Announcement

Collapse
No announcement yet.

PCD 7: Help a Newb out

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

  • PCD 7: Help a Newb out

    Hey Everyone!

    I have decided to jump head first into the deep end and learn as much as I can about speakers, x-over design, etc. I am not new to the audio world and have been designing subwoofers for a while now, but speakers are a different beast. I do have a background in physics, so most of the acoustic and electrical terms/basic ideas are not new thankfully.

    So I started off by reading through the "Speaker Building Bible" sticky and have literally read all of the posts so I decided to jump in and see what happens. To start off with I thought I would try and replicate someone else's design to make sure I am modelling everything correctly. I choose Paul Carmody's Overnight Sensations to start off with.

    I used Jeff Bagby's "Response Modeler" to create simulated frd and zma file's and extract the minphase for the given box/baffle size etc. I pulled the frd and zma data from PE and followed the tutorial written by Paul Carmody on his google site for modeling everything.

    Then I imported everything into PCD 7 and input the x-over values to see what I came up with.

    Here is where I'm not sure if I did everything correctly. The response I get, while similar, is not nearly as flat as the one shown on Paul's website. I've attached an image of the one from Paul's website and the one that I created in PCD.

    Click image for larger version

Name:	OvernightSensationSystemResponse.jpg
Views:	1
Size:	95.0 KB
ID:	1193293
    Click image for larger version

Name:	OS response chart practice.jpg
Views:	1
Size:	64.5 KB
ID:	1193294

    Any idea's why I have such a large hump from ~1200hz -3,000hz? Also, the peak at ~4600hz is not in Paul's graph either. The phase looks off too. I thought I had modeled the baffle loss and diffraction wrong in Response Modeler so I went back and re-ran all the sims but came out with the same result... Also, I'm not worried about the low end, I was having issues getting the ported box model to splice correctly so I decided to just leave it off.

    I am really enjoying this so far and hope that I can model a speaker prior to building it and have it measure the nearly the same as the model. Any help would be appreciated. Thanks.

  • #2
    Re: PCD 7: Help a Newb out

    I've never worked with the ONS so I know nothing of their xovers, and I've never used Jeff's response modeler but it does sound like a fun program. All that said... try reversing the polarity going to the tweeter? That's where I'd start. But it also looks like that hump is in the woofer, so I'd also play with the values in the xover components. It's hard to say though w/out knowing what you have there for a xover now.

    Either way, looks like you're having some fun and that's a good idea to learn IMO.
    "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
    "Audio is surely a journey full of revelations as you go" JasonP

    Comment


    • #3
      Re: PCD 7: Help a Newb out

      Originally posted by mattsk8 View Post
      I've never worked with the ONS so I know nothing of their xovers, and I've never used Jeff's response modeler but it does sound like a fun program. All that said... try reversing the polarity going to the tweeter? That's where I'd start. But it also looks like that hump is in the woofer, so I'd also play with the values in the xover components. It's hard to say though w/out knowing what you have there for a xover now.

      Either way, looks like you're having some fun and that's a good idea to learn IMO.
      I did need to reverse the polarity on the tweeter, missed that part. But like you said that didn't help the hump in the mid response. My goal is not to alter the speakers x-over, just understand why my response looks different from how it's supposed to. That way I will know that I am configuring everything correctly in the program.

      Comment


      • #4
        Re: PCD 7: Help a Newb out

        Originally posted by squiers007 View Post
        I did need to reverse the polarity on the tweeter, missed that part. But like you said that didn't help the hump in the mid response. My goal is not to alter the speakers x-over, just understand why my response looks different from how it's supposed to. That way I will know that I am configuring everything correctly in the program.
        Just so I understand what you did... You imported the factory measurements for the drivers into Jeff's program, then you input the actual xover values for the ONS into the same program to see if the response matched?
        "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
        "Audio is surely a journey full of revelations as you go" JasonP

        Comment


        • #5
          Re: PCD 7: Help a Newb out

          Originally posted by mattsk8 View Post
          Just so I understand what you did... You imported the factory measurements for the drivers into Jeff's program, then you input the actual xover values for the ONS into the same program to see if the response matched?
          Yes and No. I imported the factory measurements (that I obtained from PE) into another program first called "Response Modeler". I used this program to create simulated frd and zma files that accounted for baffle step and diffraction (using the dimensions of the OS). These are the files I imported into PCD. Then I entered the X-over values to see what I got.

          One thing that I thought might be throwing my measurements off was that it does not say whether the frd measurements that I got from PE were done on a baffle or anechoically (is that even a word)? So I'm thinking if they were done on a baffle I would actually need to subtract this effect first (to obtain the anechoic data), then model the actual OS baffle... hopefully that made sense.

          Comment


          • #6
            Re: PCD 7: Help a Newb out

            Originally posted by squiers007 View Post
            Yes and No. I imported the factory measurements (that I obtained from PE) into another program first called "Response Modeler". I used this program to create simulated frd and zma files that accounted for baffle step and diffraction (using the dimensions of the OS). These are the files I imported into PCD. Then I entered the X-over values to see what I got.

