New system is ready! Take a look

Collapse
This topic is closed.
X
X
 
  • Time
  • Show
Clear All
new posts
  • Victor
    Senior Member
    • Apr 2002
    • 338

    New system is ready! Take a look

    Here are the pictures of my new all-dipole system. Both channels are ready now and I am working on the active crossover, so I have no idea how this monster actually sounds. Both towers in each channel stand at 76.5 inches tall.

    and

    The midrange baffle has four 10-inch drivers; - they are SS25W8565 wired in the series-parallel manner. The midrange baffle is 14 inches wide with 3 inch wings folding back. The drivers are held with the back spine, so no bolts are used.

    The woofers are standard and well-known Phoenix design from S. Linkwitz with Peerless 12-inch XLS 830500 drivers for a total of 4 drivers per channel. Here the view from the back,

    The main RD-75 baffle is a curved and sloping affair. The shape was influenced by the research done by John Wittaker and the article by Tom Pirazella. It is 24 inches at the bottom and 18 inches on top counting the curved part. The straight part of the baffle is one-sided trapezoid with 18 inches on the bottom and 12 inches on top. The RD-95 baffle has no sharp edges anywhere in an effort to minimize the diffraction effects.

    Well the pictures are not all that great, but they give an idea of what the system is. It took more then a year to build this and it is huge. I somehow did not realize how big it was going to be as I was building it.

    I intend to use my homemade amps to power the system. I have on hand 4 stereo units that can deliver a clean 500 watts into 4 ohms per channel. Interestingly enough I am also looking at the digital pro-crossovers. The Behringer DCX2496 product is scheduled to appear sometime this summer and it is looking particularly attractive with 24-bit capability and lots of useful features, - all the cost of $350US. So, who knows, by the time I am done with my analog circuit I may have this Behringer box.

    The crossover points for the system are 70 and 300 Hz, all 4th order L-R. The notch filter for RD-75 is active and I am also using a fair amount of high frequency boost to EQ the RD-75 nearly flat up to 17 kHz. The boost is 10 dB and it is rising exponentially from at about 10 kHz (0 dB) and finishing at 20 kHz (+10 dB). The midrange is equalized with a 6 db/octave shelving lo-pass filter from 550 Hz and down to 20 Hz. It also has a notch of about 5 dB deep and 200 Hz wide at 320 Hz to account for the 25W driver basket resonance. Apparently there is a delay between the Phoenix woofers and the Midrange assembly. Since the midrange sits on top of the woofers I had to account for this delay electronically. So the Midrange crossover also has a delay circuit of 1.45 mS. All-in–all there are a lot of op-amps in this cross-over.

    Well, let me know what you think of it.

    Thanks,
    Victor
  • Lex
    Moderator Emeritus
    • Apr 2001
    • 27461

    #2
    Wow! 8O Great work Victor! I'm impressed.

    Lex
    Doug
    "I'm out there Jerry, and I'm loving every minute of it!" - Kramer

    Comment

    • David Meek
      Moderator Emeritus
      • Aug 2000
      • 8938

      #3
      OMG! :yeah:

      Victor, if it sounds ANYTHING like it looks, you'll have a speaker system to die for. As a guy that does cabinetry for a hobby I can say you have done some meticulous and beautiful work on the housings.

      When do we get a demo? And how do we get there?




      David - HTGuide flunky
      Our "Theater"
      Our DVDs on DVD Tracker

      .

      David - Trigger-happy HTGuide Admin

      Comment

      • Andrew Pratt
        Moderator Emeritus
        • Aug 2000
        • 16507

        #4
        Wow! I'm going to have to see about getting over to thornhill the next time I'm in toronto. Nice work:T




        Comment

        • Hank
          Super Senior Member
          • Jul 2002
          • 1345

          #5
          Victor - the cabinet work looks beautiful!! My plan is to do an RD50 version, but I wanted to put the ribbon and mid-woof stack in the same enclosure. Also, Jon and Thomas have about convinced me to go monopole. I wondering if my single cabinet plan will result in cabinets that are too big and heavy. Hmmmm...
          Jon and Thomas are better equipped than I to comment on your design. They have given good advice on xovers for line source applications.
          I'd love to hear those beauties!

          Comment

          • Dennis H
            Ultra Senior Member
            • Aug 2002
            • 3791

            #6
            Nice, Victor! Your workmanship looks beautiful.

            Comment

            • Eduardo
              Moderator emeritus
              • Jun 2002
              • 1258

              #7
              Great job Victor. Your speakers look awesome. I hope they sound as good when you get your crossovers finished.

