Thanks, yea, there's two of those (A and not A) but its in the list. It's worth considering. I'll model some tonight.
But, I'm heavily considering just doing a 2 way with the planar and a mid woofer and integrating active sub into the tower. It's probably just easier to get real 30hz performance. As much as I'd love to have a fully passive network and woofer array to get to 30hz, I'm just not sure all the work that will go into the enclosure will be worth it just to get an F3 of 30hz. I'd really like it to peak at 30hz. Gonna make these with big oak slab baffles, will be fancy hard wood. So it needs to be something that doesn't just do the job of a slightly big monitor. Gonna have that 55~60" tall power presence and can have some volume to it, so might as well make it as full range and competent for all content as I can. I won't push for 20hz, but I think 25~30hz should be doable with active integration of subs into the towers.
Very best,
Last edited:
The SD215-88 is the older shielded DVC version and is NLA to my knowledge. A is the current unshielded model.
The DS215 is the single coil Designer Series and not at all similar in spec.
The peak at 30 is not really beneficial, will likely require too big a box, and likely hinder power handling. For clean port Mach, under 17 is best, but low 20 is okay.
You will get solid 30 Hz from these, and be audible.
The DS215 is the single coil Designer Series and not at all similar in spec.
The peak at 30 is not really beneficial, will likely require too big a box, and likely hinder power handling. For clean port Mach, under 17 is best, but low 20 is okay.
You will get solid 30 Hz from these, and be audible.
The cabinet can be 3~5 ft^3 in tower form, easily 55~60 inches. Size will be there. So taking advantage of the volume as its there for use to get to 30hz or so. It's a matter of how to get there. Multiple woofers and bass reflex, passive, if possible. Or, integrate active system with its own power and have a passive 2 way and active sub in the same tower.
In a perfect world I'd love to get there with a passive network and cabinet design to get 30hz with good output. Preference to this is based solely on not needing pre-amp, sub out, etc, all that. Would work on any amp kind of thing.
But, it may become an active sub setup within the towers. I don't want separate sub(s) in the room, but am happy to integrate into the existing towers and make them larger. This just requires AVR or amp with pre-amp out, etc and then also requires external amps for the active subs, etc. Just more electronics.
Very best,
Last edited:
Definitely, it will always be near a wall, always virtually in an undesirable location without dedicated acoustic treatment, in a living room. So, "worst case." So, going for the fun of the horizontal dispersion of the planar and the beauty of live oak slabs. I have no expectation of hifi experience from these. Just trying to do something that works.
Very best,
I'm sure most of us have less than ideal listening setups, I know I do.
What I'm getting at is room gain. If you build something that has an F3 of 30hz, and shove it up against a wall. The in room response will make the thing sound super bass heavy, so consider a reduced bass shelf kind of alignment. An F3 of 30hz is into subwoofer territory.
If you're going to do the hybrid thing and separately power a built in sub, then no big deal, just adjust the gain until you like it.
What I'm getting at is room gain. If you build something that has an F3 of 30hz, and shove it up against a wall. The in room response will make the thing sound super bass heavy, so consider a reduced bass shelf kind of alignment. An F3 of 30hz is into subwoofer territory.
If you're going to do the hybrid thing and separately power a built in sub, then no big deal, just adjust the gain until you like it.
The typical weakness of a subwoofer driver is it's low sensitivity (which doesn't matter for pure subwoofer duty as it would have its own amp) and its inability to play higher in frequency to be used in a 2- or 3-way. But there are lots of subwoofer drivers that will work, and the SD215, SD270 and RSS365 would all be fine (although lower sensitivity than a true woofer).
You have been mentioning the possibility of multiple drivers since the beginning and I was hoping someone more knowledgeable than me would comment.
The low frequency of the speaker will be determined by the T/S parameters of the driver. Fs will be particularly important. Vas is typically important because most speakers have some size constraints, but in your case you can build a BIG box so you don't have that constraint. (Ignoring some of those drivers that require 8 cubic feet.)
