Hi Guys.
I´ve looked at this kind of problem before and (spoiler alert) for at compact PA-sub nothing beats a standard bas-refleks.
- at least when the only design-goals are output, freg.-range, size and complexity.
BUT, lets say that other factors are included in the design-requirement: Cooling of driver and amp-module and "novelty". For the design at hand the power is limited and thus cooling is maybe an mental-exercise, but the considerations should be valid for other designs as well. (Maybe I´m just bored with standard design).
Design-goals:
Please comment and suggest.
Kind regards Troels
I´ve looked at this kind of problem before and (spoiler alert) for at compact PA-sub nothing beats a standard bas-refleks.
- at least when the only design-goals are output, freg.-range, size and complexity.
BUT, lets say that other factors are included in the design-requirement: Cooling of driver and amp-module and "novelty". For the design at hand the power is limited and thus cooling is maybe an mental-exercise, but the considerations should be valid for other designs as well. (Maybe I´m just bored with standard design).
Design-goals:
- Driver: 12-280/W-8 (Thomann link: https://www.thomann.de/dk/the_box_speaker_12_280_8_w.htm )
- Got the driver already
- Cabinet-size max 50-60Liter internal ( including, braces, dividers, ports, ect)
- Storage-space is limited and use-case is portable in small car to "venue"
- Freq-range: 50Hz(-3dB) minimum, better 45Hz. 180Hz upper limit.
- I've analyzed quite a few relevant music-tracks in Audacity and primary bass is centered at 50-60Hz. Information at 40Hz is often 3-10dB down.
- Active driven. 200-400W available. active filters. simple electronic EQ and notch acceptable.
- DIY classD, LCC-supply, analog peak and RMS-limiter
- Somewhere between Hifi-and PA-sound.
- Sensitivity between 90 and 95dB
- "PA" as in party-music and robust, not stadium, Drum´n´bass and 1000 people 🙂
- I can accept some distortion at high levels, but not physical driver-damage.
Please comment and suggest.
Kind regards Troels
OK, reference for all other designs:
- Standard BR.
- Big ports (150cm2)
- Somewhat flat response to 50Hz
- Excursion max at 6,5mm with 200W in 8Ohm. ( Xmax is not really well-defined in the datasheet. Have not tested yet)
Re: Xmax; the most conservative definition is (Hvc-Hg)/2, giving 2.55mm for this driver.
Often, especially for PA, a more extended definition is used that takes account of the magnetic field extending a bit beyond the direct magnetic gap - that usually adds Hg/4 to the conservative amount. That would give you 5.05mm for this, so yes, your predicted 6.2-ish may give a bit much distortion.
Often, especially for PA, a more extended definition is used that takes account of the magnetic field extending a bit beyond the direct magnetic gap - that usually adds Hg/4 to the conservative amount. That would give you 5.05mm for this, so yes, your predicted 6.2-ish may give a bit much distortion.
Regarding Xmax: obviously a stupid design-exercise without some sort of data,
So, I made data 🙂
Not the most scientific method, but I think its valid (or "close enough for rock´n´roll)
Setup:
Blue dots are one-way travel and red dots are distortion. X-axis is RMS-voltage (50Hz), Yaxis is one-way-travel and distortion in %
Distortion is probably way to high because a lot of stuff on my desk was ratteling violently. Never-the-less: I dont think distortion is to LOW,
Basically I would assume that the distortion-graph would rise suddenly if travel was not linear anymore.
Distortion is rising after 35Vac (~150W/(ohm), but not sky-rocketing at 45Vac (~250W/(ohm).
Linear travel is then approx 8mm and Xmax is above 11mm.
What do you tink? - is it a valid approach, or am I missing something?
Pretty sure that 10% distortion is OK at these levels..
Kind regards TroelsM
So, I made data 🙂
Not the most scientific method, but I think its valid (or "close enough for rock´n´roll)
Setup:
- Variac connected directly to speaker.
- Speaker placed with cone up on desk
- Applied voltage measured
- Distortion measured with ARTA
- Travel measured (estimated) with ruler
Blue dots are one-way travel and red dots are distortion. X-axis is RMS-voltage (50Hz), Yaxis is one-way-travel and distortion in %
Distortion is probably way to high because a lot of stuff on my desk was ratteling violently. Never-the-less: I dont think distortion is to LOW,
Basically I would assume that the distortion-graph would rise suddenly if travel was not linear anymore.