              I know Tom Pirazella. He works for one of my customers. He and I have spoken extensively about speaker building and he's way above my head. Anyways I am going to send him a link to this post and ask him to comment.




              http://home.nc.rr.com/ejimenez

              Comment

              • sfdoddsy
                Senior Member
                • Sep 2000
                • 496

                #8
                Add me to the chorus of praise. Those should be marvellous sounding speakers.

                Just a note on digital crossovers. I'm waiting for a DCX 2496 for my surrounds and have a dbx Driverack 260 and couldn't be happier with it.

                However, I have been alerted to a potential problem when using the Peerless XLS drivers as in the Phoenix or Orion. The XLS, as you know, has a very low Qts and this requires additional EQ.

                The query is whether the digital crossovers are the right way to apply this much boost. This is one of the reasons I went with Adire DPL12s.

                But great work again.

                Steve




                Steve's DIY Dipoles
                Steve's OB Journey

                Comment

                • PMazz
                  Senior Member
                  • May 2001
                  • 861

                  #9
                  They look great!

                  Could you elaborate on the back spine system to hold the drivers.

                  Pete
                  Birth of a Media Center

                  Comment

                  • Victor
                    Senior Member
                    • Apr 2002
                    • 338

                    #10
                    Guys,

                    Thanks for the vote of approval on my new system. The motivation for this system was my desire to improve on my existing Carver Platinum Mark IV’s (which will now be relegated to the surround duty). The article by Tom Perazella made me look at the RD-75 and the race was on.

                    I believe it was Jon who called this system, when it was still in ‘discussion stage’, a ‘Chinese Menu’. It is really that, since it incorporates a lot of elements from various sources, and it is in fact a ‘little bit of this’ and ‘little bit of that’.

                    The woofers and the midrange panel came courtesy of Z. Linkwitz’s Phoenix Dipole Speaker. I simply modified the main Phoenix baffle and then used the mathematics available on Linkwitz’s site to extend the Phoenix design to what I needed in order to incorporate the larger drivers.

                    The RD-75 wing-type baffle materialized from many e-mail that I exchanged with John Whittaker, who did a fantastic job documenting the ‘ins and outs’ of the RD-75 driver behavior on the open baffle.

                    So, in the end, I simply integrated the ideas of others, more knowledgeable then me. As it happens I do not have all that much experience with speakers. My thing is the analog integrated circuit design.

                    When do we get a demo? And how do we get there?
                    Well, I have not had a demo yet. I shall report once I get the system fired up, hopefully soon. I am in the northern suburb of Toronto, so if anyone is ever in this neck of the woods I’ll be happy to demonstrate.

                    Hank wrote,

                    My plan is to do an RD50 version, but I wanted to put the ribbon and mid-woof stack in the same enclosure. Also, Jon and Thomas have about convinced me to go monopole. I wondering if my single cabinet plan will result in cabinets that are too big and heavy. Hmmmm...
                    Look here http://www.angelfire.com/nt2/flex/images.html for the pictures of the system that is similar to what you are thinking about, although it is a dipole implementation.

                    My system is definitely big and heavy. But then again, it is not that uncommon with dipoles. The ergonomics was not a design parameter, as I, from the very begging, secured the ‘ok’ from my ‘better half’. Naturally there is no free lunch and I ended up contracting for the home renovations as a price for the permission to build my new speakers. Incidentally, I think if I ever bring the subject of speakers again I may be exiled back to Siberia. So this is it for a while for me.

                    Steve wrote,

                    I have been alerted to a potential problem when using the Peerless XLS drivers as in the Phoenix or Orion. The XLS, as you know, has a very low Qts and this requires additional EQ. The query is whether the digital crossovers are the right way to apply this much boost.
                    Well, Alex Flex who posts often on the Dipole List managed to EQ his Phoenix woofers with a BSS Micro drive. Here is the link to his site http://www.angelfire.com/nt2/flex/phosubeq.html where he shows the EQ curves and gives some explanation. His curves are nearly identical to my SPICE-simulated results for those woofers. I simply used the circuit suggested by Z. Linkwitz. So I would say it is possible, but the concern you raise is a good one. Peerless XLS driver will need about 10-12 dB of boost at 20 Hz to account for the low Qtc. I looked at the capability of available boxes from Rane, DBX, BSS and the Behringer; - they all seem to be fine with about 20 dB of signal boost. Can you tell me more about the concerns that you heard?

