woofers, and a mid. I'd probably lean towards a wave-guide setup that is large enough to cross at 1.2k or lower. There's probably many good ones available, but the xo could become a little trickier. The wave-guide would have fewer diffraction problems that would be caused by all the stuff near the center. I'd be tempted to use a single woofer of good quality and make it a 2-way, rather than build a 3-way.

          What I think is important is the overall tonal balance, and the location will already be messing with that. If you were to sim a speaker, and have multiple sims with different BSC, you could start with one sim, and see how it actually sounds. Changing BSC could be accomplished with simple xo changes, and if you have the xo assembled with jumpers, you could switch from one sim to another in about 45 minutes, and compare.

          If I was really wanting to make things match, (which I might not.bother with ), I'd target the same xo points, and slopes. I would not necessarily use the same drivers.

          Buy drivers that you will have other uses for, or that will have good resale value, just in case your design isn't all you thought it could be. Also, keep in mind that PE return policy is quite long.

          Comment


          • #35
            I just want the Center to match the LR speakers. I just go with the standard W(CM)W design and I'm not really a perfectionist that need to nail every details "perfect". I'm been using the KEF Q300 for LCR for almost 10 years now and very happy with. I just thought I give the Hivi a try.

            Comment


            • #36
              Your LowPass filter seems to be giving a(n unwanted?) "lift" to the woofers around 600Hz.
              In your sim, it LOOKs like it's yielding a +6dB gain to your woofer pair, which WOULD happen if they were wired in parallel, but you get ZER0 gain when wired in series. It's sorta hard to tell, since your dB scale seems too low (by maybe -3dB?).

              Honestly, NOne of your driver simming will be correct inless you create "min-phase" files and use driver XYZ offsets (yours are all dialed in to Zero - looks like).

              All your "bad" summing ( 1k-3k & again >10kHz) should get cleaned up after you create min-ph files and enter proper "offsets".
              You can tell it's bad when the summed response actually lies beLOW your individual driver FR curves.

              The only way you can use "provided" phase data (and skip making minimum phase files) is if YOU measured all 3 driver(pair)s from the exact same mic location, under the same conditions.

              Comment


              • #37
                Originally posted by Chris Roemer View Post
                Your LowPass filter seems to be giving a(n unwanted?) "lift" to the woofers around 600Hz.
                In your sim, it LOOKs like it's yielding a +6dB gain to your woofer pair, which WOULD happen if they were wired in parallel, but you get ZER0 gain when wired in series. It's sorta hard to tell, since your dB scale seems too low (by maybe -3dB?).




                Honestly, NOne of your driver simming will be correct inless you create "min-phase" files and use driver XYZ offsets (yours are all dialed in to Zero - looks like).

                All your "bad" summing ( 1k-3k & again >10kHz) should get cleaned up after you create min-ph files and enter proper "offsets".
                You can tell it's bad when the summed response actually lies beLOW your individual driver FR curves.

                The only way you can use "provided" phase data (and skip making minimum phase files) is if YOU measured all 3 driver(pair)s from the exact same mic location, under the same conditions.
                Yes, there is a 3db lift as seen clearly on the graph below. It's similar to the Hivi Diy3.1 perfectionist mod here. https://diy.midwestaudio.club/discussion/550/hivi-diy-3-1-revisited
                Click image for larger version

Name:	ACtC-3fo_uMYasMGKI_IUE0u7Fd5DgvcE9J3DpqWzXYOEcht-V4v9ae6biFdBl5Lx_qWxodgejvh5D_Gavrg83uXvM3-4KfAf8ce0OOTCVM15kyWkIj2-35_vzW8mucHzXfJa0cECwdSl35uoCs8PMtF75Ro7A=w862-h568-no?authuser=0.jpg
Views:	32
Size:	85.5 KB
ID:	1440981

                the sim is run base on 2 woofer in series. Thank you for pointing that up. I never know that the sum lower the the actual driver is bad. How can I solve that? Btw, I do not have the phase files. Those files I got from google search is what I got. I did attached those files https://app.box.com/s/nsiq3w3c5b1aa4ymm2qjpg3nuhkm315a.

                Comment


                • #38
                  I don't think you're understanding what I'm saying...

                  and, to me, the sim looks better w/the orig. values for C9 & L6 (you're not doing yourself any favors by raising the mid-tweet Fc).

                  IF you're using all the DIY drivers (except doubling up the F6 w/2 in series), what you should really be able to do is keep the HP and BP, and just adjust the LP for a 16n(ohm) load (instead of 8n). You'd do that by doubling coils and resistors, and halving caps.
                  So L4 would go up to 3mH, C1 down to 24uF, and R1 to 4n (which you've already done).

                  You should really read something like Paul Carmody's (undefinition) tutorial on how to properly make use of mfr's F/Z data files.
                  You can't just skip half (or most) of the "prep" steps and end up w/a trustworthy sim.

                  Comment


                  • #39
                    Hi Chris,
                    I do get your point. This is really not the correct way of doing it. I am just toying around with the frd and zma files I could get hold from the internet and I'm not even sure how to use them.

                    However. I tried the sim with changes to the XO for the woofer, while keeping the rest the same value. But I got this funky looking mid-range FR. It must be the frd and zma for the DMN-A that I get from googling that is not correct? So keeping the HP and BP, and just adjust the LP for a 16n(ohm) load (instead of 8n) should be fine as long as mid and tweeter remain the same.
                    Click image for larger version

Name:	ACtC-3c3IQ0-mo_5s8iOlEGX2lfX6I2HknUu0r549NMrh00FOdC7j2Yk7N27b_NCwHCQANW_jXJVzu725XlPrvZ4CUI1Eb2AfbuNT3fjWEPo1RLlyB8Q3m66zwtmRa-WPsdQxD677ftqjyluiFATn9JM5f-A1A=w508-h356-no?authuser=0.jpg
Views:	22
Size:	34.8 KB
ID:	1441046
                    Click image for larger version

Name:	ACtC-3fkkjyrutom9QAJQNKtrtocNEsJXeXvtYzzG0zF-0tIqe7ukBgn9PE7ngvgXnCWS2lIyIHwFWKNc62DBvG51ooTGFCMRFa-b4Gs059ZPB0sWqTvxZ84pyRNWTn_LsPuxr7u11QU3LDatsD--OqYRYS6YQ=w862-h490-no?authuser=0.jpg
Views:	22
Size:	98.1 KB
ID:	1441047

                    Comment


                    • #40
                      Not sure what you're doing there, but I can SEE the FR for a single F6 (which doesn't seem right @ about 80dB) and it's being boosted by +6dB in your sim which ONLY happens w/2 drivers in parallel. When 2 are in series you get ZERO gain in sensitivity. (Having 2 drivers gives you +3dB (from doubling the cone area), but in series doubles the impedance which loses -3dB back again. Running them in parallel yields a 4n load - giving you +3dB MORE - so +6.)

                      Comment


                      • #41
                        I know what you mean. The F6's red and black line FR suppose to be overlaying each other for F6 in series setup but it appear that they are in parallel now. I put in the driver info into VituixCAD and this is the result I get. As you can see from the 2x F6 in series at the crossover setting section. This thing is driving me crazy.

                        Comment

                        Working...
                        X