Distortion is rising after 35Vac (~150W/(ohm), but not sky-rocketing at 45Vac (~250W/(ohm).
Linear travel is then approx 8mm and Xmax is above 11mm.
What do you tink? - is it a valid approach, or am I missing something?
Pretty sure that 10% distortion is OK at these levels..
Kind regards TroelsM
50Hz is near the driver's Fs, impedance is there is around 75 ohms, 17.2v around 4 watts, 35v around 16w, 45 around 27w.
6dB power increase should double excursion, 4mm to 8mm follows your chart, the 11mm seems close also.
I doubt Xmax is much more than 5.5mm, but looks like the driver has not reached Xlim at 10.5mm.
Try pushing the cone by hand and see what the limit is, and whether it is suspension or voice coil hitting the back plate.
10% THD would be fine, hard to say what % the "stuff on my desk was ratteling violently" added.
Cheers,
Art
Hi Art.
Thank you for the input.
The "power" is af cause not 150/250W due to impedance. Thats why i focused on applied voltage ( as HornResp also does).
Why do you think the limit is 5.5mm? - I dont see anything in the data that suggests a shift at that point?
Kind regards TroelsM
Hmm, all things considered in your postings, (published) driver specs + added 0.5 ohms for wiring = (~0.33 Qts'), personally would go with the pioneer's default prosound alignment, which provides enough more mid-bass 'punch' that for many (most/all?) live event goers prefer over a little lower bass 'fullness': Vas/1.44, Fb = 1.56x Fs, which HR calcs ~ -3 dB/your ~55 Hz bass mean, allowing full rated power @ 4.6 mm (18 m/s) and of course ideally need to be measured specs and averaged if multiples.
[Qts']: [Qts] + any added series resistance [Rs]: http://www.mh-audio.nl/Calculators/newqts.html
[Qts']: [Qts] + any added series resistance [Rs]: http://www.mh-audio.nl/Calculators/newqts.html
One definition of Xmax is the one-way peak excursion which produces 10% THD of the current waveform when driven at Fs, the free air resonant frequency.
Since you were using 50 Hz (not 38Hz...) and going for a high tuning, didn't realize I used the wrong data sheet for my statement in post #5, using parameters of "the box 12-280/8A" which has a longer voice coil and shorter gap than the "the box 12-280/8W" you have, as well as a higher Fs.
12-280/8A: VC length 19mm, Gap 8mm. Fs 50Hz
19-8=11/2=5.5mm Xmax.
12-280/8W: VC length 15.1mm, Gap mm. Fs 38 Hz
15.1-10=5.1/2=2.55mm Xmax, as David Morison mentioned in post #3.
Open air, little force is required for large excursion.
At 11mm one-way excursion, most of the 12-280/8W's voice coil would have left the gap, leaving little magnetic force left.
At any rate, your distortion data of the driver in open air (even if driven at Fs) will not be predictive of it's distortion under the acoustic load presented by the cabinet at the same frequency.
Last edited:
A crossover at 180 Hz is quite high, typical PA-subs/ bass cabinets are crossed at 80 - 120 Hz and have a 1 - 1.5 octave band width in exchange for more output within that limited band width. For a 45 - 180 Hz band width reflex is often the primary choice.
Liked your approach, and actual measuring is very important, even if with limited means.
Hope your microphone was practically stuck into cone, would have removed all rattling objects and much preferred you put speaker NOT on a table (or floor) but, worst case, on a bar type stool or similar, so as being slightly further away from reflecting surfaces.
maybe kludge your own IEC/DIN baffle or equivalent?
Even if only a 12" hole in the middle of any plywood/MDF piece you have around.
But in any case thanks for actual measuring.
The woofer is a bargain at that price so can´t complain too much.
A PRO woofer would have better that 0.68T flux density, that´s for sure, but they are asking too much from that 169mm magnet; I bet results would be much better if they reduced the gap surface somewhat, say going for a 2.5"/63mm polepiece and voicecoil or , say, 7.5 or 8mm plate thickness.
But hey, you have what you have, and for woofer/subwoofer duty, no big deal.