                    Pete wrote,

                    Could you elaborate on the back spine system to hold the drivers
                    The back spine system was shamelessly ‘borrowed’ from Linkwitz’s Pheonix main baffle design. Here is what Linkwitz says about the benefits of this type of mounting

                    This avoids the basket-magnet (spring-mass) resonance that is typical for a driver with the basket screwed to the front panel.
                    You can get more detailed explanation on his site. Here is the link http://www.linkwitzlab.com/m_panel.htm

                    In principle, the tension between the back spine and the inside of the front panel of the baffle holds the drivers in place. Another benefit here is that since there is no need for the bolts, you have no problems with the cosmetics of the front of the baffle and you also will not have to contemplate the grill to cover those bolts and drives.

                    The more I think about the cross-over the more I realize that the active analog crossover implementation is quickly becoming a thing of the past. Behringer DCX2496 is a product that is relatively inexpensive and it offers options that I can never built with op-amps. The necessary A/D and D/A conversion process associated with those new digital boxes seem like an issue but judging from the published specs it should not be an issue at all.

                    For some reason all the digital boxes on the market use 48 kHz sampling rate. So that any digital signal that you feed in will get resample to 48 kHz. The Sample Rate Converter that does that is very tricky and the distortion may be added particularly at higher frequencies, but still it is a small price to pay for the benefits that the digital box offers.

                    I think I will also buy the RTA software and will use it with the digital box to EQ my room. My Behringer dealer here in Toronto tells me that the DCX 2496 will become available in May-June, so most likely I will wait with the system test until then. My analog circuit is all designed but I do not want to invest in a new PCB since I can get a digital solution within a month or so.

                    Victor

                    Comment

                    • Dennis H
                      Ultra Senior Member
                      • Aug 2002
                      • 3791

                      #11
                      For some reason all the digital boxes on the market use 48 kHz sampling rate
                      The Behringer samples at 96 kHz, 24 bit; hence the model number. One thought I had about extreme dipole boosts and maintaining digital resolution is to add a simple, passive, RC filter between the digital box and the power amp to do the basic 6dB/octave dipole slope. Then you could fine tune it with the digital box's filters.

                      Comment

                      • sfdoddsy
                        Senior Member
                        • Sep 2000
                        • 496

                        #12
                        Yes, SL applies around 40dB of boost when using the XLS drivers since he EQs all the way down to 5Hz or so to minimise potential group delay problems.

                        I've not had problems with over driving my dbx 260, but I only need a 20dB boost, and possibly even less if I decide to use a monopole sub below the room resonance point as SL is now recommending.

                        Have you put your desired EQ into the DCX 2496 software? I know someone who put his desired curves into their test software and ran out of processing power.

                        Cheers

                        Steve




                        Steve's DIY Dipoles
                        Steve's OB Journey

                        Comment

                        • Victor
                          Senior Member
                          • Apr 2002
                          • 338

                          #13
                          Dennis,

                          I looked at the published Behringer specs and, yes, it does say that the system frequency is 96 kHz. It is a little hard to believe, but it seems that it is true. I would like to find out what exactly do they mean by system frequency. It also says that the digital input can accommodate the digital input stream at any sampling rate from 32 to 96 kHz. Now this is a clear statement.

                          All other products I looked at use 48 kHz rate to re-sample whatever is being fed into the digital input. This makes more sense to do from the engineering point of view. Although the higher sampling rate does considerably minimize the requirements for the anti-aliasing filter, however, the sample rate conversion process is sensitive to how far apart the input and the output frequencies are. The error that is inevitably incurred as a result of re-sampling would introduce, in my opinion, worst effects then the higher order anti-aliasing filter would. In any way I would not make such a trade off, but, who knows, perhaps Behringer engineers know sometime I do not or their Sample Rate Conversion process is free of the errors that I am referring to. Interesting indeed.

                          The AKM converters are indeed 96 kHz capable. That does not mean that they must be used at that rate. I did see in the past manufacturers making claims to higher sampling rate only because their equipment was internally capable of it, but in fact the processing was done at the lower rate. Nowhere in the DCX2496 manual does it say explicitly that the internal processing is done at 96 kHz. Other products such as Driverack 260 by DBX state clearly that the input signal is re-sampled at 48 kHz. Well, it may not matter in the end as the DCX2496 has very good THD specs anyway.

                          By the way, I like your idea of using a passive RC filter at the output of the digital box.

                          Steve,

                          I looked at the analog circuit the Z. Linkwitz is suggesting to use in order to EQ the dipole woofer with XLS drivers. The circuit shows that the boost below 80 Hz is a little less then 20 dB, not 40 dB. The digital box should be able to do it. Also the dipole EQ is taken to 17 Hz not 5 Hz. Apparently the cabinet has a resonance at 17 Hz.