            One thing that I thought might be throwing my measurements off was that it does not say whether the frd measurements that I got from PE were done on a baffle or anechoically (is that even a word)? So I'm thinking if they were done on a baffle I would actually need to subtract this effect first (to obtain the anechoic data), then model the actual OS baffle... hopefully that made sense.
            That made sense. I think you are correct about the baffle consideration, I have no idea what they used when they measured the drivers. And for more clarity, you imported the manufacturer response graphs (< that's a link) that you got on the PE forum... right? Because if so, you're absolutely correct, I can see the hump you're trying to deal with in that response. And unfortunately I don't think this would work for you to do what you're trying to do, unless you had the actual in baffle measurements. Putting the drivers in the baffle definitely changes the response quite a bit from the manufacturer's response.

            If there's a way around that, maybe someone else could chime in. I think Ben (Wolf) might know a thing or 2 about how to do this if it's possible, but unfortunately I don't.
            "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
            "Audio is surely a journey full of revelations as you go" JasonP

            Comment


            • #7
              Re: PCD 7: Help a Newb out

              Originally posted by mattsk8 View Post
              That made sense. I think you are correct about the baffle consideration, I have no idea what they used when they measured the drivers. And for more clarity, you imported the manufacturer response graphs (< that's a link) that you got on the PE forum... right? Because if so, you're absolutely correct, I can see the hump you're trying to deal with in that response. And unfortunately I don't think this would work for you to do what you're trying to do, unless you had the actual in baffle measurements. Putting the drivers in the baffle definitely changes the response quite a bit from the manufacturer's response.

              If there's a way around that, maybe someone else could chime in. I think Ben (Wolf) might know a thing or 2 about how to do this if it's possible, but unfortunately I don't.
              Yeah, that's the data I used. If they stated the baffle size I could input that into the response modeler program and remove the baffle effect, but unfortunately I don't see anything on there about that...

              Comment


              • #8
                Re: PCD 7: Help a Newb out

                Originally posted by squiers007 View Post
                Yeah, that's the data I used. If they stated the baffle size I could input that into the response modeler program and remove the baffle effect, but unfortunately I don't see anything on there about that...
                Hopefully Ben sees this thread because I'm pretty sure he knows how to do what you're trying to do. I'd love to learn too because you could model a bunch just for fun without spending all the money
                "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
                "Audio is surely a journey full of revelations as you go" JasonP

                Comment


                • #9
                  Re: PCD 7: Help a Newb out

                  Originally posted by mattsk8 View Post
                  Hopefully Ben sees this thread because I'm pretty sure he knows how to do what you're trying to do. I'd love to learn too because you could model a bunch just for fun without spending all the money
                  That's the plan!! Design a solid speaker without having to buy the drivers and x-over parts. Obviously I would tweak the design once I had built them (if necessary), but if this method works you should be able to get really close without having to actually build anything.

                  Comment


                  • #10
                    Re: PCD 7: Help a Newb out

                    "8 hours to Ben ?"

                    Comment


                    • #11
                      Re: PCD 7: Help a Newb out

                      In pcd, did you enter the x/y/z offsets that Paul used? If not, that will throw off the response and phase.
                      https://www.facebook.com/Mosaic-Audi...7373763888294/

                      Comment


                      • #12
                        Re: PCD 7: Help a Newb out

                        Originally posted by isaeagle4031 View Post
                        In pcd, did you enter the x/y/z offsets that Paul used? If not, that will throw off the response and phase.
                        Yes, but I had to guess a little on the z-offet because I was not sure where exactly the acoustic center of those drivers was. ..

                        Comment


                        • #13
                          Re: PCD 7: Help a Newb out

                          Originally posted by Chris Roemer View Post
                          "8 hours to Ben ?"
                          If I understand what "8 hours to Ben" is supposed to mean, I don't see why that needed to be said. Like I said, I've never done it so I have no idea what's involved. I always get the drivers, measure them, then import the actual in-baffle measurements. And I know that huge hump he's dealing with generally smooths out once they're in baffle. Wasn't trying to sign anyone on to design anything for anyone, just seeing if someone was willing to explain how, since I can't. Kinda figured that was maybe part of what this forum was for.
                          "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
                          "Audio is surely a journey full of revelations as you go" JasonP

                          Comment


                          • #14
                            Re: PCD 7: Help a Newb out

                            Didn't mean anything by that. Just giving the OP a "heads up" as to when he might expect an answer.
                            It's now about B-2.

                            Comment


                            • #15
                              Re: PCD 7: Help a Newb out

                              Originally posted by Chris Roemer View Post
                              Didn't mean anything by that. Just giving the OP a "heads up" as to when he might expect an answer.
                              It's now about B-2.
                              Gotcha. Sorry, I totally misunderstood what you said.
                              "The ability of any system to produce exceptional sound will be limited mainly by the capability of the speakers" Jim Salk
                              "Audio is surely a journey full of revelations as you go" JasonP

                              Comment

                              Working...
                              X