| MJK |
..... at least for me. Not a TL this time.
(Trivia question : What TV show did that line come from? You will show your age with the answer.)
Thought I would share my latest project, my son and I wired it up last weekend and it has been tweaked several times during the past few days.
http://www.quarter-wave.com/OB_Wings_Back.jpg
http://www.quarter-wave.com/OB_Wings_Out.jpg
This is an OB system (48" tall x 60" wide with the wings out) using either the Lowther PM2C, PM2A, or DX4 driver and two Dayton 15" high efficiency woofers (JBL 2226 clones) per side. I am using a DBX DriveRack PA digital crossover, the woofers roll off at 120 Hz 12 dB/octave and the Lowthers roll in at 120 Hz 24 dB/octave, and two Rotel RB-1080 200 watt amps.
All solid state and no passive BSC filter required. My calculations indicate the efficiency is about 98 dB/w/m and reasonably flat above 50 Hz. Decent bass into the 30 Hz range but no measurements yet to prove any of these numbers.
This is a very efficient system and even at fairly loud volumes the drivers barely move, you have to touch them to convince yourself they are connected. Great dynamics and bass. Best reproduction of a kick drum that I have heard from any of my own speaker designs, they can move a lot of air. Very similar sound to my Lowther ML TL design but with better bottom end, no big surprise I guess.
More details to come on my web site when I am done tweaking and find time to documented the project, |
|
|
| pinkmouse |
Interesting design. Given up on TLs then? ;)
Those pics are too compressed to do the project justice though... |
|
|
| MJK |
| quote: | | Interesting design. Given up on TLs then? |
I have not given up on TLs but I don't want to be totally one dimensional. I wrote an OB worksheet and just wanted to try it out, easy and quick project with drivers I had lying around. My wife hates them so I am not sure how long they will be allowed in the house.
| quote: | | Those pics are too compressed to do the project justice though... |
Uploaded some less compressed pictures, hope they look a little better. |
|
|
| Scottmoose |
Ha! I like it, oh yes indeed I like it. Reminds me a bit of Bert's Quasars, albeit with the additional wings and a distinct lack of BSC requirements. Very impressive Martin.
Best
Scott |
|
|
| kestrel200 |
| Not Monty Python....Bullwinkle!!!!! |
|
|
| ScottG |
| quote: | Originally posted by kestrel200
Not Monty Python....Bullwinkle!!!!! |
Thats:
"..and now here's something we hope you *really* like." (in Rocky's voice) |
|
|
| Rudolf |
MJK,
the pictures don´t show as much OB as their title promised. What about the "open boxes" behind the drivers? Was that what you meant by "OB"? ;)
Nevertheless ... it´s good to see another expert of loudspeaker theory engaging in dipoles. There still is´nt an exhaustive tool for OB design under room conditions :rolleyes:
Rudolf |
|
|
| zygibajt |
| quote: | | or DX4 driver and two Dayton 15" high efficiency woofers (JBL 2226 clones) per side |
Are those high Qts drivers or you are using active EQ? JBL2226 won't work in OB, but I suppose those Dayton are high Qts drivers.
I've been planning a similar project for a quite some time, but using two 15 inch 4 ohm car drivers per channel. Monacor has very nice 0,9-1,2 Qts, high power and long excurtion 12 and 15 inch car drivers.
Is your open baffle calculator avaiable for free?
Great project!!!
Bartek |
|
|
| MJK |
Cal gets the correct answer, vintage 1970's English humor which I watched regularly in college ..... usually with a beer or two.
| quote: | | the pictures don´t show as much OB as their title promised. What about the "open boxes" behind the drivers? Was that what you meant by "OB"? |
The front baffle is 48" tall by 24" wide. There are "sides" that extend 12" back and a pair of braces at the top and bottom to stuffen the structure. Then there are 18" wide wings mounted on hinges that can swing out to make the total front 60" wide. The assembly is open along the back and at the top, is that something other then an OB design? Maybe I am missing something or using a definition incorrectly.