                          Have you put your desired EQ into the DCX 2496 software? I know someone who put his desired curves into their test software and ran out of processing power.
                          Well, I do not have the DCX2496 yet. Also, can you control this box with a computer? I have not noticed any software, like the Driveware from DBX or the DragNet from Rane, offerings from Behringer. Do you know if such software will be offered?

                          Victor

                          Comment

                          • sfdoddsy
                            Senior Member
                            • Sep 2000
                            • 496

                            #14
                            Victor,

                            The DCX can be driven by software and it is available here:



                            Steve




                            Steve's DIY Dipoles
                            Steve's OB Journey

                            Comment

                            • Victor
                              Senior Member
                              • Apr 2002
                              • 338

                              #15
                              Steve,

                              I looked at the circuit again and I was wrong with my earlier figures and you were right. The boost is indeed around 40 dB, which makes too much for the digital box.

                              Victor

                              Comment

                              • Davey
                                Senior Member
                                • Jan 2003
                                • 355

                                #16
                                Victor,

                                The dipole and Q correction for the 830500 drivers in the Phoenix woofers is rather severe as Steve mentioned. But also keep in mind that there will probably be a relative level shift downward in level with respect to the rest of your system. This might ease the requirements a bit on the digital crossover, but it still is not the optimum way to use these units.

                                As Steve mentioned I ran out of processing power on the DCX2496 when trying to program the exact filter and equalization requirements for SL's Orion system. It might however work for your system. I will be interested to see your results.

                                Best regards,

                                Davey.

                                Comment

                                • Samppo
                                  Junior Member
                                  • May 2003
                                  • 2

                                  #17
                                  Victor: Where can one join the Dipole list? I'm about to build a pair of dipole speakers and that could come in very helpful..

                                  Also noticed the Res Sky ribbons on Alex Flex homepage. Very interesting page. Has anyone had the chance to listen to these Res ribbons or compare them to eg. B&G's? On Alex homepage I couldn't find any contact information to email and ask him.

                                  I emailed Res about half an year ago, and they said a new version of the Res ribbon is coming out before summer.

                                  Samuel

                                  Comment

                                  • Victor
                                    Senior Member
                                    • Apr 2002
                                    • 338

                                    #18
                                    Samuel,

                                    Dipole List is hosted by 'topica'. Go to http://www.topica.com/lists/dipoles/ You will need to register. Alex Flex is a frequent participant there. His opinion of Res Sky ribbon was that it did not give him the kind of SPL he was looking for.

                                    Victor

                                    Comment

                                    • JonMarsh
                                      Mad Max Moderator
                                      • Aug 2000
                                      • 15259

                                      #19
                                      Beautiful work, Victor!

                                      It's amazing what happens sometimes when I'm out of town on business for a few days! I don't know how you can stand the anticipation of waiting to setup, tune, and start listening to these beauties!

                                      Now I just have to figure out a way to be in Toronto on business some day, after you get these things up and running!

                                      Just out of curiousity, how long overall have you been working on this system, especially the cabinet contruction and finishing? I know a really big pile of work and effort when I see one.... my biggest problem these days is not lack of motivation, it's just plain lack of time..

                                      Congratulations on the excellent work!


                                      Best regards,

                                      Jon




                                      Earth First!
                                      _______________________________
                                      We'll screw up the other planets later....
                                      the AudioWorx
                                      Natalie P
                                      M8ta
                                      Modula Neo DCC
                                      Modula MT XE
                                      Modula Xtreme
                                      Isiris
                                      Wavecor Ardent

                                      SMJ
                                      Minerva Monitor
                                      Calliope
                                      Ardent D

                                      In Development...
                                      Isiris Mk II updates- in final test stage!
                                      Obi-Wan
                                      Saint-Saëns Symphonique/AKA SMJ-40
                                      Modula PWB
                                      Calliope CC Supreme
                                      Natalie P Ultra
                                      Natalie P Supreme
                                      Janus BP1 Sub


                                      Resistance is not futile, it is Volts divided by Amperes...
                                      Just ask Mr. Ohm....

                                      Comment

                                      • Victor
                                        Senior Member
                                        • Apr 2002
                                        • 338

                                        #20
                                        Jon,

                                        It took a bit more then a year and two people to complete this project. I have a friend who has a complete workshop, so the tools were not an issue. The finishing along took more then 3 month to do. My friend who is a master cabinet maker led the way. I do like how it turned out.

