Well, this is only my second speaker building attempt. My first ones, some basic towers using dirt cheap components, turned out... well... like crap (who woulda thunk). I'm passed that though; they only costed me about 30-40$ a piece (including MDF) so... yeah, not breaking the bank. Anyways, it was at the very least a learning experience for me. This time, I still don't want to break the bank, limiting the cost to under 150$ for a set of bookshelves. I've been reading up a lot lately on speaker design and I think I might have a winner here in my design. Take it easy on me though guys as this is only my second set of speakers and there's still a lot I don't understand. Anyways, without further ado, here are the parts I've chosen and decisions I've made.
Woofer:
https://www.madisoundspeakerstore.com/peerless-woofers-6-7/peerless-830657-6.5-sds-woofer/
Tweeter:
https://www.madisoundspeakerstore.c...ustics-sb19st-c000-4-3/4-dome-tweeter-4-ohms/
Crossover Frequency:
2500hz 2nd order linkwitz + zobel network
Let's start with the woofer. I chose this one because Zaph Audio gave its little brother (the 5.25" version) an extremely positive review but I wanted more sensitivity so I picked up the 6.5" version. I know just because the 5.25" was good, doesn't mean the 6.5" will be, but it was worth the risk I thought. I already have the woofers and have given them a listen and they sound loud and clean, however it's hard to know until I actually get these in a box and crossed over. Another thing about this woofer is that it has some mild come breakup starting at 4khz and ending around 6khz (about +6db peaks), however I didn't really notice it that much when I was testing the woofer. That's probably because of my relatively untrained ear, but hey ignorance is bliss I guess. Now let's talk about the tweeter. The tweeter looked good enough; that off axis resonse though...damn solid! Anyways, I mainly chose this one cause it matched the sensitivity of my woofer very well, making it so I don't need to buy an L-pad, adding about 20$ to the cost. Also, I've heard some good things about this tweeter on the Internet and think it's worth a try. So the drivers account for about 80$ of the cost (more than double that of my tower's drivers). The crossover parts account for about 50$ of the cost. Like I said above, I've settled on a crossover frequency of 2500hz @ -12db/octave. This might seem a little high, considering the cone breakup inherent in the woofer, but I really didn't notice any harshness or peakyness when I was testing it so I figured a 2nd order LW would do just fine. The tweeter's free-air resonance is 980hz, so the crossover will be more than double the fs which seems like the general rule of thumb, although I've also read 2-octaves above fs, but that's simply not realistic for me given my budget and drivers of choice. Besides, this crossover should yield around -15db or so at fs which seems sufficient. I've also designed, or rather used an online calculator, a zobel network to deal with the impedence rise in the woofer, fairly straight forward. Also, I should mention that I've thought of including a notch filter into the crossover to help with the cone breakup, but like I said it didn't seem that audible or noticeable to me, not to mention it's pretty mild given its a paper cone, so I'm going to wait on that for a while and add it later if it turns out to be necessary, but not right outta the gate. I guess I should end with the box design I've chosen. It's pretty straight forward, but the dimensions are 12x24x8" (outside) with 1/2" MDF and 1 1/4" wooden dowels to brace the box. I'll also be using s combination of fiberglass insulation and some leftover polyfill to help work out those box resonances, standing waves, and clean up the midrange. I should also mention it's a ported box tuned to 40-45hz (haven't decided yet) with a 2.75" port with a length that I don't remember off the top of my head right now. Also, I've made sure that the tweeter and woofer will be within 5.5" of eachother (center to center, the wavelength of 2500hz) to help blend the two and limit lobing.
Anyways, that about sums up my design behind this one. I kinda just want to hear other people's inputs before I got ahead and buy the tweeters and crossover parts (I only have the woofers). Keep in mind that I've got a pretty strict budget as a student, so don't recommend any options that are too expensive. Also feel free to share any design tips for the future, or that I could implement. Thanks guys!
EDIT: Also, I should mention that I was going to flush mount the tweeter because of the large 3.5" flange it has. Zaph from Zaph Audio said and has data that supports the idea of flush mounting tweeters, especially those with large, round flanges when it comes to diffraction's. However, I don't have a router to do this, so I'm buying a beauty board (birch) to go on the front that I can cut the right size hole into to flush mount the tweeter and is the exact thickness of the flange itself along with and quarter round moulding for the front edges to further help with diffraction.
the amount of times I said "also" in this most and "I should also mention" gave me a chuckle. I guess that's what happens when you post at 2:00 AM
Woofer:
https://www.madisoundspeakerstore.com/peerless-woofers-6-7/peerless-830657-6.5-sds-woofer/
Tweeter:
https://www.madisoundspeakerstore.c...ustics-sb19st-c000-4-3/4-dome-tweeter-4-ohms/
Crossover Frequency:
2500hz 2nd order linkwitz + zobel network
Let's start with the woofer. I chose this one because Zaph Audio gave its little brother (the 5.25" version) an extremely positive review but I wanted more sensitivity so I picked up the 6.5" version. I know just because the 5.25" was good, doesn't mean the 6.5" will be, but it was worth the risk I thought. I already have the woofers and have given them a listen and they sound loud and clean, however it's hard to know until I actually get these in a box and crossed over. Another thing about this woofer is that it has some mild come breakup starting at 4khz and ending around 6khz (about +6db peaks), however I didn't really notice it that much when I was testing the woofer. That's probably because of my relatively untrained ear, but hey ignorance is bliss I guess. Now let's talk about the tweeter. The tweeter looked good enough; that off axis resonse though...damn solid! Anyways, I mainly chose this one cause it matched the sensitivity of my woofer very well, making it so I don't need to buy an L-pad, adding about 20$ to the cost. Also, I've heard some good things about this tweeter on the Internet and think it's worth a try. So the drivers account for about 80$ of the cost (more than double that of my tower's drivers). The crossover parts account for about 50$ of the cost. Like I said above, I've settled on a crossover frequency of 2500hz @ -12db/octave. This might seem a little high, considering the cone breakup inherent in the woofer, but I really didn't notice any harshness or peakyness when I was testing it so I figured a 2nd order LW would do just fine. The tweeter's free-air resonance is 980hz, so the crossover will be more than double the fs which seems like the general rule of thumb, although I've also read 2-octaves above fs, but that's simply not realistic for me given my budget and drivers of choice. Besides, this crossover should yield around -15db or so at fs which seems sufficient. I've also designed, or rather used an online calculator, a zobel network to deal with the impedence rise in the woofer, fairly straight forward. Also, I should mention that I've thought of including a notch filter into the crossover to help with the cone breakup, but like I said it didn't seem that audible or noticeable to me, not to mention it's pretty mild given its a paper cone, so I'm going to wait on that for a while and add it later if it turns out to be necessary, but not right outta the gate. I guess I should end with the box design I've chosen. It's pretty straight forward, but the dimensions are 12x24x8" (outside) with 1/2" MDF and 1 1/4" wooden dowels to brace the box. I'll also be using s combination of fiberglass insulation and some leftover polyfill to help work out those box resonances, standing waves, and clean up the midrange. I should also mention it's a ported box tuned to 40-45hz (haven't decided yet) with a 2.75" port with a length that I don't remember off the top of my head right now. Also, I've made sure that the tweeter and woofer will be within 5.5" of eachother (center to center, the wavelength of 2500hz) to help blend the two and limit lobing.
Anyways, that about sums up my design behind this one. I kinda just want to hear other people's inputs before I got ahead and buy the tweeters and crossover parts (I only have the woofers). Keep in mind that I've got a pretty strict budget as a student, so don't recommend any options that are too expensive. Also feel free to share any design tips for the future, or that I could implement. Thanks guys!
EDIT: Also, I should mention that I was going to flush mount the tweeter because of the large 3.5" flange it has. Zaph from Zaph Audio said and has data that supports the idea of flush mounting tweeters, especially those with large, round flanges when it comes to diffraction's. However, I don't have a router to do this, so I'm buying a beauty board (birch) to go on the front that I can cut the right size hole into to flush mount the tweeter and is the exact thickness of the flange itself along with and quarter round moulding for the front edges to further help with diffraction.
the amount of times I said "also" in this most and "I should also mention" gave me a chuckle. I guess that's what happens when you post at 2:00 AM
Last edited:
Common guys. There's gotta be something someone wants to say about my build. I'm looking for someone with more experience than me to kinda look over my shoulder and make sure I'm doing the right things here cause, like I said, this is only my second attempt and I still don't fully understand a lot of things.
I've moved to multiway, as bound to get some input there.
Well, would a 3rd order crossover at 2100-2200hz fix that? Like I said, the woofers peakyness due to cone breakup really isn't that noticeable to me when playing it full range. I don't really have any equipment to measure it, but I think -12db at 5000hz (the highest peak +6db above the woofers sensitivity) would be sufficient. Isn't the biggest issue with cone breakup the harmonic distortion caused by it in the lower octaves? I really didn't notice much of this at all, so I figured the peaks must not be that severe. I would rather stick with the 2nd order LW crossover if possible because then the woofer and tweet are only 180 degrees outta phase which is easily corrected by wiring the tweeter inversely and yields virtually no lobing. A third order doesn't have any of those benefits; can I get away with the 2nd order?
Once you apply some Baffle Step Compensation (BSC), woofer and tweeter SPL aren't matched anymore and you need to pad down the tweeter. But a couple of resistors per crossover won't cost you 20$.
You need to suppress the breakup, otherwise it will give you listening fatigue. See here for an example of a notch with the same woofer: https://sites.google.com/site/undefinition/tarkus
I suspect that you, like many novices, tend to think only in terms of electrical slope of a crossover. It doesn't work like that, because any other roll-off due to natural behavior or diffraction will sum to the electrical slope giving easily an acoustic slope greater than the electrical slope. With the right crossover point and driver selection it is easy to accomplish an acoustic LR4 slope with a 2nd order electrical crossover, see here for an example: Zaph|Audio - ZA-SR71
Designing a crossover is not a trivial task, but it can accomplished with the right tools, and that include simulations. Have a look at post #20 here, but read also the posts before: http://www.diyaudio.com/forums/multi-way/189847-introduction-designing-crossovers-without-measurement.html
IMHO a box like yours needs at least 3/4" MDF + bracing.
Ralf
PS: have you worked out the reason(s) for your failure of the first attempt?
https://www.madisoundspeakerstore.c...bc25sc55-04-1-textile-dome-tweeter-truncated/
And another woofer for the baffle step as a .5
And another woofer for the baffle step as a .5
Hi,
Two tricky drivers to get right. The bassmid because of its peak,
the tweeter because of its impedance peak, and size meaning
it needs a higher x/o point than a typical 1" dome.
See the treble section of : Zaph|Audio - ZRT - Revelator Tower
and the bass section of : Zaph|Audio
Use some sim tools, implement BSC, aim for slightly assymetric
(to tilt the x/o intergration more on axis) 4th order L/R acoustic.
x/o point 2.5 to 3.0 KHz. FRD Consortium tools guide
See https://sites.google.com/site/undefinition/diy/amiga
and its related designs for a very nice suitable box design.
rgds, sreten.
Two tricky drivers to get right. The bassmid because of its peak,
the tweeter because of its impedance peak, and size meaning
it needs a higher x/o point than a typical 1" dome.
See the treble section of : Zaph|Audio - ZRT - Revelator Tower
and the bass section of : Zaph|Audio
Use some sim tools, implement BSC, aim for slightly assymetric
(to tilt the x/o intergration more on axis) 4th order L/R acoustic.
x/o point 2.5 to 3.0 KHz. FRD Consortium tools guide
See https://sites.google.com/site/undefinition/diy/amiga
and its related designs for a very nice suitable box design.
rgds, sreten.
Last edited:
Thanks everyone for the responses. I mostly understand what a lot of you guys are saying, but, like giralfino said, I am a novice and some of the jargon is going completely over my head. I guess what I'm saying is could you guys dumb it down for me a little. I'll definitely read up on those links and see if that helps it make more sense.
Another thing is I already have the woofers, but the other pieces can be changed out for something better/easy to work with. Any recommendations for replacement tweeters that would match the woofer well? Thanks!
Another thing is I already have the woofers, but the other pieces can be changed out for something better/easy to work with. Any recommendations for replacement tweeters that would match the woofer well? Thanks!
I don't remember saying this. Actually though, I was shooting for around a 1.2 cf box tuned to around 40hz which yields an f3 of around 41hz. I'm actually planning on using these as front replacements for my current towers in my home theater, which I coincidentally also use for music a lot hence why I'm looking for better stereo sound. Honestly I only need these things to go down to 60hz, but a little extra extension, when I don't want the subwoofer waking up the dorm above us, doesn't hurt
P.S. I use the modeling software called WinISD to design my boxes and SketchUp to plan my cuts and model what the box will look like.
Last edited:
Too eager to get something built and poor planning to boot. I didn't know a lot of the stuff I know now, although I only built those about 6 months ago, and it showed in the sound. Luckily my receiver had an EQ that I was able to use to get rid of a lot of the peaks and dips to make the sound acceptable, but it sounds absolutely horrible without it. Honestly the tweeter was the biggest downfall, a huge amount of sibilance (last time I use a cone tweeter). The woofers are actually really good given their cost and have a fairly flat response with only a small dip at 2k; they really only show their cost when you have the volume at -3db and a lot is going on.
Anyways, just poor execution and too eager to get something built. Funny thing is all the people in my dorm think they sound amazing.... but I know better and am not satisfied.
Since I'll have about a 9" baffle which means anything above 1500hz should be directed forward and the rest should wrap around right. I also read that its around half the sound pressure that is lost behind the speaker, so really I should putting a -3db pad on the tweeter right? Won't that mean that there will be a 3db peak on the woofer though from 1500hz to the crossover point then, or is the baffle only a 4.5" baffle because the tweeter is mounted in the middle making it 3000hz instead? So many questions... but I think I'm starting to understand.
Last edited:
From : http://zaphaudio.com/SR71.html
Hi,
The above shows the effect of the 9" wide madisound MD14 box.
For both an offset and centred tweeter. Sorry but I can't really
dumb it down, you have to do some proper simulations.
The other curve is the effect on the bassmid, note the hump.
Though not that great, this is cheap and easy to work with :
https://www.madisoundspeakerstore.c...5-1-fabric-dome-tweeter-54.5-mm-sq-chambered/
http://zaphaudio.com/tweetermishmash/compare.html
rgds, sreten.
Last edited:
I actually looked into that tweeter quite a bit, but Zaph audio didn't have all that much positive stuff to say about it so I shied away from it.
I understand what the graph is saying, but I'm just curious if that applies to all 9" wide baffles and tweeters or if there are significant differences from situation to situation. I would suspect that later. Either way, it definitely looks like I'll have to pad the tweeter. I'll try a -3db pad and increase it if need be; resistors are so cheap anyways. Also the response is +-3db (in relation to tweeter placement) regardless and isn't that considered pretty good actually? I'll probably just stick to center mounting it; chances are I won't be able to get the speakers perfectly placed to get proper imaging from the offset tweeter placement anyways.
Also, my baffle will have quarter round moulding as edges which should help with some of that tweeter diffraction right, basically smoothing out the response a little compared to that of the Madisound Pre-built Baffle?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.