›
Please review the site Rules, Terms of Service, and Privacy Policy at your convenience. Rules, TOS, Privacy
Get familiar with the reaction system: Introducing the Reaction System

Design shortcut. Can I get by with this?

Ron_ERon_E
in DIY
I'm trying to get a 3 way done in time for Grinnell but the driver cutouts and rebates are chewing up lots of time.  These will be ported. What I propose is to get one speakers drivers mounted so I can measure them with un-stuffed boxes and without the port installed. That way I can get the crossover designed and parts ordered.  My assumption is that it matters not whether I measure them with our without the port because it will only affect frequencies well away from the crossover points. I am correct about that?
«1

Comments

  • Ron_ERon_E
    PS It doesn't help that I was using a template with a top bearing to cut a tweeter rebate and the template moved.
  • ani_101ani_101
    Not n issue, for the nigher frequency (over 300) it is good enough. I usually measure without port or stuffing. but if you are flush mounting that needs to be done.
  • 4thtry4thtry
     I'm still struggling to learn the proper design techniques, so this suggestion may be completely wrong. :)   My understanding is that you need to establish the SPL in the 100 to 200Hz region before you can begin to design your crossover.  Then you use your crossover parts to bring everything above these frequencies down to that level.  In this way, your baffle step adjustment will not require extra parts.  So if your port or lack of stuffing has no effect on the SPL in the 100-200Hz region, then you should be OK   
  • ani_101ani_101
    But I can't measure the 100-200. All I can measure is 400 and up. So I match the 400-800 piece between the mid and woofer. 
  • ani_101ani_101
    The Right way should be Far field + near filed merged along with port response as per Jeff's paper and extend the tail.... but short cut gets pretty close, and voicing takes care of the +/- resistor values.Still don't know the right way.... haven't reached that far
  • ani_101ani_101
    BTW, and this a big BTW, I am supposed to learn how to measure and XO design from Ron, sometime....! so, pretty much disregard what I say.
    Nicholas_23
  • dcibeldcibel
    edited October 2017
    Ron_E said:
    I'm trying to get a 3 way done in time for Grinnell but the driver cutouts and rebates are chewing up lots of time.  These will be ported. What I propose is to get one speakers drivers mounted so I can measure them with un-stuffed boxes and without the port installed. That way I can get the crossover designed and parts ordered.  My assumption is that it matters not whether I measure them with our without the port because it will only affect frequencies well away from the crossover points. I am correct about that?
    Port or no port, that doesn't matter. For my ported designs, the low end response is merged from a simulation anyway, since measuring the driver output and port output seperately and properly merging them together is a real pain, and doesn't really give me any information about the speaker that I didn't already know. To that effect, you don't need the impedance to include the port either, since the crossover is up higher in freq which will be the same port or no port.

    As for the stuffing, without the stuffing you should see a strong null in the midbass which can be annoying, so I've tossed a throw pillow in behind the speaker to avoid the reflection. Cramming some stuffing in there doesn't really take much time either, but it is messier than the pillow.

    I do find it useful to mesaure without stuffing so you know where the rear wave reflection appears in the response, to make sure it is completely squashed when you finish the speaker.
    I'm not deaf, I'm just not listening.
  • Ron_ERon_E
    I will flush mount all the drivers before I measure them and Bill's thinking is what I thought too. As far as I know the only thing lining the box with foam and adding a port will effect is the very bottom end of the response. The basic SPL of the woofer remains the same.
  • WolfWolf
    200Hz is known in general as 'the loudest your reference woofer level should be capable of', even if the response increases from there. It is the output point at which mids and tweeters should be matched. I splice in box response below there anyway.

    InDIYana Event Website

  • jhollanderjhollander
    Depends on the woofer.  Some of the have rising response below 200 Hz. A sealed enclosure could roll that off and make your woofer to mid match off.
     John H, btw forum has decided I don't get emails
  • WolfWolf
    It's just in general, but not all cases are this way.

    InDIYana Event Website

  • 4thtry4thtry
    ani_101 said:
    But I can't measure the 100-200. All I can measure is 400 and up. So I match the 400-800 piece between the mid and woofer. 

