Using the specs from the data sheet, a vented box of 20 L tuned to 70 Hz looks good to me. This results in a flat bass extension down to 70 Hz -3 dB. Using your measured specs, it will have a 1 dB rise, which may help counter the baffle step loss.
A closed box with Q=0.7 would be 10 L and have bass extension down to 130 Hz. Not ideal IMHO.
A closed box with Q=0.7 would be 10 L and have bass extension down to 130 Hz. Not ideal IMHO.
We still cannot get around Hofmann's Iron law unfortunately. Not with electro-dynamic drivers at least. 🙂
Last edited:
I would suggest focusing more on the shape of the frequency response and very little about numbers.
I was Under the impression that with a Qt of .38, this driver is more suited to a BR enclosure, having more damping ability/lower compliance?
I guess sealed is preferable because it's associated with higher damping and cleaner integration with the sub(s), is that correct?
In any case as soon as I can get to it, hopefully later today, I'll test it both ways.
Another question I've got is whether I should be running impedance sweeps in free air or in a suitable enclosure? I was assuming in the enclosure was best, but it changes the factory T/S numbers a fair amount so I'm not sure.
I guess sealed is preferable because it's associated with higher damping and cleaner integration with the sub(s), is that correct?
In any case as soon as I can get to it, hopefully later today, I'll test it both ways.
Another question I've got is whether I should be running impedance sweeps in free air or in a suitable enclosure? I was assuming in the enclosure was best, but it changes the factory T/S numbers a fair amount so I'm not sure.
Measure TS-parameters free air and in a closed box of known value (to retract Vas or the equivalent air volume, related to the mechanical springiness of the suspension and the cone area). Thus you get your own, instead of the factory ones. Although those aren't that bad that often. Your own measured ones do best for simulating in WinISD, Basta! or other bass alignment apps. After that you only measure impedance 'in situ' for input in sim apps like VituixCAD or to check alignment and trace possible problems like resonances which might show up. For the latter I prefer close range acoustic measurements and extracting burst decay plots. These also provide information whether your internal damping is OK or not.
Correct, in theory Qts' = 0.403 is the break point where Vb = Vas, Fb = Fs.
Yes in that we should always be making choices from a 'what works best for the needs of the app' POV.
(Qts'): (Qts) + any added series resistance (Rs): http://www.mh-audio.nl/Calculators/newqts.html
This driver is pretty damped already, and is designed for ported enclosures, and I've only seen it used that way, so I'm going to stick it out with the ports. Even at that, I'll be lucky if I get a flat response to 80hz after all is said and done from what I've gathered. A sealed box would push me up in frequency to the point that I'd probably ditch the project and sell these drivers. I briefly considered a PR, but for now at my level of knowledge I'm going to keep it simple and solve this with the least amount of variables, if that makes sense. The good news is now that I'm looking at boxes half the size or less than I originally set out to use, I only need to buy one sheet of Baltic birch ply, which is very expensive now!
For the most part.
In all simplicity a woofer wont do much below its resonant frequency.
If you want a woofer get a woofer. Use wide baffle and lower the baffle step frequency.
People design magical boxes and plenty good examples.
Or just shoot for a woofer. 25 to 35 Hz Fs is bass.
72 Hz Fs on a data sheet. Already know its pretty
much over at 70 to 100 Hz.
Find a hitter with low Fs, load up a QB3 or BB4 alignment.
ka blam that is what it does.
On a positive note people dont get extended bass shelfs.
Put your current speaker in a room, turn up the bass knob.
you like it or you dont.
55 Hz is a high tune for me. 70 Hz reflex..not much fun
that is midbass. Run sealed crossed to a sub.
And would likely work well for such a application
72 Hz is 72 Hz
In all simplicity a woofer wont do much below its resonant frequency.
If you want a woofer get a woofer. Use wide baffle and lower the baffle step frequency.
People design magical boxes and plenty good examples.
Or just shoot for a woofer. 25 to 35 Hz Fs is bass.
72 Hz Fs on a data sheet. Already know its pretty
much over at 70 to 100 Hz.
Find a hitter with low Fs, load up a QB3 or BB4 alignment.
ka blam that is what it does.
On a positive note people dont get extended bass shelfs.
Put your current speaker in a room, turn up the bass knob.
you like it or you dont.
55 Hz is a high tune for me. 70 Hz reflex..not much fun
that is midbass. Run sealed crossed to a sub.
And would likely work well for such a application
72 Hz is 72 Hz
It looks to me like your cabinet has some extra damping (volume filled with polystuffing?) which is losing a bit of gain from the port. The dip at the port tuning frequency ought to come down towards the driver's DCR.
