Econowave Speakers

[richidoo]

OK, then I'll try longer ports. Thanks Mike

So the design process is:
measure the driver,
simulate box as a starting point.
Build the box,
tune port length and stuffing by ear.
Don't tune port length to simulated Fb.
Ignore the simulation as much as possible for sanity.

Is that what you do? Mike? Scotty? Others?  I'd love to hear your design approach to a ported box. Thanks!

Is Xmax the only reason to limit port length, or is driver resonance a barrier also? (for max flat)

Thanks
Rich

[Face]

OK, then I'll try longer ports. Thanks Mike

So the design process is:
Measure the driver.
Simulate box as a starting point.
Build the box.
Tune port length and stuffing as simmed, leaving ports a little longer than recommended.
Measure nearfield.
Fine tune stuffing and port length by measurements and ears

This is what I do...and using this program: http://audio.claub.net/software/jbabgy/WBCD.html  
WinISD is a toy in comparison.    

You may find this helpful as well: http://audio.claub.net/software/jbabgy/BDBS.html

[_Scotty_]

Rich, what we can't see from your measurements so far is any evidence of a pronounced peak in the output of the speaker at or near the tuning frequency.
What we do know is that you have a depressed midrange, psychoacoustically speaking you have speaker with a loudness contour curve built into it.  I can't see how the bass can sound right to you when it is boosted relative to the midrange.
I would be tempted to try measuring the speaker outdoors with it sitting on chair to raise it off the ground. This would lessen the effects of any room boundaries on your perception of how the speaker sounds and possibly improve your measurement accuracy.
As far as stuffing goes I have had much better results from using cotton furniture batting in the box rather than fiberglass. About an inch of cotton batting on the back and side walls seemed to be about right. Be sure to leave the front baffle undamped.  If the midrange still seems too dead cover only half of the surface area of the interior of the box and alternate the coverage of the walls. Any wall directly across from another wall should have half its surface covered with batting and the other half should be bare.
Scotty

[_Scotty_]

Rollo, thanks for the kind words, it's not broke so I'm not fixing it, yet. 
Scotty

[richidoo]

OK, thanks Scotty. I'll see if I can measure outdoors. I agree it is like a loudness button all the time. 

Another thing to consider is that I've never measured the box without the low pass filter. If the driver does have too high inductance or if the crossover design is wrong, then I could be chasing my tail for no reason.

Mike, I did switch over to WBCD after you reminded me earlier in this thread, and I like it! It gives similar results as WinISD for the box. I agree it is more powerful. I can't get the excel charts to work, even with Jeff's help, but the box calculations work. I used that for everything since the first design. I just grabbed WinISD today because I can see and post the FR charts. And the sim results are similar.

Thanks guys! I'll let you know how it works without the crossover and with longer ports, etc.

[richidoo]

I have two 11.5" ports in each speaker. Slightly less boomy in the mid bass, but even more punchy in the lowest freqs. Slightly less loud in the upper bass, but still too loud in the lower freqs. Bass speed is slow and muffled, stretching up into the midrange. The speaker sounds less of a whole, more confused now than before. I assume the ports are too long and things could be improved by shortening them. But the result will be something between what it sounds like at 6" and what it sounds like at 12", I don't like either. I don't like the exaggerated physical punch of the very low freqs, and I don't like the smearing of the group delay.

What about other variables like smaller diameter ports? Is there a sound quality difference between a narrower port and a wider port both tuned to the same freq?  Most pro audio PA monitors with 12" woofer have smaller diameter ports.

The DATS (Dayton Audio Test System) arrived so I can make impedance and T/S tests. Maybe my driver measurements are at fault. It will be interesting to see if its measurements vary much from the previous generation unit, WT3.

[Face]

Where are you crossing over?

Please don't tell me you're listening to the woofer by itself... 

[richidoo]

Tweeters are on! 

Nominally the xo is 1500Hz:

[richidoo]

I removed the foam fill, so back to ~120 liters. FG is back in, and the 12" ports are installed for now. I intend to seal the box, but I'll listen to it for a bit first with the bigger box.

I found an online calculator based on Vance Dickason's Loudspeaker Cookbook recipe. It gives very different values than the 2 simulators I tried before that use the classic formulas. Larger box, larger diameter but shallower port. I'm not planning on building these at this point, but interesting alternate view.

[richidoo]

I have found at least 4 different formulas for determining the optimum volume and port geometry for a max flat vented box. The volumes range from 136 to 194 liters for the same T/S input. Science?  

WinISD beta 0.44  
Jeff Bagby   http://audio.claub.net/software/jbabgy/WBCD.html  
Vance Dickasen http://www.mh-audio.nl/BassReflexLoudspeaker.asp
Don Keele http://diyaudiocorner.tripod.com/formula.htm

I am thinking the box needs to be significantly larger for these drivers to reach their optimal max flat alignment. Bigger is better just like sealed.

[richidoo]

I have the ports sealed up now with temporary plugs. To me this sounds so much better than ports. I'd like a little more down low, but not much.

Floor bounce occurs at 468Hz, according to this online calc
http://mehlau.net/audio/floorbounce/

[rollo]

   Rich , could you use a method of reflecting the floor bounce away from your lisetening position ? Back in the day my buddy used mylar panels on the floor to redirect the beaming tweeter  to diffusors on the side walls from his B&W 802s.
   

charles

[richidoo]

I don't know Charles, but it sounds like a good idea, but low WAF since it will need to be hefty enough to reflect at 400Hz, and angled differently than the floor it rests on. Bachelor fix.

Face, I need some PCD help. I can't see any FR in PCD. I see phase, filters, impedance plotted in the main "system response" graph where the FR should be, but I see no frequency response. Nor do I see FR in the driver specific FR/filter graphs down lower. I have the buttons pressed on the graph to show raw driver SPL and driver response, and I have all the other data entered, like offsets and such. Screen shot attached.

I measured FR and phase with HOLMImpulse and exported in several ways (linear, log, various delimiters) but no fix. I have read the FRD consortium notes on how to make FRD file. The frd file is importing into PCD and data is displayed in the SPL sheet, and the data looks OK. Any ideas?
Thanks
Rich

[Face]

I've only used Arta and the Omni mic and never had that issue.  But, the same thing happened to me when I tried to import measurements taken from Jon Marsh(HT Guide).  I don't remember off the top of my head what he uses, but it's Mac based. 

Do the samples included with the software work? 

[richidoo]

Good idea, thanks Mike. I hadn't tried the samples. Yes, the sample frd's work.

Comparing the sample frd to mine I see that my amplitudes are negative values (relative to 0dBFS) not in positive absolute values like the sample.frd.   I have ARTA so I'll try that if I can't convert the HOLM files to positive amp. values.