So, the low frequency limit is going to be the same whether you use 1, 2, or 4 drivers.
More drivers gets you more SPL and power handling. Doubling or quadrupling the surface area gets you more potential SPL. This can be because you wire two in parallel for a lower impedance ---> higher sensitivity. Doubling Sd gives you +3db and the lower impedance another +3 db (voltage sensitivity). If you wire them in series you do not get more voltage sensitivity but you still have more surface area it just takes more power to get the increased output. So if your speaker could handle 50 watts with a single driver it can handle 100 watts and move twice as much air with two drives. [This is roughly accurate, but I may be making slight errors in terminology with power, watts, etc. that others can correct.]
(a) You need to keep in mind the sensitivity and power handling of the full-range or planar here. If your dual woofers can handle 100 watts and get crazy loud but your full range is going to melt at 25 watts then the gain from the dual woofers is meaningless.
(b) You also need to keep track of the sensitivity of the woofer(s) and the full-range/planar. Your woofers stated sensitivity (on the manufacturer's data sheet) is on an infinite baffle. In a non-infinite speaker cabinet away from the walls, you will lose 4-6db of baffle step loss. Thus, if you are using a woofer with 90db 1m/2.83v you will bring down the full system to 84-86db. Assuming your full-range is higher than this, then you would "pad" it down to the appropriate level. If you used two woofers in parallel you would increase the sensitivity to 96 db 1m/2.83v and bring down the full system to 90-92db. If your full range has sensitivity higher than 92 db you would "pad" it down appropriately. BUT...if your full range has a sensitivity LOWER than 92 db (or maybe 90 db) then the drivers are not going to work together. You can pad a tweeter or midrange but you do NOT want to pad a woofer as this could dissipate a LOT of heat.
In case (a) you added SPL capability that wasn't useful. In case (b) the drivers actually don't work together in a safe way.
From a pure sound quality perspective, two subs that can be optimally placed is better because the best placement of the low frequency drivers is hardly ever where you want to put your main speakers. From a practical standpoint, having everything in one speaker and in one convenient location at the front of the listening room is nice.
Definitely an active amp, with boost or even better full DSP control, gives you flexibility that you don't have otherwise.
One thing to consider...build the mains with a full-range/planar and a woofer (like you have planned so far) but don't worry about the woofer going below 80hz or so. This still allows the woofer to go up to 350-500Hz and allows the full-range to operate in the pass-band that it is best at. Then use two subs for 30 to 80-100 hz.
Yea, the more I think about it, the more I probably should just do the towers as I want, and do a 2 way in them with a woofer and planar wideband and not worry about it less than 80~100hz. Active subs in the towers for the rest. Especially if I can swap to sealed and boost a little. Saves me a lot of pain in terms of the build (and no fun ports for the kids to put things in).
Very best,
"Active subs in the towers for the rest. Especially if I can swap to sealed and boost a little."
Are you familiar with the Linkwitz Transform? I personally do not have anything active right now, but have become aware of this recently. It is basically an active/DSP filter than can allow an undersized sealed subwoofer to extend lower than it could otherwise. You would have to model a driver with the Linkwitz Transform vs just a 3-6db boost to see the difference. A subwoofer plate-amp with boost is a convenient and cost effective solution but you should look into whether a more general (and I assume somewhat more expensive) DSP option that allows a Linkwitz transform as well as other DSP options for general room corrections is worth it. [Disclaimer - you looking into this benefits me 🙂 as I am planning on building two standalone 12-inch subs next year and have not delved into the details much yet.]
Are you familiar with the Linkwitz Transform? I personally do not have anything active right now, but have become aware of this recently. It is basically an active/DSP filter than can allow an undersized sealed subwoofer to extend lower than it could otherwise. You would have to model a driver with the Linkwitz Transform vs just a 3-6db boost to see the difference. A subwoofer plate-amp with boost is a convenient and cost effective solution but you should look into whether a more general (and I assume somewhat more expensive) DSP option that allows a Linkwitz transform as well as other DSP options for general room corrections is worth it. [Disclaimer - you looking into this benefits me 🙂 as I am planning on building two standalone 12-inch subs next year and have not delved into the details much yet.]
