I am doing my first subwoofer project, but due to a rather severe SAF I'm limited to about a 2 cubic foot box. Assuming that a 10" or 12" driver would yield good overall results in that size box I have modeled several drivers, all from PE. Since I have no experience with subwoofers at all so I don't have any feel for the how the Gain curves (from WinISD beta) will "sound". From all that I've read vented designs are considered best for HT, which will be my primary application of the sub. I was hoping the forum could advise me on my decision. Below is an Gain graph of the three PE speakers I've considered. I've included both closed and ported designs for comparison. In all cases the SPL of the closed boxes falls off faster with decreasing frequency.
- Blue is the Dayton 10" High Fidelity
- Yellow is the Dayton 12" High Fidelity
- Red is the Dayton 12" High Output. Suggested by someone on the PE tech talk.
Based on the Gain graph the 10"HF goes the deepest though only by a little and is the least efficient (1 to 3 DB depending on the freq) and the vent is a little long at 28" for a 4" port.
- Would a 3" diameter port be okay for a 10" driver?
- How do I determine if I will exceed XMAX?
- When does room gain enter the picture? My den is 2370 ft^3, but has a very large opening, equal to about 3 or 4 doorways, right next to the entertainment center. I am assuming room gain will not be a factor in my case.
- What other drivers should I look at? The AA Assassins are even less efficient so I've ignored them.
One other constraint is budget, I'd like to keep driver+amp costs to about $250, excluding shipping.
I know nothing about the drivers, but to see if you exceed Xmax, you should download WinISD Pro
The red (Dayton 12" High Output) looks the best there.
derekbannatyne - Thanks for the suggestion.
crackyflipside - Vented or closed? And why the 12" high output? Gain wise, neither of it's curves are as good (from what I understand the graph is supposed to tell me) as the 10" hi fi. What am I missing?
If you think HT will be your primary application, go with a vented enclosure AND MAKE SURE that it is tuned to a low frequency. Use a modelling program like derek suggested to help you design it.
I believe I'm the original poster who suggested the 12HO on the PE board.
Anyway, as I said over there, the 12HO was all about potential. Over here, you've added a budget constraint. With $250, looking at $120ish drivers limits you to plate amps in the 240-250 watt range. At that power input, the 10HF and 12HO are pretty much equal. The 12HO has the potential for as much as 6db more output in the 20-30hz range if you give it enough power...like 1000 watts, which is not going to fit your budget.
Also, a thought ocurred to me. On PE, you say the 2 ft^3 limit is internal volume. How did you arrive at that figure? Is that net or gross internal...the driver, port, amp, bracing aren't calculated as part of the WinISD models. Constructiong techniques can yield wildly different external footprints for a given internal volume. See GR-Research's plans for their 2.4 ft^3 internal PR
design...it's external volume is roughly 3.5 ft^3. You could knock that external down some by cutting back on the bracing and using a single thickness baffle...more tradeoffs since cabinet vibrations *may* start to color the sound at higher output levels. Looks like quite the deal though, at $129 for driver and matching PR
For a $250 budget, the 10HF and 12HO with a 240-250 watt amp are probably a coin toss after you place them in a room. They're also the best options I'm aware of that are close to your size requirements and in your budget.
Do you have an SPL meter? What volume levels do you listen at? We may be trying to get you more output than you need. Something like the DVC12 or 12HF in a 2 ft^3 sealed box will probably get you a smaller footprint and still give you decent low end extension...still 20hz usable output, but the smaller footprint will give up quite a bit of overall output.
You need to decide what your hard requirements are and where you can compromise. There are tons of options out there.