| quote: | | Are those high Qts drivers or you are using active EQ? JBL2226 won't work in OB, but I suppose those Dayton are high Qts drivers. |
The Dayton drivers have the same T/S parameters as the JBL 2226 drivers, these are low Qts drivers. This is why I am using two in parallel, I get the SPL up at low frequencies using two drivers and a boost and then roll them off with the crossover. I like low Qts drivers and there is no EQ other then the volume boost.
| quote: | | Is your open baffle calculator avaiable for free? |
It will be one of my new worksheets planned for release starting in March, unfortunately the new worksheets will not be free. The OB worksheet takes into account the driver positions on the baffle, the crossover and boost, the baffle geometry, and the floor reflections. |
|
|
| Scottmoose |
I suspect they are referring to the fact that the pictures appear to show at the top, a rear panel, which I assume is a narrow bracing piece, unless it's a trick of the camera / light.
Cheers
Scott |
|
|
| MJK |
| The back is completely open with the exception of two narrow (I think 4" wide) pieces at the very top and bottom to provide some reinforcement and square things up a little. I also have a small shelf at the very top to provide some protection to the back of the Lowther cone and allow me the option of adding a suoer-tweeter later if I desire. |
|
|
| Rudolf |
| quote: | Originally posted by MJK
There are "sides" that extend 12" back and a pair of braces at the top and bottom to stuffen the structure. |
MJK,
sorry if I´m a bit inquisitive about that: Are those 12" sides running all the way from top to bottom or am I just interpreting the pictures wrongly?
If they are real sides, do you account for any TML effects that may occur between them?
It´s certainly my fault if I equated OB with flat front face and flat back face. Thought my smilie ( ;)) did care for that.:( |
|
|
| MJK |
| quote: | | Are those 12" sides running all the way from top to bottom or am I just interpreting the pictures wrongly? |
Your interpretation is correct.
| quote: | | If they are real sides, do you account for any TML effects that may occur between them? |
No I have not, I am not sure that a standing wave is possible in this type of geometry. Can you describe where a standing wave might exist? I have seen this mentioned before but have not understood how it can occur.
| quote: | | It´s certainly my fault if I equated OB with flat front face and flat back face. |
No blame required, nothing is your fault. I understand that some people like a more rigid definition and a flat baffle is clearly an OB in everybody's book. I guess things get a little fizzy when you move away from one particular geometry and try to apply tightly defined naming conventions. |
|
|
| Corloc |
Yes, Monty Python's Flying Circus, My father in law bought the whole collection. Man, did I marry into the right family.;)
I just built a pair of OB's using Visaton B200. With only 50 hrs on them , well lets just say it's a revalation.
My baffles are 11" center with fixed wings of 7" and 9". Both at 60 degrees. 48" tall with about a 10 degree tilt back.
If you get a chance you may want to try the OB experiment again with higher Qts drivers like the B200
Chris |
|
|
| westend |
Thanks, that's an excellentr target date for my new HT setup. I should understand how to program and record with the Panasonic ES20 recorder by then also,lol.
On topic; MJK's OB efforts look exceedingly good. That OB sound is hard to get away from. Maybe I should look into an OB mains for the HT setup? What would one do to match the side and rear speakers in a 5.1 or 6.1 setup? I wouldn't want to destroy the OB image with a poor implementation of the satelites. |
|
|
| MJK |
| quote: | | If you get a chance you may want to try the OB experiment again with higher Qts drivers like the B200 |
What advantage do you believe a higher Qts driver will produce? What disadvantages? I think everythign is a trade-off. |
|
|
| Rudolf |
| quote: | Originally posted by MJK
I am not sure that a standing wave is possible in this type of geometry. Can you describe where a standing wave might exist? |
May be "standing wave" is not the proper name. With "TML effects" I meant a lambda/4 resonance. Let me explain:
You´ve got the "sides" at the back of your OB and the floor as a third side.
That forms an open box, albeit with a missing fourth side at the top. Doesn´t that work like a U baffle with a lambda/4 resonance to the open back of the "box"?
I remember increasing "boxiness" with my winged open baffles, when I folded the wings back until the angle between them became less than 60°. |
|
|
| Corloc |
Pro and cons? I'm just starting this OB adventure. I would suspect the biggest advantage is crossover point of sub. I not shure I'm going to use a sub with mine. If is did I would guess it would be around 50 to 80 hz. The B200 go as low in OB as the fe206e in MLTL, and with more athority in the midbass.
I've have read the high Qts drivers dont have the same acuracy in the upper freq., but the B200 are at least as detailed as the fe206e.
Chris |
|
|
| zygibajt |
>Martin
| quote: | | I like low Qts drivers and there is no EQ other then the volume boost. |
Can you eleborate more on this? Is there any specific reason you like low Qts drivers more then high ones?
Bartek |
|
|
| MJK |
| quote: | Doesn´t that work like a U baffle with a lambda/4 resonance to the open back of the "box"?
I remember increasing "boxiness" with my winged open baffles, when I folded the wings back until the angle between them became less than 60°. |
I don't know the answer to your question with any proven math or modeling, but here is my impression of the potential for standing waves in the cavity created as shown in your sketch and used in my design.
The two open sides, the back and the top, have in a simple form a boundary condition that requires the pressure to be equal to zero. So if a quarter wavelength standing wave exists in the vertical direction, the pressure has to diminish in magnitude so that at the very back it is zero everywhere along the back opening. If a quarter wavelength standing wave exists from front to back the pressure must diminish in magnitude to be zero along the top opening. A half wavelength standing wave between the two sides needs to diminish so the pressure along the back and top is zero. With those kinds of restrictions on the pressure profile in any standing wave. it would be maximum near the bottom along the front baffle in the corner formed with the floor.
The next point it to look at what frequencies of standing waves might exist.
1/4 wave from bottom to top
f = c / (4 x 48 inches) = 71 Hz
1/4 wave from front to back
f = c / (4 x 12 inches) = 282 Hz
1/2 standing wave between the sides
f = c / (2 x 24 inches) = 282 Hz
The first frequency at 71 Hz is in the range where the woofers are playing so I guess one could posulate it might be excited if combined with another pressure profile that summed to produced zero at the back opening.
The other two frequencies are a little harder to convince myself they exist. The excitation is coming from the Lowther which is near the top and centered, so the coupling is weak with both mode shapes.
This is my thinking about why I am not worried about standing waves in the cavity formed by the front and two sides. I have seen this claim posted before and have to wonder how much is fact and how much is pure speculation stated as fact. I am more concerned about exciting vibration or motions of the panels in reaction to the woofer motions. |
|
|
| MJK |
| quote: | | Can you eleborate more on this? Is there any specific reason you like low Qts drivers more then high ones? |
Restating the question and having read the referenced post by Thorsten, I'll offer the following comments for discussion. These are my thoughts and by no means should they be the final word on anything, I am still learning.
I agree that a higher Qts driver performs better in an infinite baffle. I would also add that this is true if the driver is being used over an extended frequency range where the response above fs is being used. But my design idea is a little different for this speaker.
The Lowther drivers have low Qts due to the large magnet's strength which produces higher SPL.
PM2C = 97.0 dB/w/m
PM2A = 97.4 dB/w/m
DX4 = 98.1 dB/w/m
The Dayton woofers have a Qts of 0.37, a fs of 39.5 Hz, and an efficiency of 97 dB/w/m. A single Dayton woofer was not going to do the job in an OB paired with any of the Lowthers. So I used two wired in parallel and boosted the SPL even more using the digital crossover. So now the low frequency pair of drivers are much more efficient then the Lowthers. Then by selecting a crossover point between 100 and 200 Hz, and rolling the Daytons off at 12 dB/octave, I produced a almost symmetric hump like SPL response. Adding this to the Lowther, crossed over at 24 dB/octave, produced a flat combined response that went down into the 30 Hz to 40 Hz range depending on the crossover settings. This is not a minimalists approach and it does not use the drivers at an optimum condition, but I think the results work. It is a wasteful design by some people's rules just like my Lowther and BSC ML TL design.
