• RE: Moving the speakers all over the room... - Davey 20:39:14 07/18/10 (8)
  In Reply to: RE: Moving the speakers all over the room... posted by josh358 on July 18, 2010 at 19:20:27

  http://www.audioasylum.com/forums/MUG/messages/15/157974.html

  • http://www.audioasylum.com/forums/MUG/messages/15/157974.htm   (Open in New Window)

  
  

  ---

  • RE: Moving the speakers all over the room... - josh358 06:27:26 07/19/10 (7)
    In Reply to: RE: Moving the speakers all over the room... posted by Davey on July 18, 2010 at 20:39:14

    I'm afraid I'm still missing something. Do you still think I'm confusing your plot of the crossover network characteristics with the acoustical frequency response of the speakers? Because as I pointed out before, I'm not, that would be ridiculous. Rather, I'm comparing the known electrical response of Magnepan's stock crossover network with a guess about the electrical response of PG's network, which as I understand it is a 6 dB/octave two-way series network.
    
    At that point, I can only speculate, because among other things I don't know the topology or values of Peter's network, but I think it's a fair bet that Magnepan designed its network as it did *for a reason* -- acoustical response, electrical interaction, driver offset, baffle cancellation, FBE and speaker wall reflection, desired tonal balance.
    
    As it happens, I see a problem in Varkdriver's measured frequency response *at the crossover point* -- precisely where the complex sum of the electrical response of Magnepan's crossover is up by 3 dB. Coincidence? I can't know for sure is what PG did, so like everything else here, this is just speculation about some of the possibilities Varkdriver might want to look at. But from my perspective, it sticks out like a sore thumb as a possibility that warrants investigation.
    
    I also know from Stereophile's response curve that Magnepan's shaping of the upper crossover network shapes the near field response of the speaker. They didn't go for maximally flat response there. In this case, *the complex sum of the crossover is clearly reflected in the acoustical output of the speaker* as demonstrated by near field measurements. Again, I am inferring that Peter's crossover network doesn't do this. I assume he was going for maximally flat in the midrange, rather than Magnepan's back-in-the-hall perspective.
    

    ---

    • RE: Moving the speakers all over the room... - Davey 07:09:26 07/19/10 (6)
      In Reply to: RE: Moving the speakers all over the room... posted by josh358 on July 19, 2010 at 06:27:26

      "Do you still think I'm confusing your plot of the crossover network characteristics with the acoustical frequency response of the speakers?"
      
      Yes, reading your responses it appeared you were. If not, good.
      
      It's difficult to infer anything about what dips/humps/slopes Magnepan was addressing in the acoustic response of the systems just by looking at the summed electrical responses.
      
      I don't think we can even guess about what type of electrical response is created by the Magnestand 3.6 crossover based on listening position acoustic in-room measurements from just one data source. Way too many variables. We'd need a schematic of the Magnestand 3.6 crossover or reverse-engineer an existing one to create a schematic. "Satie" appears to have super-human speculative capabilities in this regard so maybe he has it figured out. :)
      
      My belief is that Magnestand's approach to crossover design is pretty much completely seat-of-the-pants, and there are no acoustic measurements that show how well they achieved a "target" response...whether that target be flat or bumped midrange or whatever. The results are judged strictly by listening. Nothing wrong with that approach, but it can be a real can of worms.
      
      Cheers,
      
      Dave.
      

      ---

      • Speculative reasoning - Satie 08:49:44 07/20/10 (4)
        In Reply to: RE: Moving the speakers all over the room... posted by Davey on July 19, 2010 at 07:09:26

        I use rational boundary setting techniques and proportionation within the bounds in all my speculations. These are the basics of rule of thumb engineering, which precede the actual formulation of hypothesis and the design of experiments, and even precedes researching the science to any significant depth.
        
        This is part of a deductive and inferential process necessary to the production of a useful plan of investigation, action, and for preliminary observations and conclusions. Read some Ludvig von Mises on Epistemology of science and economics and it will give you some idea of how I think.
        
        E.g. If PG takes the trouble to say he is doing something in particular, and once he starts producing he does not alter his statements, we have a basic boundary describing the architecture of the crossover and the general approach toward his mod's design. Where he roughly places his crossover's 3db points is determinable by the slopes measured for a crossovered driver FR relative to the FR of those that are not crossed over. Where there is a problem is in the bass rolloff. But considering how the bass is tilted, dips and peaks included, it seems that the bass rolloff is earlier than in the stock crossover. Etc...
        
        BTW Varkdriver's backwave quarter wave and half wave cancellation frequencies from the 42" placement are in the 80 hz and 160 hz areas, and if he toes in the speakers, they would be more shallow and spread down in frequency by a few hz. Meaning that there is something else at play.
        
        
        

        ---

        • RE: Speculative reasoning - Davey 11:53:23 07/20/10 (3)
          In Reply to: RE: Speculative reasoning posted by Satie on July 20, 2010 at 08:49:44

          Well, I'm just a dummy that fell off the turnip truck, but I'm smart enough not to base any speculation on some sort of "description" given by Peter Gunn.
          
          Some data analysis folks have an interesting credo that goes......"give me one data point and I'll extrapolate to produce whatever data you want." It seems to me we have a similar situation here.......except the one data point is invalid....or highly questionable. :)
          
          Since you're concluding something meaningful from one invalid data point and a couple of non-descriptive photos you're much smarter than I am.
          
          Cheers,
          
          Dave.
          

          ---

          • Confirmation - Satie 13:04:06 07/20/10 (2)
            In Reply to: RE: Speculative reasoning posted by Davey on July 20, 2010 at 11:53:23

            The requirement of confirmation and definite certainty is the bane of knowledge, a waste of time, a nagging of Nabob and something you wish you did when your project blows up.
            
            I am willing to speculate on his consistency, since he never seems to deviate from what he thinks things should be. Totally inflexible on his work and offering only a choice of woods for options.
            
            Then, I saw what he did in the photo. It is very obvious that the crossover photo confirms that he did what he said he would. The parts count is telling, despite the paralleling of caps.
            
            Despite his drawbacks and doctrinaire inverse understanding of the physics, I respect what he does, both on his efforts to communicate his views and describe his work and most of all on the many positive testimonials on this forum for his work on MMG and 1.6. I do fault him for the following:
            
            1. Insisting on adhering to his MMG mod structure rather than thoroughly thinking through what would work best for the 3.6 as a three way speaker.
            2. Insisting on series crossovers and not building triampable speakers no matter what his client wants.
            3. Hiding out in silence instead of admitting a mistake and announcing the "doing it better" solution (as in an actual 3 way crossover to produce a successfully flat inroom response where needed, without comb effects).
            4. Insisting on tilting the speakers. This does not work for the bigger maggies - like the 3.6.
            

            ---

            • RE: Confirmation - Davey 17:53:41 07/20/10 (1)
              In Reply to: RE: Confirmation posted by Satie on July 20, 2010 at 13:04:06

              How 'bout drawing a schematic diagram so the rest of us can see what the crossover looks like?
              
              Thanks,
              
              Dave.
              

              ---

              • RE: Confirmation - Satie 21:08:05 07/20/10 (0)
                In Reply to: RE: Confirmation posted by Davey on July 20, 2010 at 17:53:41

                Too much effort,
                So varkdriver can read the values off his crossover component and write up the connections and what cap values are run in parallel and how they are connected to the driver posts.
                
                That said, here are two options:
                
                Since this would be a quasi 2 way design as per PG's intention, it does not really have an entirely obvious design but my initial guess would be a first order series 2 way with the bass and tweeter in parallel to a first order high pass to the midrange. This one would tend to emphasize the midrange more.
                
                A second guess is the bass and mid in a series 2 way crossover, with the tweeter and second high pass run in parallel to the midrange driver. This would tend to depress the tweeter, but not necessarily elevate the mids relative to the bass. Here the tweeter will appear to have a second order crossover below the mid high pass.
                
                

                ---
      • RE: Moving the speakers all over the room... - josh358 10:47:26 07/19/10 (0)
        In Reply to: RE: Moving the speakers all over the room... posted by Davey on July 19, 2010 at 07:09:26

        Satie I think was referring to PG's earlier description of his intentions regarding his 3.6 crossover, which I remember having read as well. I don't remember all the specifics but PG did talk about running the 3.6's as a two way system. Satie also looked at the crossover picture and noted that there was only one inductor, and that there were only four connections, and made some inferences about topology on that basis.
        
        Anyway, the point is that Satie isn't just pulling this stuff out of his hat. But I agree that there's a certain amount of speculation without the actual crossover values, or a measurement of the crossover response. My own goal was to try to offer some educated guesses about where Varkdriver should look, given that he reported having preferred the sound of the speakers in their original configuration to the new one.
        
        Have you looked at the Stereophile response measurements, BTW? From 1000-10,000 Hz, the overall shape of the dip in the summed nearfield response follows your plot pretty damn closely. The point being that I wasn't just looking at the summed electrical response in your post -- I comparing it to Stereophile's summed nearfield acoustical response. I also noticed that in the nearfield, the response in the 100-200 Hz region wasn't depressed, though it was far from flat even after taking into account the 6 dB octave bass boost. I put that together with the practical limitations of a two-way crossover to make some practical guesses about whether the crossover was an area of concern.
        

        ---

Post a Followup:

This post is made possible by the generous support of people like you and our sponsors:

Atma-Sphere Music Systems, Inc.

---

[ Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2025, Audio Asylum, all rights reserved ]

---

[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]

---