Does Sonotube + 2 Shivas = Bass Nirvana? (1 Viewer)

[Phil Olson]

Tom/Dan, (plus anyone else who'd care to comment)
I hope I got your attention!
Now that I've got my Snell SUB-550's dedicated to base enhance, (BITCHIN by the way

), it's time to think about the LFE channel via DIY.
I recall in an earlier thread there was a discussion about driving 2 woofs electrically out of phase but mechanically in phase, (push-pull?), so as to eliminate harmonics due to uneven voice coil anomiolies, (gravity etc.).
So my thought, as I am about ready to plunge into my first DIY, would be to use a Sonotube with one Shiva on top and one one the bottom wired as above. Good plan? What would be gained by having 2 up and 2 down ala Tom V's BIG MAMA .
What performance could I expect out of these? Any takers?
Phil

 

[Tom Vodhanel]

>>>Now that I've got my Snell SUB-550's dedicated to base enhance, (BITCHIN by the way ), it's time to think about the LFE channel via DIY.>I recall in an earlier thread there was a discussion about driving 2 woofs electrically out of phase but mechanically in phase, (push-pull?), so as to eliminate harmonics due to uneven voice coil anomiolies, (gravity etc.).What performance could I expect out of these? Any takers?

 

[Tom Vodhanel]

Hey, it ain't a serious talk about bass reproduction til
Stan shows up ! (Stan,along with designing most of my previous DIYs, basically taught me 99% of what I know regarding bass,and the other 1% is probably wrong)
------------
I said >>>Reducing this(reactive forces being transmitted to the enclosure), minimizes even-order harmonic distortion.>>>
Stan responds...
***Not exactly. Two different effects.***
By two different "effects"...you're referring to the 2 seperate practices being discussed right?
1)simply mounting drivers to oppose one anothers physical movement
2) mounting drivers(without regard to their physical location) to have one moving *in - toward the mag*, when the other is moving *out - away from the mag*
BTw, I know there's a technical term to describe woofer movement *in and out*...can't think of it though...?
>>>>>1) Reactive forces..you know for every actoion there is an equal and opposite reaction...
2) Driver cone movement differences inward versus outward when fed a symmetrical signal such as a sine wave.
Number one can be largely cancelled by mounting identical drivers on oppoosite phases of a symmetrical enclosure.Number two encourages even order sistortion which is not very harsh sounding.

 

[Poul]

Phil,
this is not a good plan. If you want to achieve advantages of Isobarik design, put two drivers as close as possible to each other. I'm big fun of this design, but you'll be using two drivers without increase in sound level. If it's worth a price for you, do it properly: mount them on the same wall, face to face.
With this design, you should have ½ of the volume needed for 1 Shiva.
Having Isobarik on top and bottom - 4 drivers total will give you all advantages of both Isobarik and dual firing , for the price of 4 drivers, but in volume of 1 Shiva.
------------------
The reality is poor imitation of art.

 

[Seungsoo Hwang]

>>>3) try to desing it to use one large port, rather
than 2-3 smaller ports.(Dan may disagree)>>>
Tom, by this u mean like a 10in D port or something LARGE, right? I was wondering about my 265L design, which calls for 1 10.75"L x 4"D port. I was told I could use 2 22"L x 4"D ports for the same tuning freq. Im assuming in this case the 2 ports would be better, no?

 

[Tom Vodhanel]