                                        If you recall we talked about the concept for this system sometime last summer. I am still agonizing over my decision to physically separate the RD-75 and the midrange baffle. There is about 14 inches center-to-center distance from the ribbon to the SS25W8565 midrange woofer. I hope that there will not be any detrimental effects, - what do you think?

                                        I am trying to speed up the process of my cross-over construction. I have the amplifiers standing by though. The cross-over is shaping up as a two-piece solution, - the digital box like the Behringer DCX2496 and the standalone box with some additional electronics. The Behringer has not yet gone on sale in Canada, but it is anticipated within a month.

                                        My intent right now is to use the Digital box doing the entire RD-75 duty, all of the L-R filters for the 3-way, the delay, the notch filters, the 3-way relative level adjustments, the XLS woofer dipole equalization and the room EQ. The outside box can be fitted with electronics to do just the dipole correction for the midrange and the EQ to account for the lower Qts of the woofers.

                                        I may even 'steal' the power from the DCX2496 to power up the external electronics. I can't imaging it be all that difficult, - just open up the box an find the power rails and the ground near the power supply. The external box may also contain the balanced-to-single-ended converters since my amps do not have balanced inputs, although I am sure that a custom cable can be made to simply 'neglect' the (-) feed.

                                        So, with all this in mind, if you get to Toronto sometime in the end of June, beginning of July time frame, I’ll be ready to go. I already have a good bottle of a single malt scotch to go with it.

                                        Victor

                                        Comment

                                        • Samppo
                                          Junior Member
                                          • May 2003
                                          • 2

                                          #21
                                          Thanks for the information and great link Victor!

                                          Btw. You´ve made some great looking speakers

                                          best regards
                                          Samuel

                                          Comment

                                          • tperazella
                                            Junior Member
                                            • May 2003
                                            • 1

                                            #22
                                            Victor:

                                            This is really great. I'm glad to see that I was an inspiration for you to take this route. Your dipole midbass is especially interesting. Late last year I finished a dipole midbass using 6 Peerless 831727 CC line 10" woofers per side. The results are fantastic. I hope to be writing the construction project up for AudioXpress in the near future.

                                            I also agree with using the Behringer DCX2496 to control all this. I have one on order and it should arrive next week. Also coming in the same shipment is the DEQ2496. I plan to use the DCX to set the crossovers, time delay, and basic EQ. The DEQ will be used to set up around 10 curves to "normalize" different types of source material I'll be playing.

                                            The one area where I have stayed away from dipoles has been the bass. My sub, which appeared in a past issue of Speaker Builder, uses a closet in a loft as the enclosure. It is just too hard to get low bass with a dipole. The volume displacement requirements get out of hand when you try to get low bass. With the non-dipole sub I have, the F3 is 12.5 Hz and it will do 122dB at 16 Hz with less than 10% distortion.

                                            Again, truly outstanding job.

                                            Comment

                                            • Victor
                                              Senior Member
                                              • Apr 2002
                                              • 338

                                              #23
                                              Tom,

                                              I really started thinking about this project after I read your article in Audio Express. Also, at the time, I had been using Carver Mark IV speakers, so the line source planer drivers was something I could really relate to. I am happy to hear that you like your midrange arrangement. Hopefully what I built should not be all that different from yours in spite of the lesser number of drivers and overall position of the array. It seems that it could work fine. My crossover building efforts have been ‘sabotaged’ lately so the listening tests are yet to be done.

                                              Looking at your midbass array I see that the width of the baffle you built is about the same as mine. Can you post the details of the crossover your are using for your midrange? I am particularly interested in the parameters of the shelving low-pass filter you are using.

                                              It is good to know that you are expecting the DCX2496 in a week. Here in Canada I am yet to have a firm response from Behringer rep on the arrival date. Also there were some concerns raised as to the headroom ability of the Digital crossover. It seems that it will be difficult if not impossible to generate the appropriate low-pass shelving filter that is needed for the midrange, since it must cover around 4 or 5 octaves, which puts it at 24-30 dB of gain. Certainly the EQ for my Phoenix woofers is outside the DCX2496 capabilities, as they will need even more gain at low frequencies.

                                              I did consider the shortcomings of the dipole bass. However, in the end, I wanted to have an all dipole system. Naturally, I will never achieve anything that will even remotely resemble the kind of bass you can get with the IB or sealed box, but I know I can get a very clean and powerful 25-30 Hz output from my woofers, which is fine for nearly all music material. The goal for my system was to achieve the smallest THD, with numbers hopefully approaching those of amplifiers.

                                              I am planning to build a super-sub with at least 15 Hz capability that can do more then 110+ dB SPL. Most likely it will end up being a sono-tube project with two 15-inch drivers. Right now I am risking an expulsion from paradise if I even mention this to my ‘better half’, but I know, - the time cures everything. So I am thinking of augmenting my Dipole Subs with this super-sub. The crossovers points to be determined experimentally.

                                              Regards,
                                              Victor

                                              Comment

                                              • tominizer
                                                Member
                                                • Aug 2003
                                                • 60

                                                #24
                                                Wow!!! Awesome work. And you know what's even more awesome, I'm only about 15 minutes from Thornhill And so if you ever decide to provide any sort of demo's, I will gladly sacrifice time and effort to be there................ on behalf of everyone else who is drooling on this thread :LOL:

                                                But seriously, truly very nice looking pieces. I hope all works out when you power those beauties up.

                                                Comment

                                                • Sonnie Parker
                                                  • Jan 2002
                                                  • 2858

                                                  #25
                                                  8O

                                                  Where was I ??? This is fairly old thread from May.

                                                  None the less... Victor... those babies are GORGEOUS!!!!! Awesome work!






                                                  SONNIE

                                                  Cedar Creek Cinema

                                                  DVD Collection

                                                  BFD Comprehensive Setup Guide

                                                  Comment

                                                  • Victor
                                                    Senior Member
                                                    • Apr 2002
                                                    • 338

                                                    #26
                                                    tominizer,

                                                    It is always great to know that a fellow DIYselfer lives near by. E-mail me at vlb@rogers.com and sure, I will set-up a demo for you. At this time I am deep into system set-up and do not have both channels working, but I shall have everything sorted out soon. My multi-channel volume control is still in construction stage, but I have a 2-channle version of it (temporary replacement) almost working. So in a few weeks the system will be totally up and running.

                                                    Sonnie Parker,

                                                    Thanks!

                                                    Victor

                                                    Comment

                                                    • brucek
                                                      HTG Expert
                                                      • Aug 2000
                                                      • 303

                                                      #27
                                                      Victor,

                                                      Very nice speakers. Obviously space isn't a problem in your home.
                                                      I was thinking about your intent to use the Behringer DCX2496 as the active crossover. I have a small concern.

                                                      I wonder if there'll be a level matching problem. Since you are not feeding this device a fixed line level where you can predict the maximum signal to accomodate the ADC, won't dynamic range be a problem? As you well know, anytime you insert an A to D sampler into an analog chain, theoretically to take full advantage of every bit, you want your loudest signal to approach the maximum input level of the sampler (without enabling the clipping compressor) since the number of bits used defines your dynamic range and as a result your S/N ratio.

                                                      If your loudest level for example, on a day you want to listen to quiet music, was just above the noise floor, then you certainly wouldn't be using all the bits available to define the voltage levels of the input samples. So, from the softest input level to the loudest input level, less bits would be used to define the samples and after conversion back to analog your resultant signal wouldn't represent the input accurately.

                                                      The spec claims a system dynamic range of 109dB. I guess that means they are saying that of the 24 bits, about 6 are lost to the noise, leaving about 18 to define your samples. That's certainly really good and would be great if you fed it the max level, but if you had your volume low, I wonder if it would be good enough - you might be robbing yourself of dynamic range? Just a thought.

                                                      I don't think with that many bits you will perceive any quantization error (noise), but there's always going to be quintile errors, and the more bits used in the conversion process, the less it's noticed. In this situation you are feeding a variable voltage range of signals depending on where you have your volume control?

                                                      I use a Behringer Feedback destroyer in my sub chain and the level matching is a bit of a problem, but in a sub I don't really care that much. It's not a big deal at very low frequencies because subs will generally have far higher distortion than what the A/D & D/A will introduce. And I certainly don't have to be concerned about thermal or upper frequency noise since the sub is band passed. I experimented a little one time and put the unit between one of my preamps and power amps. Yuk. sssssssssssssssssssssssssssssssssss. I hope the DCX2496 is better.

                                                      The THD+N at 10dBu is spec'd at .004%. Thats about -87dB (14-15 bits) with a 10dBu input (~3voltsRMS) if my math is right. I guess that's pretty darned good?

                                                      I was also wondering about processing delay. They claim a typical 1ms delay, which would be about a foot. If you used the DCX2496 in concert with an analog crossover as you discussed, would you have to compensate somehow? I didn't really understand how you were integrating the two together.

