Announcement

Collapse
No announcement yet.

TN Allen's waveguide measurement thread - picture/technical heavy

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

  • TN Allen's waveguide measurement thread - picture/technical heavy

    TN Allen mentioned in some thread that he had some waveguides he'd like tested, but didn't have the equipment or know how. So I offered because I've been on a waveguide hunt for a while now. I've used the Dayton 8" waveguide for a few months, so that is where some comparison in this thread comes from.

    First off, a picture of a couple of the waveguides and an 8" by 10" baffle with a 1/2" chamfer all the way around. It was a sloppy fit, just a quick jobby with a jig saw.



    The following measurements were done with an SB29RDCN-0004, a Dayton ND28F-6, and a Tang Band 28-847SE. Each waveguide was measured on axis in with each tweeter. If that was quite poor, then I moved on quickly. If it was ok, I measured at roughly 22.5 and 45 degrees. I measured at a fairly low volume at 0.6m.

    Each waveguide was assigned a letter A through E. The waveguide dimensions were quickly measured with a tape. Dimensions were all fairly similar, with 2 waveguides having a narrower throat and straight profile, and 3 having a wider throat and contoured profile. The overall diameter was approximately 165mm (6.5"). Very nice dimensions for aligning acoustic centers and matching diameter with a common woofer.

    Here we go. First up are three nice measurements with the ND28.



    And here is the same thing with a 6.8uF cap in series. Almost a perfect LR4 at 1100hz, but that's where Fs is too, so more would be needed ;)



    Next we have what I thought was the nicest of the 5 with off axis measurements. (Sorry about the reflections out at 45 degrees).

    SEE POST BELOW WITH BETTER MEASUREMENTS

    The top octave here, looks very similar to what you get with this tweeter flush on the same baffle.

    Moving on to the SB. I missed waveguide B for some reason, but previously measured it to be quite similar to A.



    And with the 6.8uF cap.



    Upper octave garbage doesn't bother me, but it is a concern. This tweeter does that in the Dayton waveguide also.

    And here we have the off axis stuff.

    SEE POST BELOW WITH BETTER MEASUREMENTS

    Finally the tang band. It was ok.



    This tweeter was insensitive to the waveguides with only mild gains in the low end.

    Just for kicks, here is the Dayton ND28 in the Dayton 8" waveguide.



    And this is a couple bad results.



    Personally, I think A and E are the nicest. I think I'd take A. I plan to do some more rigerous testing with A. I would like 3 of them :D With how cheap the ND28 is, or how good the SB is, a fine system can be made with these. C and D were not very good. A, B and E were all similar. I also think A, B, and E could accept different tweeters, because the repsonses were similar accross 3 different tweeters.

    I have plenty more measurements I can post, such as other off axis, bad measurements, SB in the Dayton 8" waveguide, etc. I can also try measuring with different filters if anyone is interested. I really think TN Allen has milled some great waveguides. I will be begging him to cut me some from this: http://www.ebay.ca/itm/Maple-Burl-Lu...item5d31b5aae9

    Feedback is appreciated. Look for Arlis' measurements coming soon also, with the RS28 I believe. Once they've been measured thoroughly, TN Allen would like to get the x-section out to the public to start making these things, so input is needed now.
    Last edited by ryanbouma; 01-05-2012, 04:17 PM.
    https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

  • #2
    Re: TN Allen's waveguide measurement thread - picture/technical heavy

    Thanks for taking this on! There are some important items that I didn't see in your write-up.

    What were your measurement conditions? Were the baffles mounted in a box and measured stand mounted or was another technique used to remove any baffle effects?

    It would be very helpful to include each drivers response on a flat baffle so it could be discerned what effect the waveguides had on the response.

    The throat to driver interface is extremely important in my experience and I didn't see anything mentioned about that. Were all the throat diameters of the waveguides the same? What was that diameter?

    If all were the same, what adjustments were made to make them fit each tweeter, which I assume had slightly different diameters?

    Where did the throat fall relative to the driver plus surround?

    Where face plates left on the drivers or removed. If left on, was there any attempt to fill gaps that might have been present due to face plate contouting?

    Any idea what the rough angle of the throat was on each waveguide?

    What was the depth of each waveguide?
    Dan N.