Stan sent me this via private email, seems he's having some trouble posting here. Hopefully, things will get worked out soon. He asked me to post it for him in the meantime.
-------Stan's post---------
*****By two different "effects"...you're referring to the 2 seperate practices being
discussed right? *****
The reactive forces are occuring regardless of how much distortion is also being produced. This is what I mean by two different things.
Suppose someone pushed me on the chest and then pulled me back, again and again. Do it real fast and I could vibrate like a sub cone. If another guy pushed on my back at the same time the first one did I would stay in one place. And if they both pulled I would stay in one place.
So a sub's cone might go out and then a reactive force takes place against the enclosure. At any SPL or THD level. Servo or nonservo.
Two identical high THD drivers can be mounted on opposite sides of an enclosure such that the reactive forces are cancelled yet THD remains. The reactive forces are still taking place, but they are equal in amount and opposite in direction. Also, the reactive force is what causes the little sunfire sub to bounce around so much.
You can partially cancel even order THD via push-pull type concepts. And leave reactive forces unchanged or cancelled depending upon how you mount the drivers. You can cancel reactive forces and still have unchanged THD. ( Kind or repititive)
*****1)simply mounting drivers to oppose one anothers physical movement*****
If you mean so that reactive forces seen by the cabinet are equal in amount and opposite in direction or sign then reactive force canellation can occur.
******2) mounting drivers(without regard to their physical location) to have one moving *in - toward the mag*, when the other is moving *out - away from the mag******
IF the driver has significant off-center characteristics then this can help the net of the two drivers to act like one less flawed woofer. It wouldn't help a vel servo much. It wouldn't help a ported sub near the port freq much since cone movement is low hence THD will be low. In that case the reactive forces would also be low since it is proportional to how much heavy stuff is moving. Cones which are light also would produce lower reactive forces if they move the same amount as an otherwise comparable cone which is heavier.
*****BTw, I know there's a technical term to describe woofer movement *in and out*...can't think of it though...?*****
Didn;t they called it the old in-out in CWO.....
*****Ok, with you so far...and the cancellation of #1, will lead to EVEN orders being reduced/minimized , right? *****
No. Enclosure vibration / hopping would be reduced if reactive forces are cancelled. THD is not really affected.
*****Ok, I'm assuming the *even order* this will encourage will occur *IN PLACE* of a combination of even and odd orders that would have occured if #2 wasn't used? Perhaps NOT a combination of even and odds....but only odds? .*****
Suppose a sub is producing 30 hz at 100 dB and the HD totals 1% The distortion harmonics would be at a level of around 60 dB, total effect. Some of that extra signal will be at 60 hz, 90 hz, 120 hz 150 hz etc. This is the HD. The even orders are 60, 120, etc They are related to symmetry issues. Getting a pair of drivers to act a bit more symmetrical is what the push-pull thing does. But each individual cone still pushs or pulls against the enclosure with the same reactive force it ever did.
******And ALL of the #2 possibilities are going to be dependent on indvidual drivers to an extent(i..e...the severity of the "non-linear-anomolies*)...? ******
The extent to which the driver reacts equally in either direction. Suppose you attach a D-cell battey to a sub driver and then cone moves out. Reverse the hook up and it will go inward. Within xmax, this leads even order distortion issues If the driver favors one direction over the other.
*****AND...all this #2 stuff, is usually going to occur when XMAX limitations are approached? Again, this is probably driver specific, right?*****
No, it often occurs at any level but the amount varies. Running out of voice coil linear travel mostly encourage odd order trouble. You can drive a cone past xmax and have sowehat okay even order HD if it continues to travel the same amount in either direction.
*****I recently built a 42"/18" sonotube, and put a HSU on each end. I tried flipping one,(reverse phase)...and then measuring it compared to simply having both drivers pointed out... IF there was a measurable difference, my equipment (mainly the SPL meter) isn't sensitive enough to verify it....i.e., the THD differences where ******
Look at the spectra of the even orders when fed a freq where the cones move alot. Good drivers will tend to have less of a change since they tend to be both symmetrical and equal to the other. Folks have measured as much as 10-15 dB or more of even order HD reduction. The trouble is that it's the odd orders thatsound really bad and the even orders are getting higher and higher shortly before the sub loses it anyway.
.
*****Ok, so this means that the advantages gained using #2 , IS driver dependent, right? *****
Yes. And it will vary with freq and travel, etc
******I know we exchanged a TON of email about this a couple months ago Stan, but you pretty much lost me on a few things back then(imagine that). ******
Okay! Some points:
Sub amps when clipped produce lots of all harmoincs.
Sub drivers produce lots of odd order HD when the linear travel is exceeded. No matter where you mount them this won't change.
Sub drivers produce even order HD related to how equally they travel inward versus outward. This is what can be partially cancelled HD-wise though it's not a huge benefit to do so.
Sub drivers produce reactive forces when the move regardless of the HD level. mostly it's just moving mass vesrus movement.
I hope this helped a bit...
Stan

 

[Tom Vodhanel]

Poul sez...
Phil,
this is not a good plan.

 

[Tom Vodhanel]

>>>Tom, by this u mean like a 10in D port or something LARGE, right?

 

[Tom Vodhanel]

My prior comment denoted by the (******)
*****By two different "effects"...you're referring to the 2 seperate practices being
discussed right? *****
Stan's relpy.
>>>The reactive forces are occuring regardless of how much distortion is also being produced. This is what I mean by two different things.

