Nooby speaker design questions

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm thinking about building a speaker, small two-way, sealed, with sub, listening position will be somewhat close.

I've looked at a few mid-bass drivers, and modeled them in Unibox. Here are my questions:

1. Most show a pronounced impedance peak. Like 5X higher than where it flattens out. What can I do to smooth this out? Would cone loading via tuned resonant stubs work? The frequency is at 65 Hz or so, but that is probably irrelevant.

2. Is there a relation between the modeled -3 dB frequency on the low end, and the crossover frequency I should select? Stated another way, if I'm shooting for a crossover frequency to a sub of 100 Hz, what should the -3 dB frequency be for the mid-bass driver?

3. Are there general guidelines for selecting an "optimal" crossover frequency between the mid-bass driver and the tweeter? How does this relate to driver characteristics? Stated another way, if I wanted to target say 1.7 kHz as a crossover frequency, what should I be looking at when picking drivers? My paper tiger is an MTM config for the moment.
 
I will try to answer #2 and #3....


#2 - assuming your sub has an adjustable low-pass filter? Then your 2-way should have an F3 of less then 100-Hz. I would try the 2-way as is and let the natural roll-off of the box handle the low end. If the bass (from the 2-way) is too distorted, start adding a high-pass filter to the 2-way to block the bass frequencies. Then adjust the sub to make up the difference. FWIW, I think 100-Hz is too high for a subs, I think it muddies the deep male voices, I like 60-Hz or lower.

#3 - Go over to http://www.madisound.com/catalog/product_info.php?products_id=1750 , get the 6th edition of Loudspeaker design cookbook.

Generally speaking, you want to be 2-3 octaves above the tweeter Fs
 
1: that's part of the physics of the driver, no point in trying to change it, deal with it by having the correct box size

2: For for 100Hz xover to sub, ideally, the mid woofer should fall off in a predictable way below 100Hz . This can be controlled box size, and is easy to predict, fire up Unibox (You need to suss how the xover slope of the sub will affect the combination).

3: the ideal is to ensure the driver has a smooth response one octave either side if the crossover frequency, (freq x2 or x .5),
but this can change depending on the quality of the driver (smoothness of fall off), what xover slope you're using (2nd order or higher xovers make the result a bit more controllable), what you're trying to achieve, how brave you are etc....

for 1.7 kHz mid- tweeter xover, you're looking at a smooth fall off above that point for the mid, metal cones have another peak that needs to be dealt with separately;
and a resonant freq of < 900Hz and smooth fall off for the tweeter (most ferrofluid tweeters look pretty good in this respect)

Have look at few good designs (such a Zaph's) and take it from there. Any particular drivers in mind?
 
All,

Thanks for the assistance. I appreciate it!

1: that's part of the physics of the driver, no point in trying to change it, deal with it by having the correct box size

Hmm. I modeled this in Unibox with a Qtc of 0.707. Amplitude response is smooth, but impedance shows a pronounced peak. User error, maybe?

2: For for 100Hz xover to sub, ideally, the mid woofer should fall off in a predictable way below 100Hz .

This is an area where a bit more specificity could help me. All the drivers I've modeled "fall off smoothly", from what I can tell. In a sealed box, with a Qtc = 0.707, one driver is at -3dB at 110 Hz, another at 80 Hz, another at 65 Hz. Interestingly, F3 seems to inversely correlate well with driver cost. Rolloff slopes are about the same, -18 dB/octave, from what I can tell from Zaph's charts.

I suspect that I want the driver to have a F3 point well below the crossover frequency to the sub. But, I don't know this for sure. Doing without the crossover probably won't do anything to reduce distortion at low frequencies.

I suppose I could port the box for more extension, but then I need to figure out what the challenges of integrating that to a sub will be.

3: the ideal is to ensure the driver has a smooth response one octave either side if the crossover frequency, (freq x2 or x .5),
but this can change depending on the quality of the driver (smoothness of fall off), what xover slope you're using (2nd order or higher xovers make the result a bit more controllable), what you're trying to achieve, how brave you are etc....

for 1.7 kHz mid- tweeter xover, you're looking at a smooth fall off above that point for the mid, metal cones have another peak that needs to be dealt with separately;
and a resonant freq of < 900Hz and smooth fall off for the tweeter (most ferrofluid tweeters look pretty good in this respect)

Have look at few good designs (such a Zaph's) and take it from there. Any particular drivers in mind?

OK, after reviewing Zaph's site, I selected drivers for this exercise based on low distortion performance and price, to a lesser extent. Those drivers were:

Peerless 830882 (now replaced by 831882)
Seas W15LY
Seas W15CY
Seas W15CH
SS 15W8530K01 (priced out of my budget, but gotta see what the best might do)

All drivers are 5.5", and I'm looking at an MTM config for the moment. Of these, the W15LY and W15CY are probably the best mix of distortion, extension, and price. I want to like the Peerless, but I believe the Peerless lacks extension in a sealed box.

For tweeters, I was looking at the Seas 27TDFC, Peerless 810921, and perhaps an Excel tweeter. The best of these would appear to be the 27TDFC.

One thing: you could do an TL for the midwoofer, which smooths the impedance peak.

Interesting. Usually TL's are designed for low frequency extension and are large. I need small. Can those be tuned to specifically deal with an impedance peak, at perhaps a smaller size?
 
"I suspect that I want the driver to have a F3 point well below the crossover frequency to the sub' - yep

"impedance shows a pronounced peak. User error, maybe?"
no, there's alway going to a peak at Fs (and another at Fb if you go ported), don't worry about it.

One thing you can try is to port the mid box anyway, if you don't like how it combines with your sub you can seal the ports. I've done this with the now obsolete Peerless 5.5" 850488, which lets me tweak at will, depending on how much bass I'm in the mood for :), I think the new Peerless drivers are similar in the low end (but have a more pronounced peak at the high end). The only disadvantage of this is that you may need a slightly larger cabinet.
Seas or Peerless, I don't think you can go wrong, I'd choose by the best combination of smoothness of response, bass extension and efficiency.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.