For instance, money no object for a small design...I'd probably choose the Titanic 10" in 2 ft^3 tuned to 20hz. Yields an f3 of 19hz. Not quite the maxSPL capability of the 12HF, but it goes lower and has a flatter overall response. Match it with the OAudio 500 watt amp for 108db @ 20hz. Of course, tuning a high excursion woofer that low in that small of a box is tough...a 3" port may not breathe enough so your budget goes up even more to add a suitable PR
or two, unless your design aesthetic allows external ports. Should be competitive with the SVS PB10, but the PB10 would probably be cheaper. :-)
FWIW, SVS only uses a 3" Precision Port in the PB10 and it has more excursion than any of the drivers mentioned here, but I don't think anyone complains about port noise with it...even though the rule of thumb says even the 12HO is under ported with only 240 watts input below 30hz. The 10HF and Titanic are pushing the recommended speed limit at 240 watts...the Titanic goes off the charts with 500 watts.
Isn't this fun? Good luck. :-)
I just saw your reply at PE, I'd figured that after a day and a half since my last post the topic got buried so moved it to AVS active since topics stay above the fold. Actually, I'd mentioned the budget limit as the last thing on my PE post, but it can grow somewhat, as can the box volume though my oversight on the port volume will eat that up. Noobiness. I'd set the limits just a little low because inevitably someone will bring up an option just beyond the limit, so I gave myself some room to grow.
1000 watts is way beyond the budget and unless subs are horrifically inefficient I can't imagine needing that much power, but then I have zero experience with a subwoofer.
The 2ft^3 volume was determined by making a 3ft^3 cardboard mockup and presenting it with all the enthusiasm I could muster to my wife. :eek: Several trimmings later it was down to 2ft^3, which still garnered a :( , but I think I can talk my way out of a 10% or so increase (we shall see). I used 2ft^3 as input to WinISD, but completely ignored the add on from the port volume, which eats up my wiggle room. Ooof.
I'll check out the GR research and the PB10 subs, and the 12DVC. The 12HF is the yellow curve in my original graph and it did not seem to be better except below 20hz and when comparing sealed boxes. I recall it was also very touchy on port length, so tuning without an SPL meter could be tough. The Titanic was rather pricey but I'll look at it again.
There's a PB10 in the B-Stock right now for $389...it still has an external volume of 3.5 ft^3, but it looks like one he## of a a sub for $400. See the extensive review @ hometheaterhif.com
Get that SPL meter and check your typical listening levels. With 240 watts (Dayton SA240) the 12HO will hit 108db @ 30hz, 101db @ 20hz. The 2 ft^3 sealed 12HF is 103db @ 30hz, 97db @ 20. So figure you get at least 100db @ 20hz with even modest room gain from the sealed 12HF. Actually, probably more since WinISD is modelling ground plane, not in room...should pick up at least 6db just from corner placement. The point being, if you're normal listening peaks are near 100db or lower, you'll probably be just fine with one of the sealed designs...which will be easier to build smaller. Even when showing off, I barely register peaks above 100db...normal listening with wife present probably peaks in the low-mid 90s. YMMV.
The DVC12 produces as good or better output numbers as the 12HF in only 1.75 ft^3 sealed. It's not as cosmetically pretty as the HF, but it is slightly cheaper and is built in the USA if that sort of thing matters to you. Years ago, when it was first introduced, someone even put it in a 1 ft^3 sealed box and got very respectable in room numbers with it. Used to be on the web somewhere, but I've searched several times over the past couple of years and can never find it.
|Years ago, when it was first introduced, someone even put it in a 1 ft^3 sealed box and got very respectable in room numbers with it. Used to be on the web somewhere, but I've searched several times over the past couple of years and can never find it.
Not sure if you're referring to it, but that sounds like my lil' green box. I have some pics, but not digital. Used to be on Rudy's Dayton archive, but I'm not sure what became of that. I still have the box, but it's currently black and has grown a twin
. The 12" DVC is still doing a great job for a friend in ~5ft^3 reflex tuned to 18Hz and driven by an Adire AVA250.
I think the RS drivers are a nice step up, though.
Ding! Ding! Ding!
Jack, I think we have a winner. If memory serves correctly, the DVC12 box I'm thinking of was green. Any chance you still have those original measurements around? That would probably be of more interest to the OP than photos.
I agree the RSHF is worth the extra $20 over the DVC, but the OP seems to have an exterior volume dictated at under 2.5 ft^3. Given the DVC12 needs about .25 ft less internal volume for the same Q and SPL as the RSHF, it seemed worth mentioning. At these sizes, that .25 represents a significant % savings.