                                                      I don't know, but do you really like digitizing that rather pristine connection between your pre-amp and power amps? I kinda like your idea of doing it completely analog.

                                                      Anyway, interesting stuff. I'll be interested in reading how it goes.

                                                      brucek

                                                      Comment

                                                      • Victor
                                                        Senior Member
                                                        • Apr 2002
                                                        • 338

                                                        #28
                                                        Brucek,

                                                        You are raising many interesting points in your post. First of all let me illustrate the signal chain of my system. I have both the DEQ2496 and DCX2496. They are ‘digitally’ connected through the balanced AESBU line with DEQ in the front so to speak. The signal chain is as follows:

                                                        - An analog pre-amp (heavily modified Bryston BP-25)
                                                        - DEQ2496
                                                        - DCX2496
                                                        - Analog post-EQ for the subs only
                                                        - 20-channel passive volume control
                                                        - Power amps

                                                        You are quite correct when you raise a concern regarding the input voltage level. For the DEQ this level is either 4-5 Volts or 9-10 Volts depending on the position if the switch in the back of the unit. This switch simply inserts the amplifier with a gain of 2 into the signal path. Naturally, independent of the switch position the analog output of the DEQ and the DCX will be at maximum if the input level is at maximum, i.e. input and output levels are the same in amplitude.

                                                        I take care of the input voltage issue by placing an analog pre-amp (Bryston BP25) before the DEQ. The pre-amp also serves as an input source selector. Since Bryston BP25 has balanced outputs, - I used them to feed the DEQ inputs. I set the input voltage levels with Bryston BP25 volume control while monitoring the input level as displayed on the DEQ meter. It is quick and easy. Most recordings are more or less the same in level so I do not have to touch the input level too often. The DEQ input is therefore maintained at the comfortable -3-6 dB level on average. In this way the A/D converter uses its full dynamic range.

                                                        I connect my CD Player (heavily modified Denon DCD2526) to DEQ through SPDI/F line. This way the signal is kept in digital domain all the way to the DCX and no input level issues to contend with.

                                                        My analog sources are:
                                                        - The 5.1 analog output from my computer sound card Delta 410 FireWire
                                                        - Soon to be acquired SACD and/or DVD-A player

                                                        All in all I think that I solved the input voltage level issue.

                                                        I am forced to use the post-EQ analog circuit for my dipole subs. You see, the subs require compensation for the low Qtc of the drivers, also the 6 dB/oct dipole roll-off compensation, and a notch filter. I can do all but the dipole roll-off in DCX2496. The problem here is lack of headroom since at low frequencies the total boost is nearly 35 dB and the input voltage level is high. So I made a special cable with XLR connector on one end and RCA on another and I take the sub analog output from DCX and feed it into my external post EQ box that only has RCA plugs.

                                                        I place a simple resistor divider at the sub’s output of the DCX and attenuate the signal by about 11 dB. Then I feed this signal into an analog low-pass shelving filter, - my post EQ box. I use a rather high +/- 24 Volts power supply to prevent any possibility of clipping. Using this scheme along with the integral level control built-in in DCX, it is possible to equalize the level down to what it must be, which is a few Volts at most, without any loss of resolution. Experiments with various sound tracks showed that I only need to turn down the level control in the DCX by at most 3 dB. Well, I am still tweaking!

                                                        So at this point I have 3 signals, the ribbon signal, the midrange and the sub signals all going into my monstrous volume control box. I built 20 channel level control box using 5 motorized Alps 4-gang pots. I can electronically control each or all or any combination of those pots with universal remote control. Each pot handles one full sound channel. For example, my front speakers are 3-way, so I need 2 pots to control the Right and the Left channels. My rears are 2-way, so 2 more pots are used to control them. Additionally I have the Center channel, so one more pot is used there. There are unused sections in each potentiometer. Well, who knows, I may turn my 2-way rears into the 3-way some day.

                                                        This volume control is in fact the Master Volume Control. It controls the overall level of all signals just before the power amps. As you can see, I use the built in level control facility in DCX, which is +/- 15 dB, in order to equalize the sensitivities between the sub, the midrange and the ribbon. I use at most +/- 3 dB range in order to accomplish this task. I do not use the DCX to control the volume, - that would be a mistake.

                                                        The processing delay of less then 1 mS does not seem to be a problem. Certainly it is not a problem with music since all channels are equally delayed. I do not think it’d be a problem with the video, but I have yet to test this.