Also, there's something going on at 500Hz that might warrant attention. Likely a panel resonance, or maybe an internal standing wave.
Chris
Yes you're right Chris, I added a bunch of cotton and acoustastuff to tame a second port peak at 150 that was wreaking havoc, and to tame some internal resonances. It did the job on those issues, but it's not ideal and is definitely siphoning off too much energy.
But that setup is history now, I'm recalibrating with a much smaller enclosure, less damping, and different port tuning. I haven't had time to set up and measure yet, might be able to get to it tomorrow but Saturday for sure, I've got my fingers crossed.
I'm on the fence about what I'll do if I can only get a nice response down to 80hz, to me that's really high as White Dragon was inferring. Like I've said before, my little standmounts with 6.5" Satoris probably hit in the low-mid 40s, and I don't want to live without that juicy fat area 40-80hz, there's a lot of beautiful stuff happening there!
But that setup is history now, I'm recalibrating with a much smaller enclosure, less damping, and different port tuning. I haven't had time to set up and measure yet, might be able to get to it tomorrow but Saturday for sure, I've got my fingers crossed.
I'm on the fence about what I'll do if I can only get a nice response down to 80hz, to me that's really high as White Dragon was inferring. Like I've said before, my little standmounts with 6.5" Satoris probably hit in the low-mid 40s, and I don't want to live without that juicy fat area 40-80hz, there's a lot of beautiful stuff happening there!
Are these on stands?
It might be possible to do a floorstanding ML-TL which could give more LF output.
Chris
It might be possible to do a floorstanding ML-TL which could give more LF output.
Chris
I've wondered about that or a TQWP, but I'm not comfortable enough at this point to proceed with confidence, or to be sure this driver is actually a good candidate. To answer your question, they've turned into a giant stand mount, working at 40L now, might end up closer to 30 if that doesn't measure well. My answer to Andrew Jones' latest haha.
Most comments on BR and MLTL loading seem valid to me. But since you plan to use a sub, I’d aim for less low reproduction instead of more. I believe I mentioned a B6 arrangement, which can be calculated or modeled on quite a few acoustic alignments. You‘ll have to use a highpass anyway, because of the rather high Fc you will get. I still would vote for such an approach.
How low does f3 have to be for a speaker to work well assisted by a woofer? Does i have to be half of the crosover frequensy? I thought 80-100 hz flat would be plenty good.
Cheers!
Cheers!
Sounds like good advice Mark. What is a B6 arrangement? I'm figuring Fc to be around 1200hz, are you talking about a highpass on the woofer as well and choking it off at say 80 hz? If so why would I do that except to reduce some cone distortion? sorry about all the question marks, they're sincere questions.
Hmm, coming back to this we already know that the driver's Qt is high enough once a little series resistance is accounted for to meet a Vb = Vas, Fb = Fs and since we're always trading efficiency for BW, how much are you willing to trade to go low without building a huge BLH?
I assume none based on 'flea power', but these type amps typically have a high output impedance (low DF) which while it interacts with the driver's specs a bit differently than a simple series resistor we can design cabs using it; for example if it's a matching impedance it could have up to 8 ohms, which can raise this driver's effective Qt = 0.84! and higher if there's any wiring, XO components added resistance.
This makes the cab much bigger and in theory tuned much lower, but for max bass it must be tuned to actual Fs in box, which in turn can be quite 'boomy'/'one note' without some form of EQ/critical vent damping, so looking like it's best to use sealed, though need to know its actual output impedance to be sure.
Attachments
Well GM, that's a challenging read for me but I'm trying to understand it all. To answer the question, yes you're right, I really can't trade much in to go low, so I just have to work with whatever suits this driver best in the long run rather than compromise SQ. I'll shoot it sealed and ported tomorrow, 60-70-80hz port tunings, different amounts of damping. I'm starting at about 40L but will then test it at 30. Just nearfield and port for now I think, to see what puts me in the best position.
Here you’ll find the original paper, the link also is on Keele’s site. Imho with such a high ported tuning it is mandatory to highpass the woofer, because the excess cone movements you get when not will ruin your midrange performance. And I mean ruin in capitals.
I like the idea of a high pass on the woofer, I've thought about putting them into my current speakers too, cutting them at 50 at least. That little 6.5" Satori really moves with the low stuff, I think this would clean things up well. I'll model it in VCAD when I get to that point, but any general idea of the cap value I'd need? I'm assuming it's just a cap, or would I need a shunt coil too?
Model the response in VCad or Basta! and tune the cap. You’ll end up with several hundred uF, which in turn would lead to electrolytes. I’d prefer an active circuit around one opamp and some buffering.