    Comment


    • #3
      Re: TN Allen's waveguide measurement thread - picture/technical heavy

      Originally posted by dlneubec View Post
      Thanks for taking this on! There are some important items that I didn't see in your write-up.
      Hi, I anticipated most of these questions ;)

      The first thing I did was measure the throat and depth dimensions. I plan on posting those here, but need to make a table. Coming.

      As for the flush responses, here is the Dayton ND28 on a very similar baffle.



      And here is the SB in the Dayton 8" waveguide, but you'll have to look at the manufacturer's spec sheet for the flush response, I haven't done it.



      Note the peak around 1500hz for the ND28, that shows up in the waveguide response also I believe.

      Nothing was done to seal the tweeter to the throat or fill in gaps. The were just soft fitted and centered as best I could. The end user could decide to fuss over that. I wouldn't. They were a decent fit anyways. Well, the SB had a little overhang iirc, I'll check on that when I re-measure A or E.

      I agree with your experience that the throat is very important. I'm unsure of the angles, perhaps TN Allen knows the angles. I can attempt to measure the angle. I'll ping TN Allen first though. I'll also try and get a nice photo of it.

      I think my table of dimensions will help answer these questions better.

      As far as measurement conditions, the waveguide was soft fitted into the baffle. As meantioned in my OP, sloppy. This was just a preliminary run through. I think diffraction is showing up in the measurements. A roundover would likely be better than a small chamfer. These were NOT in a box, although I have a box for this baffle if you think that would help. A large piece of fiberglass insulation was placed behind the baffle. Elevated about 1m above the floor. Measure about 0.6m from the baffle surface. All were highpassed at 500hz, and gated just under 1000hz for extra safety.

      Next time I measure A and/or E, I plan to get out around 1m and get outside. Or at least clean up the living room and get further from boudaries. Lots of Christmas junk around the house ;)
      https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

      Comment


      • #4
        Re: TN Allen's waveguide measurement thread - picture/technical heavy

        Oh, and I did remove the faceplate from the ND28, and it was a perfect tight fit with waveguides C and D, but the results were much the same, and were no good. For A B and E, they didn't fit well at all.
        https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

        Comment


        • #5
          Re: TN Allen's waveguide measurement thread - picture/technical heavy

          On waveguide A, if you were to add a small coil to make a simple 2nd order electrical filter, the response could probably be made level on down to 2k, and then roll off. I'd try something around .4 mH to .6 mH, give or take. Also, if you want to tame the peak on the top end a little, my spreadsheet shows a 1 uf cap in series with a .1 mH coil, and about a 6 or 7 ohm resistor. I'm guessing on the resistor value. Note... Added a couple notch measurements below.


          Originally posted by ryanbouma View Post
          And with the 6.8uF cap.



          Here's a non waveguide tweeter measurement where I added two different notches to show the effect on my dome tweeter. It looks to me like the green trace is the one that would work. With a bigger resistor, the effect can be lessened.

          Comment


          • #6
            Re: TN Allen's waveguide measurement thread - picture/technical heavy

            Thanks for the work. Thsi stuff can take significant time. I appreciate it.

            I am unsure if what I am seeing is a good thing or not. Trying to look at it as if the tweeters came that way off the shelf, I would consider them as any other for a specific purpose. The WG's did not seem to reinvent the wheel as far as response goes, other than the bennies of the known lower end improvement in a couple of flavors.

            Makes a good argument for using larger mids or crossing to a larger woofer.

            Now: Looking at 30,000 feet: The added complexity. What am I gaining by using a WG in a 6" bookshelf, or even a standard 3-way without having the mid in a WG which would have to be enormous.

            Makes me wonder vs. a small tweeter array? Short line source?

            This is good data. I am just wondering how it compares to other options and is it USEFUL data other than showing the normal advantages of a WG?

            Do you have any impedance measurements?

            I am starting to wonder how a Scan 2604/833000 with a Fs of 475 and 94dB eff. would do in a moderately sized WG crossed to a 10" pro woofer directly. Just what I needed...ANOTHER design to start bouncing around in my head... Thanks a lot!
            .

            Comment


            • #7
              Re: TN Allen's waveguide measurement thread - picture/technical heavy

              Thanx for the invaluable measurements. Roman recently did a DIY small guide with the VIfa DX25 that suggested by listeners did extremely well. This tweeter known for crappy HD products below 2khz seems to be a great candidate for larger waveguides. The cheap price, high power handling and efficiency really sweeten the pot!

              Comment


              • #8
                Re: TN Allen's waveguide measurement thread - picture/technical heavy

                Originally posted by rpb View Post
                On waveguide A, if you were to add a small coil to make a simple 2nd order electrical filter, the response could probably be made level on down to 2k, and then roll off. I'd try something around .4 mH to .6 mH, give or take. Also, if you want to tame the peak on the top end a little, my spreadsheet shows a 1 uf cap in series with a .1 mH coil, and about a 6 or 7 ohm resistor. I'm guessing on the resistor value.
                I actually considered that. I'll leave that sort of thing up to the end user. The notch on the top end doesn't concern me. I'd just use the DSP in my avs to deal with it. Also, top octave stuff has never concerned me nearly as much as 1 to 3khz stuff ;) You're absolutely right though, that response could look really nice.

                As for using the SB on this waveguide, I feel it's a little bit wasted. The Dayton 8" for $10 gives a killer response. Only drawback is the cheap look of it, and large size. These are a nice size, and if someone can mill them into a baffle, or from solid (like I would like to have done), then they'll be very sweet.
                https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

                Comment


                • #9
                  Re: TN Allen's waveguide measurement thread - picture/technical heavy

                  Originally posted by mzisserson View Post
                  Thanks for the work. Thsi stuff can take significant time. I appreciate it.
                  No kidding. And I feel like I've only scratched the surface with these. TN has mentioned sending more. And I plan to try more tweeters and beat the s&*t out of these for a while more.

                  Originally posted by mzisserson View Post
                  I am unsure if what I am seeing is a good thing or not. Trying to look at it as if the tweeters came that way off the shelf, I would consider them as any other for a specific purpose. The WG's did not seem to reinvent the wheel as far as response goes, other than the bennies of the known lower end improvement in a couple of flavors.
                  It's not really meant to reinvent the wheel. Only find out what these specific waveguides can do. TN Allen does not consider them complete yet. Although I'm trying to keep this technical and unbiased, I actually am starting to believe these are complete and could be a finished product. Soon we'll know. What it also does is show that a $12 dome tweeter that normally crosses around 2500hz can now cross at 1500hz.

                  Originally posted by mzisserson View Post
                  Makes a good argument for using larger mids or crossing to a larger woofer.
                  For me, that basically is the argument. You can now go bigger and get higher sensitivity, with very cheap parts. That's a personal reason, there are other technical reasons.

                  Originally posted by mzisserson View Post
                  This is good data. I am just wondering how it compares to other options and is it USEFUL data other than showing the normal advantages of a WG?
                  It's not that it shows anything compared to the normal advantage of a WG, rather that this homemade waveguide works like a waveguide should. And TN Allen intends to just throw the x-section out there to the public, for you and I to consume. What a deal!!!!!!!!

                  Originally posted by mzisserson View Post
                  Do you have any impedance measurements?
                  No. Would you like some. I could get some.

                  Originally posted by mzisserson View Post
                  I am starting to wonder how a Scan 2604/833000 with a Fs of 475 and 94dB eff. would do in a moderately sized WG crossed to a 10" pro woofer directly. Just what I needed...ANOTHER design to start bouncing around in my head... Thanks a lot!
                  Ha ha ha. I had that same thought about 6 months ago. I ended up with something a tad different. See my recent Challenger thread in the Project Gallary section. You'll do it, I'm sure ;) Let us know how that goes.
                  https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

                  Comment


                  • #10
                    Re: TN Allen's waveguide measurement thread - picture/technical heavy

                    Originally posted by Mayhem13 View Post
                    VIfa DX25
                    Hmm, I'll have to take a look at that tweeter, and possibly snag a couple.
                    https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

                    Comment


                    • #11
                      Re: TN Allen's waveguide measurement thread - picture/technical heavy

                      Originally posted by ryanbouma View Post


                      It's not really meant to reinvent the wheel. Only find out what these specific waveguides can do. TN Allen does not consider them complete yet. Although I'm trying to keep this technical and unbiased, I actually am starting to believe these are complete and could be a finished product. Soon we'll know. What it also does is show that a $12 dome tweeter that normally crosses around 2500hz can now cross at 1500hz.