Yes, was trying to avoid DSP/active. I have lots of that. 20+ subs. Silly stuff.
My latest sub just finished last week:
My weak point is passive crossover. I'm so used to active and its my crutch. So I keep pondering every new build if its possible to be a good option to do more passive crossover work.
This is a good opportunity for me to make something fairly large and simple for our living room. Used everyday for junk TV and random music for the family. I never use it. But they use it all day. Still, the wife likes the natural wood so I have big oak slabs to turn into something fancy and it just lends itself to being big tower speakers. If I can remove the separate subs and speakers and just have two lovely oak towers to handle it all, everyone will be happy.
So I know I'm a broken record here. I just love the idea of a wideband planar and some woofers and use the cabinet like bass reflex to get it down to about 30hz, so it will get pretty solid range without a sub. But, it just may not happen. It may be better to just do a 2 way and put integrated subs in the towers. It's ok if its not ideal for the room, it's less in the room and that's ideal for the room and situation, and they can be large as they'll be big oak slabs and my wife doesn't care how big it is when it looks like that, she likes that look, so its a great opportunity to make some big classy towers and try to make them as full range as I can without getting too wild (they don't need 20hz, but 30hz would be great).
Very best,
Minor point of difference, what you probably have here is boundary reinforcement as long as it isn't modal, which it likely will be above a certain frequency.. Room gain is a pressurisation thing and it doesn't happen in practice.
Ok,
So, coming back to the passive approach.
GRS PT6816-8 Planar wideband x 1
Cross this at 700hz to the Dayton woofers
Dayon DC250-8 x 2
Wired in parallel, 4ohm
Brings that 88.4db sensitivity +3 for the extra cone and +3 for current from parallel
4 ft^3 net internal bass reflex
30hz tuned (4" diameter port for air velocity to be under 18m/s to 30hz)
Here's 70 watts:
Need a little help with the crossover, got a nasty null up high.
Attaching the Xsim for play.
Very best,
So, coming back to the passive approach.
GRS PT6816-8 Planar wideband x 1
Cross this at 700hz to the Dayton woofers
Dayon DC250-8 x 2
Wired in parallel, 4ohm
Brings that 88.4db sensitivity +3 for the extra cone and +3 for current from parallel
4 ft^3 net internal bass reflex
30hz tuned (4" diameter port for air velocity to be under 18m/s to 30hz)
Here's 70 watts:
Need a little help with the crossover, got a nasty null up high.
Attaching the Xsim for play.
Very best,
With 10" woofers, I'm thinking 12" wide total baffle for the entire height of the front baffle. 55" tall. Planar wideband at 42" down from the top, it's 8" tall" so 5" more baffle over it to the edge. Centered. 2 x 10" woofers spaced below this, centered, probably 2" below the planar. Port will be 2" below the woofers. Edges of baffle rounded over with 3/8th inch round over bit probably. Or something like that. I like the idea of the woofers lower down, for the look too, to get the port lower to the bottom and open up the middle of the baffle to see the lovely oak grain, but I realize that's also probably bad for the design. So they can be tightly grouped. I'm not locked in on any dimensions, other than it will be tall likely.
Very best,
The first thing to try is to invert the polarity between them. Looking at the response trend you're going for, I found the widerange wanted a steeper slope. The bandstop filter added to your existing woofer filter has only a small effect.
That was for practice. The 10" shouldn't be used into breakup like this and the widerange should be able to go lower.
That was for practice. The 10" shouldn't be used into breakup like this and the widerange should be able to go lower.
Awesome, thanks, yea I was stuck and wanted to go lower on the planar, closer to 700~800hz range. Just didn't know what I was stuck on.
Thanks so much, that looks great!
Implemented the circuit, seems about the same, so I think I got that part.
