First project (revised)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I factored in the baffle response and removed the resistor. How's it now? :D


I am not so sure that you accounted for the baffle step. An spl of 90 dB/1m at about 1kHz radiating 2pi and with a standard input voltage of 2.83V would have dropped a couple of dB after introducing bsc. Perhaps I am missing something. I do see there is means in VituixCAD to account for it but I have not tried it yet.
 
That looks good. You just have to eliminate that dip, by adjusting your lowpass.

You could also add a small value cap (0.1uF to 0.5uF) parallel to low pass inductor, for a steeper rolloff of the woofer.

Now tweak your lowpass values a bit on the woofer to eliminate the dip at 1.2kHz, and you'll have a solid starting point.

Later,
Wolf

Yes you have come quite a way in a fairly quick time :up:

I am not so sure that you accounted for the baffle step. An spl of 90 dB/1m at about 1kHz radiating 2pi and with a standard input voltage of 2.83V would have dropped a couple of dB after introducing bsc. Perhaps I am missing something. I do see there is means in VituixCAD to account for it but I have not tried it yet.

Should I revert to how it was in post 13? I'm not sure I understood post 16 properly because it seems the dip isn't what I want :p
 
I quickly tweaked your crossover a bit . You can further adjust the values to your liking.
 

Attachments

  • Untitled.png
    Untitled.png
    487.8 KB · Views: 141
Just another Moderator
Joined 2003
Paid Member
Having a closer look I agree with lojzek that the baffle step doesn't appear to have been added in. I had a play with your drivers a week or so ago, but didn't post because you seemed to be working your way through it. with the baffle step factored in the woofer is around 85db not 90db.

I went a bit further and also adjusted for a vented enclosure and came up with the attached. for me the tweeter is the problem, with quite a ragged high end response. I'll put your crossover into my sim when I get home and see how it compares to what you posted, using my baffle adjusted frd.

lojzek, there is a tool in vituixcad to model the baffle response, but you have to export the frd and then use that new one in the sim. No feature to just put in the baffle effects directly.

edit: I'll also do mallikreddyk's as it looks better to me than the one I have attached :)
 

Attachments

  • dayton_sim.png
    dayton_sim.png
    97.5 KB · Views: 145
Just another Moderator
Joined 2003
Paid Member
This is what I get using the attached baffle adjusted woofer response with the crossover from post #17.

Note that the attached frd was created by first subtracting (division) The 20L closed box sim of the driver response from the factory traced curve, then adding (multiplying) by the 10L vented response sim to the flattened factory curve.

I then used vituixcad to simulate the baffle step with baffle dimensions of 227mm wide by 280mm high, and exported the response, which is the zip file attached.

The last attachment shows the comparison between the origin dayton frd (blue), the same frd with the 227 X 280mm baffle diffraction added (black) and the adjusted for reflex alignment curve with the 227 X 280 baffle diffraction added.

Tony.
 

Attachments

  • dayton_sim2.png
    dayton_sim2.png
    108.3 KB · Views: 139
  • dayton_rs150_vented_baffle+0.zip
    5.4 KB · Views: 44
  • dayton_box_model.png
    dayton_box_model.png
    107.1 KB · Views: 132
  • dayton_diffraction.png
    dayton_diffraction.png
    62.5 KB · Views: 133
  • diffraction_compare.png
    diffraction_compare.png
    25.9 KB · Views: 73
Note that the attached frd was created by first subtracting (division) The 20L closed box sim of the driver response from the factory traced curve, then adding (multiplying) by the 10L vented response sim to the flattened factory curve.

WinterMute,

Are these functions available in VituixCad, or did you do the calculations with a different tool? I am very much interested to learn how you did that. and the reasons to pick 20L for subtraction
 
Just another Moderator
Joined 2003
Paid Member
Your welcome Lojzek! :)

mallikreddyk, the functions are available in VituixCad with the calculator tool, but I have to confess that I find it a bit counter intuitive (I probably just need to read the manual or watch one of Kimmo's excelent videos) so I actually did it in speakerworkshop.

It was an idea I had recently. I thought to myself, if we can add and subtract baffle effects from a curve, why couldn't we subtract and add box alignment as well. Some manufacturers actually state the baffle type (often IEC or Infinite) that the measurement was taken on, as well as the box size that the driver was in. I remember seeing a few that said that they used a 20L box, so I figured that may be a somewhat defacto standard in the industry.