                      For me, that basically is the argument. You can now go bigger and get higher sensitivity, with very cheap parts. That's a personal reason, there are other technical reasons.



                      It's not that it shows anything compared to the normal advantage of a WG, rather that this homemade waveguide works like a waveguide should. And TN Allen intends to just throw the x-section out there to the public, for you and I to consume. What a deal!!!!!!!!



                      No. Would you like some. I could get some.



                      Ha ha ha. I had that same thought about 6 months ago. I ended up with something a tad different. See my recent Challenger thread in the Project Gallary section. You'll do it, I'm sure ;) Let us know how that goes.
                      I will check out the Challanger. I have never been a huge fab of WG's. However I am slowly starting to see some light now that I am understanding their purpose.

                      I agree with your "argument" on wg's. Like it even. I also like the way you have it in perspective and are realistic about it vs. some around here who think it is the answer to cure world hunger. It makes more sense when some is explaining things in a fairly non-biased manner, showing both the good and bad, offering work around and ideas with them.

                      I have a pair of 83300's in the Fatbellys'. I may have to experiment. The consensus seems to be from those who have heard them on the forum they do offer a bit more resolution and smoother presentation than the traditional DX 25. But one they cost as much as a pair of DX 25's. Could get interesting! A WG woudl push the Fs down on a DX25, and that is the largest advantage of the 833000.

                      Yes, I would like to see impedance measurements. With quite a few data points if possible.
                      .

                      Comment


                      • #12
                        Re: TN Allen's waveguide measurement thread - picture/technical heavy

                        Originally posted by ryanbouma View Post
                        I actually considered that. I'll leave that sort of thing up to the end user. The notch on the top end doesn't concern me. I'd just use the DSP in my avs to deal with it. Also, top octave stuff has never concerned me nearly as much as 1 to 3khz stuff ;) You're absolutely right though, that response could look really nice.

                        As for using the SB on this waveguide, I feel it's a little bit wasted. The Dayton 8" for $10 gives a killer response. Only drawback is the cheap look of it, and large size. These are a nice size, and if someone can mill them into a baffle, or from solid (like I would like to have done), then they'll be very sweet.
                        The x-over an a waveguide can be a pain. If adding a coil does what I think, it would make the waveguide more practical since x-over would be inexpensive and straight forward. Some of my filters have 7 or 8 parts. If you have a some coils in your parts box, try it and see.

                        Comment


                        • #13
                          Re: TN Allen's waveguide measurement thread - picture/technical heavy

                          If the opportunity arises, sure. You don't sim, but most here do. A trace of my plot and the manufacturer's impedance plot would allow someone interested to check hundreds of possible filters, against any woofer in their data base. For you, I'll try and sneak in a 0.45mH shunt coil ;)

                          First priority is to define its performance with measurements bare.
                          https://www.youtube.com/channel/UCm2...oSKdB448TTVEnQ

                          Comment


                          • #14
                            Re: TN Allen's waveguide measurement thread - picture/technical heavy

                            With the acoustic offsets already done for you by the depth of the guide, a 2nd order filter can be accomplished with a single component........provided the rest of the response is ok. That's where the fit up comes into play and really at this junction, it's truly plug and play only. Some tweeters just don't work well in guides. Ring radiators seem to do very well if the wide surround can be matched to thE throat well. The 'phase plug' seems to produce excellent off axis results....the Vifa XT comes to mind. It too isn't that clean down low and could be greatly improved with waveguide loading.

                            Comment


                            • #15
                              Re: TN Allen's waveguide measurement thread - picture/technical heavy

                              Originally posted by ryanbouma View Post
                              If the opportunity arises, sure. You don't sim, but most here do. A trace of my plot and the manufacturer's impedance plot would allow someone interested to check hundreds of possible filters, against any woofer in their data base. For you, I'll try and sneak in a 0.45mH shunt coil ;)

                              First priority is to define its performance with measurements bare.
                              There's no need to post the response if it's not close, but if it puts the response close to a desirable rolloff, you could simply confirm that the filter will be simple.

                              Comment

                              Working...
                              X