If you gave the DVC12 to a friend, what drivers are in the twins now?
Either way, once Marc gets the wife indoctrinated with a small sealed sub, he could start working towards an upgrade to a bigger, ported box with either driver...both RSHF and DVC12 start looking good at 3 ft^3 and up, tuned to 20hz. It could happen. ;-) When I met my wife, she thought a 19" TV viewed from 15' away was perfectly adequate...after getting used to watching movies in my dedicated theater room, she agreed that HT space was as important as any other criteria we used when looking for "our" house. Usually, she and the real estate agent sent me straight to the potential candidate's bonus room to decide whether we spent any time viewing the rest of the property. Did I mention how lucky I am? :-) Sorry, Marc.
|Jack, I think we have a winner. If memory serves correctly, the DVC12 box I'm thinking of was green. Any chance you still have those original measurements around? That would probably be of more interest to the OP than photos.
Just about a 14" cube of 3/4" birch ply, so ~1ft^3 after some cross-braces.
|I agree the RSHF is worth the extra $20 over the DVC, but the OP seems to have an exterior volume dictated at under 2.5 ft^3. Given the DVC12 needs about .25 ft less internal volume for the same Q and SPL as the RSHF, it seemed worth mentioning. At these sizes, that .25 represents a significant % savings.
Yeah, although either the DVC or the HF are a bit high Qtc-wise in this box. The 12" HO actually looks best from that perspective, but any could use something like the 350w Rythmik amp with the LT circuit. I ran all three in 29L/350w for fun, nothing radically different down deep.
|If you gave the DVC12 to a friend, what drivers are in the twins now?
Not appropriate ones, Adire DPL12's, Qtc =~0.95. I'm experimenting with LT-type curves with my Behringer DEQ2496, and I have a few of those around.
|Just about a 14" cube of 3/4" birch ply, so ~1ft^3 after some cross-braces.
Funny. What's that joke about giving me what I asked for, but not what I wanted. :D
I thought you published some in room frequency/spl measurements of your green DVC12 way back when. Maybe I'm confusing it with another memory, but I remember being fairly impressed by such a small box and tried to hang on to the memory in case I ever needed it in the future.
|Funny. What's that joke about giving me what I asked for, but not what I wanted.
I thought you published some in room frequency/spl measurements of your green DVC12 way back when.
Ahh... ;) I don't recall measurements, but I may have done some rough ones with a RS meter. No idea where that stuff might be.
if you haven't picked a direction yet, it's wothing noting that PE has the DVC12 on sale for $80 through 17:00 EDT today. Maybe even think about a pair in a 2 - 2.5 ft^3 sealed box at that price.
Also, SVS is blowing out their PB10-ISD for $379 until they start building/shipping the 10" NSD driver.
You read my mind, or planted the thoughts. The PB10-ISD, on your recommendation, was very tempting, but it's footprint (21" deep) was problematic. So... I picked up the Dayton 12"DVC and a 300w Dash amp, though only in the last 10 minutes of the sale because I was thrashing amongst those two and a high efficiency car woofer with a Linkwitz Transform. Arg. I hate being rushed, but the deal looked good and I think 3 cubic feet is sellable. Will an LT work on a vented box?
You can take a look at the sub I built the link is in my signature. The 12HF is great only one issue is WOTW DTS but the Dolby Track is smooth sailing,
since I've never seen anyone use an LT on a ported design, I'm going to guess it's not a good idea. I would imagine that it's technically possible, but provides an unsatisfactory solution.
Just tried playing around with a LT on the DVC12 in 3ft^3 tuned to 20hz powered by 300 watts. Using the before parameters from the DVC12 in 3 ft^3 sealed and a modest target of 20hz with a Q of .5 afterwards, the model exceeds Xmax anywhere from 24hz - 37hz and then again quickly below 20hz. But it's ruler flat to 20hz. :-) In the end, the model doesn't show a change in excursion limited SPL with or without the LT in circuit.
Where are you hoping to improve? The DVC12 in 3 ft^3 @ 20hz, has an f3 of 27hz and an f6 of 20hz. Even without room gain, I don't know that you can do much better for around $200 parts. You'll quickly exceed Xmax below 20hz so there's really no room for equilization and you're already essentially flat to that point.
Oh no. I was not getting results like you were so I double checked my parameters. The XMAX was way off. So I reentered the params, this time from the detail sheet, and I am still not getting the your results but now the gain curve (Transfer Function Magnitude) is humped up between 70hz and around 170hz and falls off sooner that previously. Nuts. I think I made a mistake.
The specific form of EQ known as "Linkwitz transform" (basically notch and shelving filters) is for sealed. Other forms of eq are used for reflex to change FR, typically high-Q high-pass filters to extend response/lower f3.
...mistake buying the 12"DVC, is what I meant. I think that without some of the electro mechanical parameters the gain and SPL curves can look a lot better than they really are, so this led me astray. To be honest, the price did too. So let's see what I can achieve with the DVC and maybe some simple EQ. For the 12"DVC I think I now have the parameters correct, though I an still not getting the same results as Brent. In particular I can't get the "ruler flat" result, even ignoring unacceptable consequences such as excessive excursion. However, the SPL numbers are so high at 300W that (if WinISD is to be believed) it looks like running at 50W or so should provide enough volume and leave headroom for EQ in the lower frequencies. Couple of questions / assumptions on EQ:
- The intent of the EQ circuit is to alter the input curve (from the signal source) to compensate for some of undesirable parts of the driver's or driver+box's natural curve?
- EQ circuit is placed before the amp?
- If the EQ is an active circuit I need to power it somehow.
- Are LTs usually active circuits?
- My reading on LTs leads me to believe that it does two things: reduces signal strength for those frequencies that are "above the line" and increases for the below the line, where the "line" is the desired output level. How the compensation curve is specified is, for me, TBD
- Assuming any EQ shapes the input curve I don't see why any EQ type can't be used on any driver+box configuration, though some may be more effective than others in achieving the desired output.
- Will the cost of most simple EQ circuits have been better spent on a superior driver in the first place?
Is my noobiness just too much to stand?
Yes eq circuits like the LT are made up of active devices (opamps)and are placed before the amp and yes a power supply is needed for the opamps. LT type EQ is not suitable for ported boxes because it would add too much boost to the output below the port tuning frequency and could potentially damage the driver due to overexcursion and limit its power handling. The proper type of EQ for ported boxes include a hi pass filter to help protect the driver from overexcurion while also extending the useful bass reponse. The two circuits are implimented differently . Would just buying a better driver be answer? Maybe ! No your noobiness is no probs you are asking some very good questions.
I seriously doubt you made a mistake buying the DVC12 for $80, especially given your overall stated budget constraints. Even at regular price, it's one of the best performance/cost compromises available to the DIY market, IMO. A few years ago the DVC12 and its near clone, the original Adire Shiva, were THE 12" driver for the DIY market. There are a few 12" drivers available today that offer a little more output (need more than 300 watts), maybe a little less distortion, but they're also moderately to considerably more costly and usually require substantially larger cabinets. Search the web for DVC12 and Shiva projects and you'll find a lot of happy people.
Which of your models is not agreeing with mine? WinISD can pretty accurately recreate PE's suggested model as well as any other DVC12 alignment I've seen over the years, no matter what s/w they used.
My "ruler flat" comment was for the LT applied to a 3 ft^3 ported box. The transfer function was down something like .8db @ 20hz. However, as I also stated, excursion limits were a problem well above tuning with the 300 watt input.
Build the 3 ft^3 tuned to 20hz without any EQ and you should be very happy, especially if this is your first sub. I don't understand what you find "undesirable" about an f3 of 27hz, f6 of 20hz with SPl output of 105db@20hz, 109db@30hz. Even without any room gain, that's going to sound very nice. It's even more sweet when you consider you'll probably have less than $300 in the whole sub, depending on your aesthetic choices.
You can safely ignore the response rise you're seeing above 70ish hz. Pretty much every driver WinISD models has a raw response that resembles that. Remember, you're going to be using a low pass crossover somewhere around 80hz...and you can add it to the WinISD model to see how it will effect your output. There's also going to be natural and high pass Xover roll off effecting your mains as well...hopefully, you end up with something fairly complementary...until the room throws the whole thing out of whack. :)
SPL readings in WinISD are calculated at a distance of 1 meter, unless you change it. SPL drops off at the rate of 6db for every doubling of distance. So assuming you're listening position is 4 meters (12ish feet) from the DVC12, you'd reduce the WinISD number by 12db. This is the theoretical model, but your room will play into this so you may experience more or less actual rolloff.
I sense a little concern about SPL output so I'll offer this. My colocated dual NHT1259s (original from mid 90s, not current offering from Madisound) model in WinISD for a theoretical max SPl of 102db@20hz, 109@30hz...f3@32hz, f10@20hz. My typical listening level measures around 95db peak on a RadioShack analog SPL meter using something like the THX intro on "The Incredibles" with both mains and subs included. My "show off" volume might bump that into the low 100s. Only rarely during the last 10+ years have they made unpleasant noises that might indicate they're nearing or reaching they're limits and that was most likely frequency dependant. To put 100db in perspective...a typical lawn mower is 100db @ 3ft according to this page http://trace.wisc.edu/docs/2004-About-dB/
I always wear hearing protection when mowing the lawn, woodworking, etc...uncomfortably loud otherwise. As the saying goes, YMMV.
Originally Posted by riverwolf
build the 3 ft^3 tuned to 20hz without any EQ and you should be very happy, especially if this is your first sub. I don't understand what you find "undesirable" about an f3 of 27hz, f6 of 20hz with SPl output of 105db@20hz, 109db@30hz. Even without any room gain, that's going to sound very nice. It's even more sweet when you consider you'll probably have less than $300 in the whole sub, depending on your aesthetic choices.
WOW now that is what you call good bang for the buck! :)
It's late, so I'll be brief. I never got a "ruler flat" curve on a sealed+LT. And when I try 3 ft^3 tuned to 20hz it yields an f3 of 32hz and an f6 of 22. Must be the driver data, or I messed up somewhare. Here's the 12DVC data for WinISD pro. You'll need to rename it to *.wdr. More later.
Dayton 12 Inch DVC.txt 0.8330078125k . file
It must be late...
|My "ruler flat" comment was for the LT applied to a 3 ft^3 ported box. The transfer function was down something like .8db @ 20hz. However, as I also stated, excursion limits were a problem well above tuning with the 300 watt input.
is quoting yourself kind of like talking to yourself.
Your DVC12 parameter attachment file is coming up "not found on server".
Which version of WinISD are you using? I'm using WinISD Pro Alpha 0.50a7.
For comparison, .50a6 yields an f3 of 26hz and the same f6 @ 20hz. So the models do change slightly with versions even though I'm using the exact same parameter file for both.
this is what my WinISD model looks like. Hope it helps.
By "undesireable" I meant excessive excursion, and any other potential problems, not the f3 and f6 values. I'm using WinISD Pro Alpha 0.50a7. Subtle differences in the parameters - mine relative to yours:
On the graph, did you measure f3 and f6 relative to the 0 DB red line? I took mine from about 1 DB, at 80hz, on the premise that the preceived decline in DB (i.e. f3 and f6) is relative to where the subwoofer kicks in. If I take the 0 DB line as the reference then f6 is 20hz, which would require I EQ down anything above 50hz. Also, at a 20hz tuning the vent length is 25 inches. That will require a bend to fit in the box. That okay?
BTW - the driver and amp arrived today. Nice! Just have to complete the design.
I was referencing WinISD's 0db line. Just for fun, I ran the same alignment in Unibox and got almost identical numbers. I think Unibox predicted an f3 of 24hz, but overall SPL output was the same.