I like low Qts drivers because they typically have big magnets and are very efficient. You can increase the effective Qts by adding series resistance tailoring the system SPL response. It is easy to increase the effective Qts of a driver but not so easy to reduce the Qts of an already high Qts driver. There is a lot of adjustment that can be done on a low Qts driver from the system level. Again, this goes against the typical use of Lowthers and some Fostex drivers and it is wasteful of the high efficiency in the raw driver. Guilty of non purist thoughts! But I believe it produces a very high quality speaker design, others might argue and they are welcome to approach the problem differently. |
|
|
| Rudolf |
| quote: | Originally posted by MJK
1/4 wave from front to back
f = c / (4 x 12 inches) = 282 Hz
1/2 standing wave between the sides
f = c / (2 x 24 inches) = 282 Hz
...
The other two frequencies are a little harder to convince myself they exist. The excitation is coming from the Lowther which is near the top and centered, so the coupling is weak with both mode shapes. |
MJK,
thanks for your detailed derivation. That 282 Hz is just in the range of resonance many H baffles have to cope with. And those resonances do exist:
http://www.linkwitzlab.com/models.htm#C
You state that you are crossing the Daytons between 100 and 200 Hz, and rolling off at 12 dB/octave. Taking you are crossing @150 Hz and the resonance peak is 9 dB high, that peak would still lurk with -3dB at 282 Hz, if not compensated.
I don´t say that this has to be detrimental to your design. I just was wondering whether those side panels in your OB design do serve any intended acoustic purpose.
Rudolf |
|
|
| MJK |
| quote: | | That 282 Hz is just in the range of resonance many H baffles have to cope with. And those resonances do exist |
I agree, but isn't his H baffle closed at the top? If it is then he has two short TL's and the resonances at 282Hz are a real concern. I did some work for John K modeling H and U baffles a few years ago and this is indeed what happens.
But if the top is left open then I am not sure that this resonance will be excited nearly as strongly, if at all, and the response peak will be much lower. The open top and back are the key variables that will determine what resonances exist in the cavity formed by the baffle and two sides. |
|
|
| Scottmoose |
| quote: | Originally posted by MJK
I agree, but isn't his H baffle closed at the top? |
Indeed they are. Not that the Orion is a bad speaker of course (I know you're not implying that, I hasten to add)... in fact, they are the best I've ever heard, from anything, and so they should be, cosidering the fact that they are far, far out of my price range, though very reasonable for what they achieve. Doesn't mean they can't be improved upon of course -I wonder what SL would think of your new MathCad OB worksheet once it's released -he's always looking for new ideas and ways to improve things. Perhaps we could tempt him onto the forum.
Best
Scott |
|
|
| MJK |
| Thinking about the open top some more, maybe it would be interesting to run an acoustic FEM model of the geometry to see what types of 3D modes and the corresponding frequencies would be predicted. Something for the detailed write up (when I get time) ....... |
|
|
| Rudolf |
| quote: | Originally posted by MJK
But if the top is left open then I am not sure that this resonance will be excited nearly as strongly, if at all, and the response peak will be much lower. The open top and back are the key variables that will determine what resonances exist in the cavity formed by the baffle and two sides. |
Martin,
thank you VERY much for that statement. I know I should have ripped the tops off my H baffles and measure the change by myself, but I´m the guy who needs enforcement by theory to do the obvious in real life. :rolleyes:
Let´s see who is first with published data: your FEM model or my measurements of a ripped H baffle. :D
Rudolf |
|
|
| MJK |
I am sure you will be first, my backlog of interesting stuff to do is huge.
If you rip the top off of your H frames, are you going to be left with a woofer in a very small baffle? I wonder if the 1/4 wave resonance, and effective distance betwwen the sources, is what helps produce the bass. If the baffle is small you may eliminate your objectional resonance at the expense of your bass output. |
|
|
| Rudolf |
Well, my current baffles actually are Z baffles.