I'm curious now though if I can push it lower than crossing around 1~2khz and instead get closer to 800hz or 700hz even. Since the planar's mid-range dips lower than its treble, maybe its not possible without a wonky "V" response profile. I really like this current slope. I suppose there's no real advantage going from 1.2khz or so crossing down to 700hz crossing, other than more from the planar.
Very best,
Thanks so much, that looks great!
Implemented the circuit, seems about the same, so I think I got that part.
I'm curious now though if I can push it lower than crossing around 1~2khz and instead get closer to 800hz or 700hz even. Since the planar's mid-range dips lower than its treble, maybe its not possible without a wonky "V" response profile. I really like this current slope. I suppose there's no real advantage going from 1.2khz or so crossing down to 700hz crossing, other than more from the planar.
Very best,
Last edited:
Without baffle simulation with correct size and driver positions.
The effort is complete waste of time.
Phase means nothing and the dip means nothing. Since non of it is correct.
On a real baffle you will find the 90 dB sensitivity area of the 6816 is 500Hz to 2K.
And to even get a remotely flat response it requires a specific small baffle shape.
They are practically designed to be on a small baffle or no baffle. The odd driver
response makes complete sense after modeled, since the usable bandwidth completely flattens
out.
Why everything shown in the Sim shown makes no sense its a 90 dB driver the peak is 100 dB
Your sim shows usable bandwidth to be non existent at 75 dB and the non usable peak to be 85 dB
Makes absolutely no sense.
Ive seen countless sims on the net with " Xsim" and non of them seem to include off axis info.
Maybe people seem to feel it is useful, so far have seen nothing from it that is impressive or usable.
Unless it has just been example after example of just inaccurate models done by users.
The learning curve for proper simulation is extensive and would advise spending time with Virtuix Cad.
Being it includes baffle sim and easily workable.
Real world phase response according to driver placement on a specific size baffle is incredible important.
And also must include off axis info.
GRS does show driver SD so a respectable model could be built but they dont show the actual radiating area
to even enter the shape in the datasheet. Since no questions asked about the driver SD and size.
Absolutely sure the xsim model is a waste of time.
Driver Sd is 51 cm2 and the area to enter in a model would be 34mm wide by 150mm tall.
The effort is complete waste of time.
Phase means nothing and the dip means nothing. Since non of it is correct.
On a real baffle you will find the 90 dB sensitivity area of the 6816 is 500Hz to 2K.
And to even get a remotely flat response it requires a specific small baffle shape.
They are practically designed to be on a small baffle or no baffle. The odd driver
response makes complete sense after modeled, since the usable bandwidth completely flattens
out.
Why everything shown in the Sim shown makes no sense its a 90 dB driver the peak is 100 dB
Your sim shows usable bandwidth to be non existent at 75 dB and the non usable peak to be 85 dB
Makes absolutely no sense.
Ive seen countless sims on the net with " Xsim" and non of them seem to include off axis info.
Maybe people seem to feel it is useful, so far have seen nothing from it that is impressive or usable.
Unless it has just been example after example of just inaccurate models done by users.
The learning curve for proper simulation is extensive and would advise spending time with Virtuix Cad.
Being it includes baffle sim and easily workable.
Real world phase response according to driver placement on a specific size baffle is incredible important.
And also must include off axis info.
GRS does show driver SD so a respectable model could be built but they dont show the actual radiating area
to even enter the shape in the datasheet. Since no questions asked about the driver SD and size.
Absolutely sure the xsim model is a waste of time.
Driver Sd is 51 cm2 and the area to enter in a model would be 34mm wide by 150mm tall.
Thanks, I'll look into Virtuix Cad and try it out.
The GRS planar is often used for open baffle application. I was going to do that at one point. Might still. But having two of them on hand, figured I'd see how they might work into this. If not, I can swap out to something else. Just liked the idea of trying them for fun.
All of this is just to get an idea of what's possible. Once something looks good, I can build it and then measure it and produce impedance and FR info to then do crossover work on so its more exact to the actual speaker at that point.
Very best,