You shouldn't have any cone excursion issues with 300 watts in the 3 ft^3 @ 20hz alignment. It's actually a very good match...the minimum excursion limited power is still 360 watts above 20hz. Your amp has a rumber filter with a cutoff of 17hz, so you're in no real danger below 20hz, either.
Why do you feel the need to pull down the >50hz response? If it is really there, it's only a 1db peak at its worst; you won't be able to hear it. If you drop a 4th order L-R low pass at 80hz into the simulation, you're already down 10db by 100hz.
Elbows are fine, try to find the long sweep elbows to minimize port restriction if you can. What diameter port do you plan to use? If you check your air velocity graph, a 3" starts to exceed the generally recommended 34m/s around 25hz at full power. Some will disagree, but in the interest of keeping your external volume as small as possible, I think you'd be ok with that compromise. A 4" port will consume nearly 3x the volume. Picking up a 3" double flaired Precision Port would be my suggestion. You can avoid the elbow with the 3" PP if you can come up with an 18" internal dimension...3"x14.68 according to PP's calculator.
Well, I'm confused about the f3 and f6 predictions. The definitions of those terms refer to a "baseline", for which I have not found a definition. 0db sounds like a good baseline, but clearly the Transfer graphs rise above the 0db line. What does that mean and why should any part of the curve above 0 db be ignored? Depending on the slope of the Transfer Function curve at any point a 1db difference could be a many Hz difference. The slope of the Transfer curve 3db down from its peaks (around 80hz to 100hz) is around 1 (45 degrees) so f3 will be 1hz higher per 1db. I know these are just predictions and I am splitting hairs, but things have become a little less clear to me.
The 50hz to 80hz EQ was just to flatten the curve, but is probably more hair splitting.
Your right about the cone excursion, though below 20hz it quickly goes nuts. Rather amazing actually. Even 50w will push it past XMAX at around 13hz. A rumble filter is definately a good thing.
Only the height of the box will exceed 18" internally, but a top port sure sounds like a crud catcher. A screen or maybe an elavated cover is needed. I'll look into the PP.
Lastly, what is a good sealer strip to use for the woofer and the plate amp?
My understanding of WinISD, or any simulation program, is 0db indicates an input to the driver at equal levels at all frequencies. The transfer function then shows you how accurately a given alignment will recreate that signal. In an ideal world, everything would measure at 0db. Going positive with respect to 0db indicates a peak of some sort in the alignment.
I'm not suggesting that you ignore just any response you see that goes above 0db. I'm simply saying that the very broad and mild 1db hump that you're seeing with this model shows up nearly identically in every 12" driver I run, ported or sealed. I don't know if it's a flaw in WinISD .50a7 (.50a6 doesn't do this) or a more accurate depiction of what happens when a 12" driver reaches a frequency above the influence of the box/port loading. Many DIY subs were designed and happily enjoyed without any EQ of the upper end response before .50a7 was released to show us that we might have an amplitude peak around 100hz.
Regardless of the source, even if this hump is truly there, it will be swamped by peaks/dips introduced by the room. Even in an ideal environment, 3db is generally considered the smalled increment the human ear can differentiate. Might have something to do with why speaker tolerances are generally spec'd as +/- 3db over a given frequency range. So, the 1db peak you're seeing is essentially inaudible except by test gear. And you're still ignoring the effects of a lowpass filter that I assume you're going to use on this subwoofer. It will start rolling of your response about an octave below whatever frequency it's centered at, hence my example with a 4th order xover at 80hz being 10db down at 100hz, even with the 1db peak.
Can you introduce feet to your cabinet and make the port bottom firing? If not, a small grill of some sort over the top port, ala SVS would be appropriate.
I beleve my DVC15 came with a weather strip type gasket attached. Does the DVC12 not? To be honest, and I know it's heresy, but my sealed 1259s don't have any gasket material at all between the driver and cabinet and I've never detected any air leaks. Otherwise, foam weather stripping or any similar material is fine. PE, Madisound, etc. usually offer similar stuff in their speaker building supplies.
BTW, the slopes of speaker curves are measured in db / octave. 1st order = 6db/octave, 2nd order = 12db/octave, etc.