View Full Version : DIY Speaker Project
DaveInVA
05-11-2011, 11:26 AM
http://www.braveknight.com/images/New8Speakers.jpg
http://www.braveknight.com/images/New8Done.jpg
Over the years I've accumulated a lot of drivers from other projects that either never got done or got modified etc. As I am on a limited income and a cheapskate by nature and have enjoyed building speakers since I was 14 I thought I'd build some new speakers for my home "office" using mostly stuff I have kicking around already.
A carpenter friend of mine built the cabinets for me to spec. 1.72 cf after accounting for the space the drivers etc take up. That gives a port tuning frequency of 30.58 hz with a 6" 2.5" port using the Dayton DSC200-8 8" drivers I had left over from other stuff. At first I was going to go 2 way but I didn't have any tweeters that could go low enough to mate with the 8" within its smooth area of response. I did have a pair of Dayton 2" Dome midranges so I mated those with a pair of Audax 1" textile dome tweeters.
The cabinets are a true 1" thick MDF and the speaker holes are rabbeted so the drivers mount flush to help eliminate edge refractions and the edges of the front baffle are rounded over for the same reason. I usually use rock wool for cabinet packing but I dont have any and nobody in town carries it anymore. It seems to work better for reducing back waves at lower frequencies than poly batting which is what I ended up using. I did line the inside of the cabinets with Bitumen though.
After some serious subjective listening and comparing with my DCM Time Windows I did some tweaks and some response measurements. I ended up reducing the the midrange/tweeter crossover point from 5000 hz down to 3000 hz (that eliminated a grittiness the 2" dome had) and padded the midrange by 4.5 db and the tweeter by 3 db. It now has a very smooth overall response. Sounds extremely nice with no hint of harshness in the vocal range and very non fatiguing. These Dayton 8's have always been a favorite of mine because of their performance at a good price. Very dynamic sound and deep clean bass. Very nice sounding project for little money. After about 20 minutes of AB'ing with the DCM Time Windows The Time Windows got relegated to my "junk" room for the first time since I bought them back around '74.
Dave
Eminence1963
05-11-2011, 11:38 AM
Over the years I've accumulated a lot of drivers from other projects that either never got done or got modified etc. As I am on a limited income and a cheapskate by nature and have enjoyed building speakers since I was 14 I thought I'd build some new speakers for my home "office" using mostly stuff I have kicking around already.
A friend of mine built the cabinets for me to spec. 1.72 cf after accounting for the space the drivers etc take up. That gives a port tuning frequency of 30.58 hz with a 6" 2.5" port using the Dayton DSC200-8 8" drivers I had left over from other stuff. At first I was going to go 2 way but I didn't have any tweeters that could go low enough to mate with the 8" within its smooth area of response. I did have a pair of Dayton 2" Dome midranges so I mated those with a pair of Audax 1" textile dome tweeters.
The cabinets are a true 1" thick MDF and the speaker holes are rabbeted so the drivers mount flush to help eliminate edge refractions and the edges of the front baffle are rounded over for the same reason. I usually use rock wool for cabinet packing but I dont have any and nobody in town carries it anymore. It seems to work better for reducing back waves at lower frequencies than poly batting which is what I ended up using.
After some serious subjective listening and comparing with my DCM Time Windows I did some tweaks and some response measurements. I ended up reducing the the midrange/tweeter crossover point from 5000 hz down to 3000 hz (that eliminated a grittiness the 2" dome had) and padded the midrange by 4.5 db and the tweeter by 3 db. It now has a very smooth overall response. Sounds extremely nice with no hint of harshness in the vocal range and very non fatiguing. These Dayton 8's have always been a favorite of mine because of their performance at a good price. Very dynamic sound and deep clean bass. Very nice sounding project for little money. After about 20 minutes of AB'ing with the DCM Time Windows The Time Windows got relegated to my "junk" room for the first time since I bought them back around '74.
Dave[/QUOTE]
*bigthumbup* Nice speakers and a nice writeup to boot*Hi5*
I think Dayton has some nice sounding drivers for the price*bigthumbsup*
Putting those DCM's to rest, *eyepop*should say something about the Dayton Brand*grin*
DaveInVA
05-11-2011, 12:00 PM
Thanks! Dayton does have some great values. I dont however like their 1.125" soft dome tweeters though. These Audax units are much better and they were only a few $ more. I also didn't think the difference would be so vast between these and the DCM's either. They are no slouches for what they are. These new speakers are more dynamic. The DCM's main thing to me was their ultra smooth midrange but they were always a little weak in the bass area and needed subs to sound balanced.
Dave
*bigthumbup* Nice speakers and a nice writeup to boot*Hi5*
I think Dayton has some nice sounding drivers for the price*bigthumbsup*
Putting those DCM's to rest, *eyepop*should say something about the Dayton Brand*grin*[/QUOTE]
Socal Sam
05-11-2011, 09:25 PM
Speaker project! Interesting layout. Mind telling us about your design?
I'd like to know:
Why the mirror image?
What are your crossover points?
How is the response at the crossover boundaries?
Efficiency?
Overall Freq Resp.?
Nominal impedance?
Inquiring mind(s) want to know.*Hi5*
DaveInVA
05-12-2011, 07:37 AM
Speaker project! Interesting layout. Mind telling us about your design?
I'd like to know:
Why the mirror image?
What are your crossover points?
How is the response at the crossover boundaries?
Efficiency?
Overall Freq Resp.?
Nominal impedance?
Inquiring mind(s) want to know.*Hi5*
I think mirror imaging in this type of speaker gives superior imaging.
As far as the layout I wanted to get the midrange and tweeter as close together as possible for better near field integrity. The woofer doesn't matter as much as the wavelengths coming from it are much longer.
At first just to test it out I used a pair of Dayton crossovers I'd had bougth years ago and used briefly in another project but didn't like their sound. In fact ever since I always build my own crossovers now. It has 675/5000 hz xover points.
Even though 5000 hz was well in the recommended range of the midrange there was a brittle grit I didn't like so lowering it to 3000 hz tamed it. I built new crossovers using Motor Run caps and ferrite chokes for the woofers. I used what I had and the ferrite core chokes have a much lower DCR than the air core chokes I used for the rest of the drivers. I used Polypropylene caps for everything else.
The tweeter and midrange are more efficient then the woofer (87db) so I used an old rheostat ahead of the xover network components for the tweeter and the midrange and adjusted so that the output was equal at the new xover points (800 & 3000hz). I measured the rheostat after each measurement and replaced it with the closest resistor ahead of the xover components for that driver.
The woofer and tweeter are 8 ohms nominal but the midrange was 6 ohms so I used a 2 ohm series resistor to keep it closer to 8 ohms to match the xover.
So the nominal impedance is 8 ohms and about 87 db efficiency. The port tuning frequency is 30.58 and its very nearly flat to that. There is even significant output down to about 28 hz. These Dayton woofers are a very good value and performance.
The Audax tweeter is rated at about +/- 2b to 22khz. A manual sweep with a signal generator and calibrated mic at 1 meter showed no big peaks or deviations across all the xover points.
I will also put felt rings around the tweeter and midrange to further reduce edge and baffle defraction. I am also building a sub that will double as a stand that will use 2 of these 8" woofers in a 3.44 cf cabinet and run them to about 180 hz. This will keep the excursion down on the single 8" also. These drivers are only $20 each when you buy 4.
Dave
Socal Sam
05-12-2011, 09:19 PM
I think mirror imaging in this type of speaker gives superior imaging.
As far as the layout I wanted to get the midrange and tweeter as close together as possible for better near field integrity. The woofer doesn't matter as much as the wavelengths coming from it are much longer.
At first just to test it out I used a pair of Dayton crossovers I'd had bougth years ago and used briefly in another project but didn't like their sound. In fact ever since I always build my own crossovers now. It has 675/5000 hz xover points.
Even though 5000 hz was well in the recommended range of the midrange there was a brittle grit I didn't like so lowering it to 3000 hz tamed it. I built new crossovers using Motor Run caps and ferrite chokes for the woofers. I used what I had and the ferrite core chokes have a much lower DCR than the air core chokes I used for the rest of the drivers. I used Polypropylene caps for everything else.
The tweeter and midrange are more efficient then the woofer (87db) so I used an old rheostat ahead of the xover network components for the tweeter and the midrange and adjusted so that the output was equal at the new xover points (800 & 3000hz). I measured the rheostat after each measurement and replaced it with the closest resistor ahead of the xover components for that driver.
The woofer and tweeter are 8 ohms nominal but the midrange was 6 ohms so I used a 2 ohm series resistor to keep it closer to 8 ohms to match the xover.
So the nominal impedance is 8 ohms and about 87 db efficiency. The port tuning frequency is 30.58 and its very nearly flat to that. There is even significant output down to about 28 hz. These Dayton woofers are a very good value and performance.
The Audax tweeter is rated at about +/- 2b to 22khz. A manual sweep with a signal generator and calibrated mic at 1 meter showed no big peaks or deviations across all the xover points.
I will also put felt rings around the tweeter and midrange to further reduce edge and baffle defraction. I am also building a sub that will double as a stand that will use 2 of these 8" woofers in a 3.44 cf cabinet and run them to about 180 hz. This will keep the excursion down on the single 8" also. These drivers are only $20 each when you buy 4.
Dave
I'm always humbled by the thought process that goes into home brews. Your project is well reasoned and developed.*bigthumbup* How would you describe the sound you are getting? Maybe if you can compare to a JBL 4311 or Pioneer HPM-100 that everyone has heard, that would help us place your speaker.
Your comment at the end about "keeping excursion down" is also very interesting. For a single driver, EMF in the wire translates to a certain amount of VC motion less loss due to heat, induction, etc. If the signal is split between two identical drivers, is the signal split in half such that each VC moves only half as much?
DaveInVA
05-13-2011, 07:10 AM
I'm always humbled by the thought process that goes into home brews. Your project is well reasoned and developed.*bigthumbup* How would you describe the sound you are getting? Maybe if you can compare to a JBL 4311 or Pioneer HPM-100 that everyone has heard, that would help us place your speaker.
Your comment at the end about "keeping excursion down" is also very interesting. For a single driver, EMF in the wire translates to a certain amount of VC motion less loss due to heat, induction, etc. If the signal is split between two identical drivers, is the signal split in half such that each VC moves only half as much?
I've not heard the JBL's or Pioneers in a long time so I can only do actual comparison between them and my DCM Time Windows that I have owned since new. I am used to listening to electrostats so I find very few cone speakers I like because I am used to a very clean midrange. Unlike the electrostats all box speakers have some degree of "cabinet sound" In addition to listening do the drivers you are also listening to cabinet resonances and defractions etc. The same drivers will sound different in different cabinets even if they are the same dimensions but different materials, bracing etc. The Time Windows have a very clean mid range and produce virtually no listening fatigue for me. Their week point is the bass and mid bass. They need subs to sound balanced and thought they sound really good they just don't have the dynamic mid bass I like. I was not trying to build a giant killer from these new speakers as I built them around leftover parts from other projects. I almost never build 3 way speakers because I don't like having a crossover point in the midrange but I didn't have any tweeters on hand that went low enough for the woofer. The woofer can go high enough but not without imparting a chestiness and hollowness to the sound do to the mass of the driver and probably some cabinet resonances. I still need to play with the crossover point between the mid and tweeter some more. The sound is very dynamic. Being used to electrostats I am very sensitive to things like slow bass, chestiness in vocals (and other cabinet resonances) and harshness in the midrange, especially in female vocals. I've listened to these for a couple of days straight now and they have caused me no listening fatigue at all yet. Vocals are well balanced, mid bass tight and no desire to move the tone controls off "flat" on the Sony STR-7055. The only complaint is even though the bass is very authoritative and tight if I really crank it the single 8 just isn't enough so the sub cabinet with 2 matching 8's will blend in seamlessly and take a load off the main cabinet. They don't have a dark sounding midrange like many vintage speakers have to me. I also have a pair of EPI 100's
When ever you double identical drivers you get 3 db more efficiency so each cone has to travel less. That is one of the pluses for a line array because the cones are staying in a lower distortion region of cone movement.
I am going to use 2 of these 8's in a separate cabinet for just under about 180 hz and then put a high pass on the current main cabinets so it's single 8 will not have the cone travel from the bass and should help it produce mid bass cleaner.
On my electrostat system I have 2 mid bass towers I built that each have 8 - 5.25" drivers and they only do 60 to 230 hz. Using that many small light weight drivers instead of a single heavier large diameter unit still gives a large total cone area to move lots of air but it does it faster so mid bass impact is more realistic. Also the drivers are barely moving during all but the most demanding music. I built these to replace the Time Windows on my home office system and I had lots of leftover parts to choose from and I needed a stress reducing project.
Dave
Socal Sam
05-13-2011, 11:49 AM
I can tell you are as unhappy with "store bought" speakers as I am. I like dynamic drivers. I like full range drivers but there is not enough range to satisfy like multi-driver setups. I am also interested in zero phase and frequency shift which full rangers, electrostats, and magneplanars deliver. None of them have the impact or immediacy of a dynamic three way that can push some air. That being said, I have yet to hear a three-way box speaker that I am truly happy with. Unhappiness leads to change so my new interest in DIY. The search goes on.
DaveInVA
05-13-2011, 12:06 PM
I can tell you are as unhappy with "store bought" speakers as I am. I like dynamic drivers. I like full range drivers but there is not enough range to satisfy like multi-driver setups. I am also interested in zero phase and frequency shift which full rangers, electrostats, and magneplanars deliver. None of them have the impact or immediacy of a dynamic three way that can push some air. That being said, I have yet to hear a three-way box speaker that I am truly happy with. Unhappiness leads to change so my new interest in DIY. The search goes on.
http://www.braveknight.com/images/ESLMIDSUB.jpg
Thats why I did my main system this way. I get the best of both worlds. The 5.25" drivers are fast enough to blend pretty seamlessly with the Electrostats at 230 hz and the Dayton Titanics 15's only come in at 60 hz. Now I just need a bigger room. In my last house they were in a room 22 X24 ft with a 12' ceiling. This 18 ft Octagon shaped room sucks big time. I don't really have any rooms decent for audio here..
I am also unhappy with factory made speakers also though I wouldn't mind owning Martin Logan Statements. But as I can never afford anything like that I built my own DIY version.
DIY speaker building is the way to go. Not just for the cost savings but you can tailor it the way you want it. To me anyways its also a lot of fun and you learn a lot, even from the ones that dont turn out as planned.
Dave
Socal Sam
05-13-2011, 04:09 PM
http://www.braveknight.com/images/ESLMIDSUB.jpg
Thats why I did my main system this way. I get the best of both worlds. The 5.25" drivers are fast enough to blend pretty seamlessly with the Electrostats at 230 hz and the Dayton Titanics 15's only come in at 60 hz. Now I just need a bigger room. In my last house they were in a room 22 X24 ft with a 12' ceiling. This 18 ft Octagon shaped room sucks big time. I don't really have any rooms decent for audio here..
I am also unhappy with factory made speakers also though I wouldn't mind owning Martin Logan Statements. But as I can never afford anything like that I built my own DIY version.
DIY speaker building is the way to go. Not just for the cost savings but you can tailor it the way you want it. To me anyways its also a lot of fun and you learn a lot, even from the ones that dont turn out as planned.
Dave
Nice Dave! I can tell you have been on the Yellow Brick Road for a long time. I'm just getting started with DIY and have three pairs done already. My fourth pair will be a 12 inch Yamaha and 5 inch Fostex full range 2-way. Like you, I have all sorts of drivers, parts, and cabinets to play with. I'm experimenting with cabinet porting, dipole, and full range drivers with passive subs and tweeters. I'm not at all knowledgeable about crossover design but I do have some working theories that I think are at the fringe.
For example, I am treating the woofer as an air foil by equalizing the compression moments on both sides of the cone. IMO, any cab that constricts air flow causes air pressure change which changes the substrate through which sound waves travel. (Of course, this is how a port tube amplifies bass but I will not be using a port tube.) This is not an entirely original strategy, Jamo hints at this in the R-909 whitepaper but does not go into detail because with OB, there is no need.
vBulletin® v3.8.5, Copyright ©2000-2012, Jelsoft Enterprises Ltd.