 

[Phil Olson]

Wow, this is turning into a great thread! Let me see if I've kept up.
Imagine a Sonotube with 1 Speaker in the top pointing out and one in the bottom pointing out, (both facing out in opposite directions). Both +'s are hooked together and both -'s. The opposing mechanical forces keep the sub from jumping all over the room and causing distortion, (odd or even?).
Now if I take the top woof and have it point inward then reverse the +/-, I've kept the same mechanical effect and in addition, helped cancel even HD since one is firing into it's basket, while the other is firing out of it. But as Stan says, it's probably not much.
Another thought. Suppose I put four sides on the Sonotube and ran a silicone bead so that each side would be attached to the tube, (0, 90, 180, 270 deg.). This could help stiffen the tube as well as add mass. The open spaces between the tube and the four corners could be filled with low expanding foam or sand or fiberglass, whatever. Now the nice part is that these sides could be 3/4" plywood all ready for a great looking finish, making a great looking woof. Are there any drawbacks?
Also, what if I built a 2 woof design to get my feet wet and to check out performance. I could always buy two more later and just upgrade the tube right?
Just a thought. From your post Tom, I figured a 30" Sonotube about 4' high for four drivers right? I think that I may need to get the wife used to a smaller one first; for awhile anyway.
Phil

 

[DanWiggins]

Few things:
1. Tom's correct about bigger vents being better. The reason: turbulence. In a pipe, you'll have turbulent flow at the boundaries (or the walls of the port).
Since a larger diameter pipe has a higher area/circumference ratio than two smaller diameter pipes with the same overall area, you're better off with two ports.
All that given, though, I've NEVER seen a situation where a 4" diameter port was a significant limitation. I'd stick with a single 4" vent for a single Shiva. Even in a big EBS box, your vent velocity will not break Mach 0.1.
Plus, with longer vents (like you'd use for twin 4" vents), you WILL have resonance issues to worry about. Realize that a 22" long vent will start to have pipe modes at 154 Hz. That's a little close to the typical XO point of 100 Hz for my comfort. I'd like to see a good octave between the lowest resonant mode of the vent and the XO point.
2. "Reversed" subs, a la M&K. Basically, you hook two drivers up so they are in acoustical phase. One driver is connected to move back into its basket on a positive going pulse, and the other is conencted to move forward out of its basket in the same positive going pulse.
Obviously, this requires one of the drivers to be mounted "backwards". For a tube, think of the top Shiva firing "up" (that is, the cone is facing the ceiling, and the magnet is towards the floor). The other end of the tube has a Shiva mounted the same way: cone towards ceiling, magnet towards floor. And the two drivers are wired out-of-phase.
The result is that when a positive pulse is fed to the system, the top Shiva moves up (towards the ceiling), and the bottom Shiva moves down (towards the floor). Acoustically, they are in phase.
Now, THEORETICALLY, this will reduce distortion. However, unless the drivers are mounted close together, and in the same radiating plane, you're not going to gain much. Why?
Distortion is a set of harmonics that are not desired. That is, they are of higher frequency than the fundamental. For example, the 3rd harmonic of the low E on a bass guitar is at 120 Hz.
So, distortion in the bass and lower midbass range is of pretty high frequency; in fact, we can see that the 5th harmonic of 80 Hz is 400 Hz! Realize that the wavelength of 400 Hz is only 33.9 inches. Thus is the path between your two drivers is 33.9", the distortion components will actually ADD together when you're on the plane halfway between the drivers!
Now, run the same process for 2nd - 5th harmonics of frequencies below 100 Hz, and we can see that there will be all KINDS of distortion products that will add, rather than subtract.
And, the off-axis nature of your listening position will change the frequencies that add/subtract to each other. As you can see, it's NOT an easy thing to calculate the total distortion products within a room!
Remember, distortion is a high frequency issue. Your crossover will NOT help out, and it's tough to physically mount drivers such that the distortion components always subtract.
Now, mounting like M&K does (forward firing driver, with a reverse mounted driver directly below) does help; however, it still has problems at the higher frequencies.
Best bet: choose a driver with low distortion. All other things being equal, longer Xmax means lower distortion for a given SPL level. There are some drivers that do have lower distortion than others, but for the most part, distortion is inversely proportional to swept volume capability. So if you're looking to reduce distortion, then buy drivers with the most swept volume.
Hope this helps,
Dan Wiggins
President
Avatar Audio

 

[Poul]

TV,

Quote:

1. If you take the top woof and have it point inward then reverse the +/-, you didn't change anything here.
2. The jury is still out here for effects of (proper) Isobarik design, IMH (and subjective)O it makes very significant improvement.
------------------
The reality is poor imitation of art.