Xlim must be high, because to magnet thickness you also must add that the plate is bumped, so you "should" not hit the back plate even at high power, as long as the cabinet provides reasonable loading.
Maybe next time you will use some bridged PA amp to drive it instead of a Variac to feed it any frequency you want, keep us updated with your design.
Yes, hopefully price and quality follows each other, BUT the driver you link to is 5-10-times more expensive. - and for the purpose it would never be justified.
Thanks for the input
TroelsM
Thanks you for the comment. I agree that the load in free air will be different. Maybe I'l build a test-cab and rund the same test. - that should reveal something right?
Do you agree that as long as distortion is not too high is matters less whether the VC is in the gab, partly outside ( or in orbit around Jupiter) 🙂
Sorry for the dumb questions, - I'm not used to testing this.
Kind regards TroelsM
Hi. Thanks for the HR -file. I see your point, but I would have liked -3dB to be a little lower. That said, I see that a lot of pro subwoofers are closer to your spec, than the ones I was chasing 🙂 .
Maybe I need to spend the time testing "how far down" I can hear with standard pop-music at loud levels.
TroelsM
Indeed! The real Altec told me that for driver measurements, 'near-field' meant < 1/4" [~6.35 mm] of the dustcap, though of course with today's high Xmax/Xmech drivers it will need to be [much] longer in some cases.
Last edited:
Yes.
But I guess that as long as it´s within the conical area between the cone edge and the dustcap/voicecoil it counts as being "close field".
Sound still has to exit from "inside the conical piston/chamber" through the roughly 11" diameter "front window" whose perimeter is cone edge and inside that place we can assumes "all as one" at those lowest frequencies.
But I guess that as long as it´s within the conical area between the cone edge and the dustcap/voicecoil it counts as being "close field".
Sound still has to exit from "inside the conical piston/chamber" through the roughly 11" diameter "front window" whose perimeter is cone edge and inside that place we can assumes "all as one" at those lowest frequencies.
Greets!
You're welcome!
In retrospect, at high enough SPL our hearing flattens out to where the added LF could very well be desirable due to sheer air pressure modulating our body's internals.
Another thing to factor in, the lower the fundamental, the wider its BW, so sufficient HF is required for max impact, which in turn if the HF is elevated our brains tend to 'fill in' at least some of the missing LF, ergo might want to experiment with the understanding that IME the 'fair' sex will 'cringe' [or worse] if not at vanishingly low distortion.
Agreed, the < 1/4" [I edited the OP] is just a 'one size fits all' rule-of-thumb to ensure a max accurate measurement, though also applies for [Altec's] max 'near-field' = > 3/4" inside the diaphragm proper as defined by the suspension joint, same as their 3/4" horn/speaker, perforated cinema screen clearance.
Bass reflex may be concidered 'simple' but so often takes care & testing to get a design that is working optimally.
Bracing - Port placement to encourage convection cooling - Port tuning frequency measurement & adjustment.etc
Could you post your Hornresp input screen as that would allow further analysis of your design?
The port looks long compared to the cabinet. Will the 150mm CSA port fit?
If a hi-pass filter is needed to allow party levels make sure this does not compromise the target response too much - Once the port unloads the driver, excursion rises rapidly- so a Hi pass just below the tuning frequency will be needed which can truncate the bass response.
This is one feature of a tapped horn that I like, The high pass can be set at a lower freuency than the speakers natural roll off, and does not intefere so much with the pass band.
Bracing - Port placement to encourage convection cooling - Port tuning frequency measurement & adjustment.etc
Could you post your Hornresp input screen as that would allow further analysis of your design?
The port looks long compared to the cabinet. Will the 150mm CSA port fit?
If a hi-pass filter is needed to allow party levels make sure this does not compromise the target response too much - Once the port unloads the driver, excursion rises rapidly- so a Hi pass just below the tuning frequency will be needed which can truncate the bass response.
This is one feature of a tapped horn that I like, The high pass can be set at a lower freuency than the speakers natural roll off, and does not intefere so much with the pass band.
You have the speaker, might as well build the cabinet- you can always upgrade.
That said, I don't expect the 12-280/8W to have less than 10% THD past 4 or 5mm excursion, be interesting to see the results.