What I will do is: build a provisional real H baffle with the same drivers (10") and same outer dimensions where I can screw the top on and off. That way I can see exactly to what extend I "eliminate your objectional resonance at the expense of your bass output". |
|
|
| Scottmoose |
In the case of your current Z baffles, you might loose some, (assuming that there is some loading going on), but probably not as much as with the drivers mounted vertically on an H baffle. In the Z, it looks to me as if you're going to get a boost from the floor-boundry condition, so the loss of the top might not be quite as significant. Just my tuppence worth.
Best
Scott |
|
|
| Rudolf |
Scott,
I see your argument.
The Z baffle without the top will behave even more unsymmetric and unpredictable than it does already. So results would be arguable. That´s why I will do a "formal" H baffle, hoping to get a better insight. Got some chipboard in the basement and hope I can find some time there too.;) |
|
|
| Scottmoose |
| I know that feeling... |
|
|
| thalis |
If other parameters could be kept unchanged, the low damped drivers (high q) are better for an OB regarding only the low freq cut off alone, since a -3dB point at the driver’s fs could be obtained if its Qts equals 0,7, but the problem with one dimension/parameter is that it can’t be changed alone, and there are also other parameters that one has to look after except physical low end extension. As Martin already said, low q also means (it comes with / it is a result of such a design) big motors, low mms and high sensitivity in general.
Q is also related with step response – a driver with a Qt of 0.3 will have a much better step response from a 0.7 one, in an OB alignment. Part of what is perceived as ‘quick, precise and without coloration bass’ in open baffle designs (another part comes from in box standing waves that pass through the cone material, and the smallest part comes from box construction resonations, as I believe) is a result of the lack of a resonating air volume which in every case worsens the driver’s free air step response (step response is depended on the driver/enclosure Q, which in the case of the OB is at about the driver’s Qts, and in the case of a box it is always bigger). If OB is used with a driver of 0,7 Qts then the resulting step response is even worse than that of a 0.3Qts driver in a closed 0.6Q system enclosure. Lastly, part of what is sometimes perceived as ‘thin bass’ in some circumstances is exactly because of a very low total Q and a resulting very high, subjectively ‘not physical’ damping, and perhaps extremely low Qts drivers should not be used in OB – but a value of about 0,4 as used here is a good compromise, I think.
So Martin just ads another woofer and gets his desired low frequency response and a better defined bass together with drivers adequate to extend well and also gets micro-details and micro-dynamics which also come with a low cone mass material and good motor, as well;-) As always, it is about tradeoffs and personal demands and likings.
Regards,
Thalis. |
|
|
| Rudolf |
I promised to do some measurements to look after the effect of pulling off the top cover from a H baffle.
First a picture of the assembly:
It shows two 15" drivers in a H frame of 82 cm height and 40 cm inner width. Depth is 20 cm in the back and 22 cm in the front.
I did measurements with the mike in the center of the back plane, 20 cm from the loudspeaker baffle.
The top cover was split in two halfs. One covering the front top and the other one covering the back. |
|
|
| Rudolf |
For MJK primarily:
The red line in the diagram shows the H baffle with the top cover on. Black line is with the back part of the top cover lifted off. It´s hardly recognisable, because the green line, showing both covers lifted off (as in the photo) is at almost the same position.
What I see is, that removing the top of a H baffle doesn´t change the character of the baffle much (in this assembly).
I know that I have to do some more measurements to complete the picture. So what measurements should I do next? I will have to confine measurements to the nearfield for the next sessions, so please don´t ask for more than 1 m distance.
BTW: SPL levels are not calibrated, but the mike is. |
|
|
| MJK |
| quote: | | What I see is, that removing the top of a H baffle doesn´t change the character of the baffle much (in this assembly). |
I agree. And to take it one step futher, there cannot be any standing waves between the sides of the baffle in any of the measurements because changing the end condition would have had an impact on the mode and the SPL response. So I conclude that in the frequency range you measured, I would not be worried about standing waves impacting the sound. This could also be checked by plotting the driver impedance where a standing wave would appear as a discontinuity (wiggle) in the amplitude and phase. |
|
|
|