 

[Tom Vodhanel]

>>>I think we have a major confusion of terms here. If I understand Phil correctly, he wants to put two Shivas at opposite sides of the tube, and connect them out-of-phase (i.e. plus of one driver connected to minus of another one and v.v.).

 

[Tom Vodhanel]

>>>Wow, this is turning into a great thread! Let me see if I've kept up.>Imagine a Sonotube with 1 Speaker in the top pointing out and one in the bottom pointing out, (both facing out in opposite directions). Both +'s are hooked together and both -'s. The opposing mechanical forces keep the sub from jumping all over the room and causing distortion, (odd or even?).Another thought. Suppose I put four sides on the Sonotube and ran a silicone bead so that each side would be attached to the tube, (0, 90, 180, 270 deg.). This could help stiffen the tube as well as add mass.> The open spaces between the tube and the four corners could be filled with low expanding foam or sand or fiberglass, whatever. Now the nice part is that these sides could be 3/4" plywood all ready for a great looking finish, making a great looking woof. Are there any drawbacks?>Also, what if I built a 2 woof design to get my feet wet and to check out performance. I could always buy two more later and just upgrade the tube right?>Just a thought. From your post Tom, I figured a 30" Sonotube about 4' high for four drivers right? I think that I may need to get the wife used to a smaller one first; for awhile anyway.
Phil

 

[Phil Olson]

Dan,
Thanks for the info on the port resonances. I handn't even thought about it. From your dessertation(sp?) on HD, I think that I'll stay with standard wiring/firing. I'll be placing an order as soon as I figure out the design.
Tom,
My thought on the four sides would be to make it look good, (WAF), as long as no one saw any drawbacks to it. I'll probably go with dual endcaps to avoid flexing as well.
I WAS thinking about a new tube if going from 2 -> 4 Shiva's, assuming that I could reuse the drivers from the first design into the second.
In addition, since I have the Snell SUB-550's on the sides, my guess is that I would only need the Sono design for the LFE, (and the area below 40 Hz if I change my sides from full to small Xovered at 40 Hz). What would a 2 Shiva design do? 120+? I would probably use a 18 or 24 inch Sono. I guess I could also use a 30 inch and make it short, but I'd be concerned about vent length.
Phil
P.S. Thanks, all for some very stimulating input!!

 

[Johnny Giant]

Where are you coming up with 136 db from four 12" drivers at 20 hertz. ??? You must be taking it from a computer program. I doubt this is accurate, you will end with alot less in the real world. The 1w/1m measuremets are usually taken at a 100 hertz or higher. This inflates the performance on a computer program. You need to measure with a 20 hertz tone in the room with an SPL meter that is accurate at 20 hertz from 1 meter. I think you'll find your 136 db ouput to be a fantasy.

 

[Seungsoo Hwang]

dan says>>>Plus, with longer vents (like you'd use for twin 4" vents), you WILL have resonance issues to worry about. Realize that a 22" long vent will start to have( pipe modes at 154 Hz. )That's a little close to the typical XO point of 100 Hz for my comfort. I'd like to see a good octave between the lowest resonant mode of the vent and the XO point.
So Dan, you would stick with a single tube in a case such as this? Is there a noticable difference if I used 1 10.75" port vs 2 23" ports? If not, I may as well go with one to cut down on all avaliable costs.

 

[Tom Vodhanel]

JG jumps in...
>>>Where are you coming up with 136 db from four 12" drivers at 20 hertz. ??? You must be taking it from a computer program.> I doubt this is accurate,> you will end with alot less in the real world.

 

[Tom Vodhanel]

>>>I WAS thinking about a new tube if going from 2 -> 4 Shiva's, assuming that I could reuse the drivers from the first design into the second.>In addition, since I have the Snell SUB-550's on the sides, my guess is that I would only need the Sono design for the LFE, (and the area below 40 Hz if I change my sides from full to small Xovered at 40 Hz). What would a 2 Shiva design do? 120+?

 

[Tom Vodhanel]

Dan sez...
>>>Plus, with longer vents (like you'd use for twin 4" vents), you WILL have resonance issues to worry about. Realize that a 22" long vent will start to have pipe modes at 154 Hz. That's a little close to the typical XO point of 100 Hz for my comfort. I'd like to see a good octave between the lowest resonant mode of the vent and the XO point.