                                                        The DCX and the DEQ units are specified at 112 dB S/N with full range input. These are very respectable numbers. They are not, however, representing the 24-bit capability that A/D converters are designed to do. Furthermore, I am sure that this number is also A-weighted, i.e. taken with a brick-wall 20 kHz filter. This is not the real world measurement, so we must discount this 112 dB figure as a simple marketing ploy.

                                                        In the real world it is very difficult to obtain a S/N number close even to 100 dB. This is 16 bit resolution. There are many reasons for that, chief among them is power supply noise and totally inappropriate use of op-amp based circuits. To top it of we must understand that those 24-bit converters are in fact Delta-Sigma types and as such they use over sampling techniques. This produces internal noise that is impossible to control. As a result I have yet to see the LSB (least significant bit) on my scope. In fact 3-4 lower bits are completely masked by noise. Sony’s Pulse-Current D/A is an exception and I have no idea how the Sony engineers manage to make a clean DAC all the way to nearly 20 bits.

                                                        So considering those factors, getting 16-bits is all we really can do. The question can be is this good enough? I think so. The digital processing that can output 100 dB S/N is completely transparent, particularly when you consider that even my monstrous speaker system will still have close to 0.5-1% THD. On the other hand, the comfort of changing parameters with digital boxes far outweighs any minute THD-related gains that an analog circuit may produce.

                                                        I will make an analog cross-over for my speakers sometime next year and will compare it to DCX2496 implementation.

                                                        Regards,
                                                        Victor.

                                                        Comment

                                                        • brucek
                                                          HTG Expert
                                                          • Aug 2000
                                                          • 303

                                                          #29
                                                          All in all I think that I solved the input voltage level issue
                                                          Oh for sure. I had some wacky ideas in my head of feeding each line source directly into the EQ and Crossover, but it would have been ugly. Using the BP25 to switch and level control is a great idea.

                                                          I love your volume control.

                                                          I wonder if you've listened to the noise level from your ribbons with all this hooked up yet? You must be anxious to hear what the resultant noise floor is like after these extra conversions. As you say, the spec's bare out the expectation of 'no problem' with noise. My trepidation on this comes not only from my usual suspicion of published specs but from my experience with a Behringer Parametric EQ - DSP1100 (BFD) unit that has pretty good published specs but was a poor performer in the noise floor department when I put it before a power amp. As a sub EQ it is indispensable, I love it, but for regular duty it had too high a noise floor. It's specs weren't up to the level of the two units you're using though.

                                                          I wish they would be consistent with specs - particularly regarding noise. I love the spec for my DSP1100. It reads, greater than -94dB, unweighted, 20Hz to 20KHz. Oh brother, if it's unweighted, what the heck is with the upper 20KHz limit. No doubt they've got a perfect low pass filter killing all the noise energy above 20KHz. This is worse than A-weighted, which uses 22K I believe. That noise above 20K will certainly have an effect in the audible range. This 20K limit no doubt accounts for their rather impressive THD also. Anyway, the unit had too high a noise floor for audiophile types in my opinion, given the filtered -94dB... This is the only reason I raised the flag on the noise issue when I saw the units you were using. The new specs on your units look real good though and I expect they'll be great.

                                                          I do not think it’d be a problem with the video
                                                          Not a chance. Even with processing time of two units of 2ms is only ~2feet, it wouldn't effect video/voice sync. I had just thought you were running some sort of parallel digital/analog deal that might effect the sound timing.

                                                          I wonder if the phase shift caused by any EQing will affect the sound. I know with the DSP1100 EQ the filters add a lot of phase shift. When I was originally playing with my EQ filters I used the Behringer design software. It nicely displayed the resultant filter curve and phase shift curve. Lots of phase shift came with heavy EQ. I guess since you're independantly feeding the ribbon/Mid/Sub sections it won't matter.

                                                          Thanks for taking the time to fill us in on all the details....good stuff......Don't forget to tell us how it sounds.

                                                          brucek

                                                          Comment

                                                          Working...
                                                          😀
                                                          😂
                                                          🥰
                                                          😘
                                                          🤢
                                                          😎
                                                          😞
                                                          😡
                                                          👍
                                                          👎
                                                          Searching...Please wait.
                                                          An unexpected error was returned: 'Your submission could not be processed because you have logged in since the previous page was loaded.

                                                          Please push the back button and reload the previous window.'
                                                          An unexpected error was returned: 'Your submission could not be processed because the token has expired.

                                                          Please push the back button and reload the previous window.'
                                                          An internal error has occurred and the module cannot be displayed.
                                                          There are no results that meet this criteria.
                                                          Search Result for "|||"