    That's the rub.  Your FF measurement starts to bounce up and down below 400Hz due to room reflections.  Hence the need to merge NF and FF in an attempt to get a better picture of the SPL level in the 100 to 200Hz region.  The NF measurement needs to be adjusted for baffle step so that when you attempt to merge it with the FF, the slopes match up over an extended range of frequencies.  But the FF measurement is bouncing up and down below 400 Hz, so the merging process is at best a good guess created by striking a line through the center of the FF's bouncing response.    
  • 6thplanet6thplanet
    edited October 2017
    I think you'd want some sort of internal damping control else you might have some squiggles from in box reflection show up in your measurements. Like dcibel said, it doesn't take really any time to throw some FG, foam, or a pillow in there to midigate that.
  • Ron_ERon_E
    These will go right down to the wire. I've got all the rebates cut on one cab which I've lined with foam and I should be in a position to measure it tomorrow.  If things go well I can get the crossover designed and parts ordered before PE closes
    .

    Ron
    kennykNicholas_234thtryJavadSsquamishdroc
  • KillaKilla
    Nice! As far as the 200hz deal goes. If you use Jeff Bs frequency response blender merging far/near field response and including baffle step is a breeze. I don't see many people using this program and it works so well.
  • PWRRYDPWRRYD
    I've been using Jeff's blender and DLR's WinPCD since they came out.  They do work very well.
  • PWRRYDPWRRYD
    Nice looking speaker Ron.  Is that the aluminum Designer series woofer you won?
  • Ron_ERon_E
    Yes, it's the DSA215-8 along with Peerless 830985 mid and da25bg08-06 tweeter. I measured everything and ordered the crossover parts today. They should be here Thursday morning. I use SoundEasy which lets you sim baffle step and merge it with far field measurements. Hopefully I'll in good shape by Friday afternoon.

    Ron


  • Ron_ERon_E
    The mid and tweeter measure pretty well. These measurements were taken without a mic cal file and my mic has a little fall off on the top end.



    Ron
  • Ron_ERon_E
    I spoke too soon.  I hadn't looked at the impedance which turned out to be too low. I cancelled the parts order and I've decided not to try to kill myself trying to get these ready for Grinnell. 

    Ron
  • PWRRYDPWRRYD
    Aw man that sucks!
  • jhollanderjhollander

    :/  I'd bring you parts if I was attending. Maynard could probably set you up as well.

    I think there's a rich tradition of soldering parts in the hotel room.

    JavadSbrek81S7910
     John H, btw forum has decided I don't get emails
  • Ron_ERon_E
    If it was just the crossover I might have persisted but there's some cabinet stuff to work on too and it gets hard for this old guy to spend hours on end working in the shop. One of the problems I'm up against is the mid I chose is 4 ohm and it's efficiency is marginal. I really should have built a test box before I started. I'm looking into using an 8 ohm Fatal Pro mid with 91 db efficiency if it will fit. I might bring one of the pretty boxes along. Really sorry to hear you're not going to make it John.

    Ron
  • Ron_ERon_E
    edited October 2017
    I think I've solved the impedance problem. I reversed the order of the high pass and low pass filters that made up the band-pass on the midrange and the impedance went from under 2 ohms to barely under 4.



    Ron
    kennykPWRRYD
  • ani_101ani_101
    What parts do you need? Maybe I have them and can get them on Friday.
  • Ron_ERon_E
    I'm good Ani.  There are just too many other things that have to be done to the cabinets before they would be ready to go. I have a tweeter rebate that's in the process of being reworked and the ports need to be installed and tuned. I've already made some mistakes by being in a hurry so I'm just going to save them for a later date.

    Ron
  • PWRRYDPWRRYD
    I rushed the project that I took to my first Iowa and it was painfully obvious. 
  • ani_101ani_101
    All my projects are rushed. If it weren't for the DIY gatherings, I wouldn't have completed any... though I have have plenty of speakers, which are sitting in raw MDF!
    PWRRYDhifiside
  • Ron_ERon_E
    I got the ports tuned and installed so now I can compare the sealed measurements with the ported measurements. The upper range of the woofer track pretty well between ported and sealed. The levels don't quite match between measurements but are pretty close. When I dropped the ported response measurements into the sim I developed from the sealed box the results are less than optimum so I will rework the crossover and see if it can be improved. The ported response was very flat until I added the simmed box diffraction

    Crossover sim comparing ported vs sealed FR using crossover designed using sealed measurements

    Ported FR FF, NF, Port, and simmed diffraction

    Sealed FR FF, NF and simmed diffraction


    Ron
  • 4thtry4thtry
    Ron,  thanks for posting detailed info comparing sealed verses ported FR measurements.  This is very interesting.  Would I be correct in concluding, for this design, that it is necessary to install both the port and the stuffing, prior to taking measurements, in order to select the best crossover parts?   Or am I misinterpreting your analysis? 
Sign In or Register to comment.