Next I modeled the driver in vituix cad in a 20L box and exported the SPL curve.

I also did a 10L sealed one which is roughly a 0.707 alignment for that woofer.

I imported both of these into speaker workshop and overlayed them to see what matched best and decided the 20L one seemed close enough. Note that to get the best match I had to scale the 20L sim up by 0.8db. See first attachment.

I also simmed the 10L reflex alignment and exported that, and imported into speakerworkshop. Note that if you adjust the first one you need to adjust the 2nd by the same amount, so I increased it by 0.8db

Next I made a copy of the original dayton measurement and then used speaker workshops calculate/combine menu to first divide by the 20L sealed response, and then multiply by the 10L vented response. (attachments 2 and 3) This gives the new response in attachment 4.

I exported the new vented alignment response, and used that in the baffle simulation in vituix cad as the half space response (attachment in my previous post). That response was then exported, and used as the driver response in the main VituixCad simulation.

Obviously this will only work if you can get a good match between a standard sealed simulation of the driver and the actual traced curve. In this case I would say it is good (as has been the case for peerless drivers that I have traced).

Some drivers I traced I could get no resemblance at all (MCM was an example) and it was likely that they had been measured open baffle or even baffle-less.

So that's it. I'm not sure if anyone else has been using this technique, I've not seen it posted anywhere before, but I thought it was a good way to play around with drivers you may be considering buying for a speaker project to get an idea of how they would look with your chosen alignment, somewhat more realistic than a spliced in box sim :)

Tony.
 

Attachments

  • 20L_comparison.png
    20L_comparison.png
    23.4 KB · Views: 68
  • divide.png
    divide.png
    4 KB · Views: 59
  • multiply.png
    multiply.png
    4.1 KB · Views: 50
  • result.png
    result.png
    21.6 KB · Views: 58
Last edited:
This is what I get using the attached baffle adjusted woofer response with the crossover from post #17.

Note that the attached frd was created by first subtracting (division) The 20L closed box sim of the driver response from the factory traced curve, then adding (multiplying) by the 10L vented response sim to the flattened factory curve.

I then used vituixcad to simulate the baffle step with baffle dimensions of 227mm wide by 280mm high, and exported the response, which is the zip file attached.

The last attachment shows the comparison between the origin dayton frd (blue), the same frd with the 227 X 280mm baffle diffraction added (black) and the adjusted for reflex alignment curve with the 227 X 280 baffle diffraction added.

Tony.

I quickly tweaked your crossover a bit . You can further adjust the values to your liking.

Having a closer look I agree with lojzek that the baffle step doesn't appear to have been added in. I had a play with your drivers a week or so ago, but didn't post because you seemed to be working your way through it. with the baffle step factored in the woofer is around 85db not 90db.

I went a bit further and also adjusted for a vented enclosure and came up with the attached. for me the tweeter is the problem, with quite a ragged high end response. I'll put your crossover into my sim when I get home and see how it compares to what you posted, using my baffle adjusted frd.

lojzek, there is a tool in vituixcad to model the baffle response, but you have to export the frd and then use that new one in the sim. No feature to just put in the baffle effects directly.

edit: I'll also do mallikreddyk's as it looks better to me than the one I have attached :)

With the baffle response factored in it took a little longer to correct but I think I did it :)
 

Attachments

  • doneish.PNG
    doneish.PNG
    123.3 KB · Views: 71
Having a closer look I agree with lojzek that the baffle step doesn't appear to have been added in. I had a play with your drivers a week or so ago, but didn't post because you seemed to be working your way through it. with the baffle step factored in the woofer is around 85db not 90db.

I went a bit further and also adjusted for a vented enclosure and came up with the attached. for me the tweeter is the problem, with quite a ragged high end response. I'll put your crossover into my sim when I get home and see how it compares to what you posted, using my baffle adjusted frd.

lojzek, there is a tool in vituixcad to model the baffle response, but you have to export the frd and then use that new one in the sim. No feature to just put in the baffle effects directly.

edit: I'll also do mallikreddyk's as it looks better to me than the one I have attached :)

I noticed yours matched the phase better than mine. Should I use that instead?
 
Just another Moderator
Joined 2003
Paid Member
It looks good. Since both mine and yours are from traced curves I wouldn't sweat the phase tracking too much. It is a guess at best what the offset should be, so either one could be closer or further from the ideal tracking, which will only be known with proper measurements :)

Tony.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.