Fast, fun, Inexpensive OB project
Please go to post #455, http://www.diyaudio.com/forums/multi...#post2752617to see the latest version of this project
Some time ago last year a speaker building buddy of mine in California decided to see what he could do in a simple, inexpensive, open baffle project. He came up with a 2 way passive system on a medium size baffle. He has been trotting them out at shows across the country and baffling a lot of folks who would never have thought so simple a concept could sound so good.
For your pleasure, here are the build specs on the Manzanita 12s:
To build each speaker you will need:
Total project parts cost is under $150 US per side, excluding wood. Or less than $300 for the pair. The drivers and parts are easy to find in the US, and should be as easy to find in Europe. As long as you keep the crossover values the same, feel free to substitute your favorite brands. Obviously, you will use the same drivers!
There may be plenty of forum denizens who will poke holes in this design without ever building it. No worries, that’s part of the DIY sport. Those who do build it will enjoy it, holes or not.
Why build this speaker? A few reasons:
Why NOT build this speaker?
Really, this project is so easy you can start after lunch and be finished in time to kick back, and enjoy the blues and some brews before dinner. The speakers throw a large soundstage, don’t suffer typical box coloration and have good tonal balance. The recessed midrange makes them very easy to listen to. But the midrange fills in nicely in the typical living room that is often a little over bright. With a good solid state or tube amp of 20 watts or more, you’ll have plenty of power. Even an AMP6 10 watt T-Amp runs these speakers to quite respectable volumes. A 20 watt tube amps has no trouble, despite the low sensitivity.
Continued in next post.....
What can we learn by building this speaker?
First the baffles:
They are 15” wide by 27” tall. Chipboard, MDF, plywood etc. will do just fine. Mirror image tweeter placement for left and right.
Some sound deadening (simple automotive spray) should be put on the baffle back. Typically they sound balanced with the baffle face being from 30" - 38" from the rear wall and at least 36" from side walls. Only about 83 dB sensitivity at 1 kHz and -3 typically from around 40 Hz to 22 kHz on axis up to 14' of listening distance. You will want to elevate the baffle off the floor with a target height for the tweeter of 42", depending on your listening position.
Let’s look at the crossover.
Drivers are electrically in phase. Woofer exhibits a 2nd order acoustic roll off at the top and no roll off on the bottom down to the mid 30's in a small to medium size room. (up to 2800 cu ft – 80 cu meters or so). They will sound best near field (8' to 14'). The tweeter has a forth order acoustic roll off around 1350 Hz. A pair will deliver clean sound on most recordings up to around 94-98 dB.
Why the big inductor? A 12mH inductor on an 8ohm driver is a first order low pass of about 105Hz. That’s seems awfully low to cross to a dome tweeter, right? But the idea here is to tilt the woofer response down to make up for the bass losses suffered in an open baffle design. If left alone or crossed at a more normal point, the 12” driver would have very little bass on an open baffle of this size. We need to “boost the bass.” The only way to do that with a passive filter is attenuate the high end. That’s where that part about throwing away certain things comes in. We throw away efficiency in the midrange to keep the low-end balance; otherwise the tonal balance of the 12 would be far too bright.
The big inductor does this FR tilt for us by pushing down the top end response - it’s an old trick for open baffle speakers. If you’re not using a box for the bass, you have to give up something. In this case we trade away efficiency.
The crossover on the tweeter looks a little more normal, a simple 2nd order high pass with attenuation. But there is some staggering of values from textbook values. This stagger helps bring the tweeter into phase with the woofer, among other things. The 25R resistor pads the tweeter down to the woofer’s midrange level.
The side wings of the baffle are asymmetrical to help smooth frequency response. Tweeter placement is also chosen for smooth response and to mate well with the woofer. Note that left and right baffles should be mirror images. The tweeters will be toward the inside edge when the speakers are in their playing position.
There are refinements that can be made to the basic design without much trouble. The drivers can be flush mounted, if you have the tools and the skill. Flush mounting should help refine the midrange. Baffle edges could be rounded over, which should also smooth out the mids, but maybe at the expense of a little bass response. And of course the crossover could be tweaked, but it will take a lot of experimentation or experience to get it better.
The 12” Peerless will jump up and grab you from time to time. On a hot vocal or piano the woofer cone can be pushed into break-up. You’ll hear it, no mistake. But it’s not all the time and isn’t too distracting. Fans of the Mamboni and EnABL treatments will likely find fertile ground in the Peerless 12. I suspect the EnABL process would clear up just about the entire bad behavior of the woofer.
So I advise you to go for it. Build a pair with scrap wood, just to get a feel for it. If you love them, make a prettier baffle. If you don’t love them, well, you’ll still have a set of nice drivers for your next project.
Interesting design, I think the idea is very similar to where I have been going with my latest OB designs. If the spec's for the drivers are easily found on the Internet, it should be a simple simulation to run even including the passive crossover components. If I have time tonight I will see what I predict for a SPL response using one of my OB worksheet with passive crossovers.
Wow, thanks very much. I'm intrigued by your description of the Manzanita's. The low cost/ easy build aspect surely is attractive to me.
The recessed midrange and occasional breakup mode that you described may not be a problem after all, but it 'flies in the face' of the conventional stuff that I've read and you also commented on.
I just had an interesting thought... IF I did build it and wanted to experiment with the drivers later in a different design, ie a 3way, I could maybe copy Omni's (g'day mate) 'Frankie'!
Thanks 'Pano', I'm sure things are great in Kihei! best wishes , grant
I'd love to see what your sims show. There should be a pretty big hole in the mids, but you just don't notice it. Seems like average room response fills it in.
There was a bit of measurement done in the hotel room in Denver, dropped fast below the mids 30s, IIRC.
Hey Grant. Things are lovely here on the south shore of Maui, summer is over, so the weather is just right. :)
Yes, if you don't dig the project - but I think you will - you can always use the drivers elsewhere. They are both very nice. The Peerless is said to work well on a 3 cu ft sealed box or very well in a bigger ported box.
This is not a tooty-fruity, over thought design. It's simple, it works, it sounds great. Those who choose to build it will be surprised at how easy it is to get a great, musical speaker.
Here are links to the driver specs:
Seas 27FTDC Tweeter
Panomaniac, this is very interesting topic. Thank you.
I think that this project is for me :)
If I can find those XO parts, I will try Manzanita 12".
Sorry, my english is not perfect...
I have the simulation running but something does not look right. Basically the tweeter has an efficiency of about 75 dB/W/m with the series resistor. You specified "25 ohm 5W resistor", could this be 5 ohm 25 W? Is this a typo?
You should easily find all you need except maybe the big inductor. It's a hog. :)
Love the simplicity of this, my only concern is the efficiency, but you can't have it all....
I think 25 ohms is about right actually, if the 12mH causes a drop off of 6dB/oct, at the xover freq the woofer should be around 20 dB down. Of course this will be offset by the rising response beacause no zobel is used, the calculator here
gives 13.7dB attenuation using the quoted min Z of 6.5 ohms + 25ohms in series
5 ohms only gives around 5dB, which would probably be way too hot in this speaker...???
|All times are GMT. The time now is 09:34 AM.|
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio