Tweet
Introducing The Arias - my Black Box design. RS150-4 + XT25SC90-04
It is with great pride and minor disappointment that I show off my first design – the Arias. The pride comes from the fact that I’m extremely proud of these speakers, and I’m thrilled with how they sound and the overall result.
My disappointment stems only from the fact that I am introducing these on this board and not at InDIYana 2013. These were my entry to the Black Box design competition, but I was unable to attend the fest.
So, I present them here for your enjoyment and pleasure. I hope someone out there takes a chance on building these, because I feel they represent an extraordinary value and present the listener with a wall of superb sound.
My main design goal for my first set of speakers was to build something for my cousin who will be graduating high-school this
summer and will be attending college to study music. She’s an opera singer and was completely floored when she traveled to Austin for a visit last summer and I was able to show her what her operas sounded like on a great set of speakers.
So, I wanted to build her a set of speakers to give her the entry point into the world of hi-fi audio and enjoy the Arias she loves in the manner they were presented to the original audience. Hence the name - “Arias”.
I spent quite a bit of time on driver selection, trying to decide between paper or aluminum woofers, tweeter selection, etc. It was around the time I was making my driver choices when Wolf announced the Black Box competition, and I decided to make my design fit into the parameters of that competition.
It took quite a bit of research, simulation, and smart shopping, but I was able to fit just under the $150/pair requirement.
For the woofer, I had settled in on the Dayton RS150 because of the great reviews over on Zaph’s site and other places, the availability of FRD and ZMA files, and that I could sim the speakers down to the upper 40’s in a ported enclosure. The way the Black Box competition set up, I ended with the RS150-4 version to save a little cost. The RS150-4, at $35.62/ea, is about $4 cheaper than the RS150-8.
For the tweeter, I really wanted to try out a ring-radiator because of the way they are known for clarity and crispness. This was going to be important for the listening of female voices and classical music. I searched for a while, but the end choice was easy – the Vifa XT25SC90-04. I was willing to go for one of the larger siblings of this tweet, but the Black Box competition forced me to go slightly cheaper. It is available for $21.90 at Madisound, which is about $3 cheaper than Parts Express.
To build the cabinets, I wanted to stick within the Black Box competition parameters, but I also wanted to make this design easy to build. So, I settled in on the same dimensions as the Dayton 0.38 cu. ft. pre-made cabinets available at PE. These are 8”w by 14”h by 10.5”d, not including the grilles. I made my boxes out of ¾” MDF, but the design should easily work in those pre-made cabs as well.
The woofer is centered on the baffle, and the centers of the tweeter and port are 2.5” from the top and bottom respectively. All drivers and the port are centered horizontally. The baffle has a ½” roundover because that’s the router bit I had available. The port itself is made of 2” PVC from the plumbing section of a big-box store and is mounted half-way into the baffle from the back side. Then, I used a ¼” roundover bit to create a flare for the port opening. The total length from the back of the port to the front of the baffle is 6 ¾“. With this, the box measured out to be tuned at about 48Hz. This was a little lower than I had intended, but it works pretty darned well. I added two internal braces that are ¾” by 2”. One is mounted horizontally behind the tweeter and the other is mounted vertically behind the woofer. Both are centered on the side panels of the design. I’m not sure just how much these braces are doing inside the panels, but I had the leftover wood, so I added them. I did chamfer the back of the baffle around the woofer mounting screws, but I’m not sure that this is necessary.
The cabinet’s back wall, top, and some of the sides are lined with PE’s 260-516 acoustic foam. I purchased only one sheet of this for both cabinets, so that’s why the entire side walls are not covered – there wasn’t enough. I also added about 4 handfuls of Poly-Fill pillow stuffing to the cabinets. The front baffles are removable – I used triangular pieces of wood in the corners of the box to create a place for T-Nuts. The baffles then mount with flat-head furniture bolts I found at home depot. I recessed the heads of those bolts to be flush with the baffle. Note that the woofer and tweeter are also recessed flush. To connect to the outside world, I used PE’s 260-283 round terminal cups.
I simmed out the crossover with Jeff’s wonderful PCD tool, and came up with a LR4 design centered around 2500Hz or so. Doing some smart shopping landed me over on Erse Audio’s site, and I placed an order for a bunch of components, not knowing if the sims would hold up in real life or not. Well, they were close, but not quite right. Armed with my Deal-of-the-Day Omnimic, I set out to find the best combination of crossover point and reverse null.
Sparing you the details of all the combination of parts that I tried, I’ll simply provide you with the crossover values.
The woofer uses a simple second-order parallel topology with a tank circuit to tame the RS150’s breakup. The circuit is able to obtain an LR4 slope with the crossover point still centered right around 2500Hz. The inductor value required for the project is 1.4mH. This is NOT a common value, but it’s easy to obtain if you end up purchasing the same inductor I did. I used the Erse Audio 1.5mH ALQ 20AWG air coil (part number EAC13-20-1500), and I unwound the coil exactly 12 times. This number of un-windings resulted in both of my samples hitting the 1.4mH number. Granted, I was able to double-check my work using an Audio Precision, but the same procedure netted the same result twice. If you purchase a different coil than the one I mention above, your results may vary.
You may wonder why I chose the air-coil over an iron-core, and there are actually two reasons. First is cost. Again, I was staying within the Black Box rules. The second is the higher DCR of the air-coil provided a better integration between the woof and the tweet. The lower DCR of the iron-core (yes, I tried it) allowed the woofer to play louder and I was personally not as happy with how this sounded. Again, your mileage may vary, but this was my result.
I added a 0.22uF capacitor (Erse Pulse-X 630V, PN = MPX63-03-0.22) across the inductor as a “tank” circuit to really clamp down on the breakup of the woofer. I tried several other values of capacitors, but 0.22uF was by far the most effective at taming the breakup in my design. You may be able to get away with 0.20uF, but I know that 0.18uF does not provide the results I desired. The shunt capacitor value is 12uF (Erse PEx 250V, PN = MET25-05-12.0/PB). Other than the tank capacitor, I used mylar caps throughout to save costs. I have not tested with any other cap types, so your mileage may vary. In series with the shunt capacitor, I added a 1-ohm, 10W resistor (Erse EWR10-05-1.0/PB). This provides a resistive path for ultrasonic frequencies, and will help prevent some amplifiers from going unstable due to the tank circuit.
The tweeter circuit uses a third order parallel topology with an attenuation pad to obtain an LR4 slope. The circuit starts off with a 3.6 ohm series resistor (Erse EWR10-05-3.6/PB) to provide uniform attenuation throughout the tweeter band. Following is a series 6.0uF, 250V mylar cap (Erse MET25-05-6.00/PB) and a shunt 0.200 air-coil inductor (Erse EAC15-20-200). The third stage is another 6.0uF series capacitor (Erse MET25-05-6.00/PB). Finally, I placed a 30-ohm 10W resistor (Erse EWR10-05-30.0/PB) across the tweeter terminals to help tame the resonant frequency of the tweeter a bit.
I had to pay very careful attention to how I mounted the crossover components because the port for the box is very long, and basically takes up most of the center of the box, leaving very little room to work with. I mounted the components to a RadioShack perfboard and paid close attention so that I could mount them along the sides of the box’s port. The woofer’s air-coil inductor must stand up vertically, or it wouldn’t fit. Close attention must also be paid to the capacitors. Check out the picture to see how I did it. Note that the wire entering the center of the perfboard is the entry from the terminal cups.
In putting them together, the largest challenge came from determining which post of the tweeter was positive. I was only able to do it by watching the reverse null with the Omnimic.
Again, I hope someone gets around to building these. I’m very pleased by them and I feel that most anyone would be as well.
Thanks!
Tyger23
Cut list for a 14" by 48" sheet of MDF:
The crossover design:
The final measured design, showing a very deep reverse null. Note - I'm not sure I took the measurements right when I tried to sum the woofer and port responses, so I really don't trust anything below 300Hz in this picture. Maybe it's right, but I just don't feel comfortable saying it is without further measurements:
My disappointment stems only from the fact that I am introducing these on this board and not at InDIYana 2013. These were my entry to the Black Box design competition, but I was unable to attend the fest.
So, I present them here for your enjoyment and pleasure. I hope someone out there takes a chance on building these, because I feel they represent an extraordinary value and present the listener with a wall of superb sound.
My main design goal for my first set of speakers was to build something for my cousin who will be graduating high-school this
summer and will be attending college to study music. She’s an opera singer and was completely floored when she traveled to Austin for a visit last summer and I was able to show her what her operas sounded like on a great set of speakers.
So, I wanted to build her a set of speakers to give her the entry point into the world of hi-fi audio and enjoy the Arias she loves in the manner they were presented to the original audience. Hence the name - “Arias”.
I spent quite a bit of time on driver selection, trying to decide between paper or aluminum woofers, tweeter selection, etc. It was around the time I was making my driver choices when Wolf announced the Black Box competition, and I decided to make my design fit into the parameters of that competition.
It took quite a bit of research, simulation, and smart shopping, but I was able to fit just under the $150/pair requirement.
For the woofer, I had settled in on the Dayton RS150 because of the great reviews over on Zaph’s site and other places, the availability of FRD and ZMA files, and that I could sim the speakers down to the upper 40’s in a ported enclosure. The way the Black Box competition set up, I ended with the RS150-4 version to save a little cost. The RS150-4, at $35.62/ea, is about $4 cheaper than the RS150-8.
For the tweeter, I really wanted to try out a ring-radiator because of the way they are known for clarity and crispness. This was going to be important for the listening of female voices and classical music. I searched for a while, but the end choice was easy – the Vifa XT25SC90-04. I was willing to go for one of the larger siblings of this tweet, but the Black Box competition forced me to go slightly cheaper. It is available for $21.90 at Madisound, which is about $3 cheaper than Parts Express.
To build the cabinets, I wanted to stick within the Black Box competition parameters, but I also wanted to make this design easy to build. So, I settled in on the same dimensions as the Dayton 0.38 cu. ft. pre-made cabinets available at PE. These are 8”w by 14”h by 10.5”d, not including the grilles. I made my boxes out of ¾” MDF, but the design should easily work in those pre-made cabs as well.
The woofer is centered on the baffle, and the centers of the tweeter and port are 2.5” from the top and bottom respectively. All drivers and the port are centered horizontally. The baffle has a ½” roundover because that’s the router bit I had available. The port itself is made of 2” PVC from the plumbing section of a big-box store and is mounted half-way into the baffle from the back side. Then, I used a ¼” roundover bit to create a flare for the port opening. The total length from the back of the port to the front of the baffle is 6 ¾“. With this, the box measured out to be tuned at about 48Hz. This was a little lower than I had intended, but it works pretty darned well. I added two internal braces that are ¾” by 2”. One is mounted horizontally behind the tweeter and the other is mounted vertically behind the woofer. Both are centered on the side panels of the design. I’m not sure just how much these braces are doing inside the panels, but I had the leftover wood, so I added them. I did chamfer the back of the baffle around the woofer mounting screws, but I’m not sure that this is necessary.
The cabinet’s back wall, top, and some of the sides are lined with PE’s 260-516 acoustic foam. I purchased only one sheet of this for both cabinets, so that’s why the entire side walls are not covered – there wasn’t enough. I also added about 4 handfuls of Poly-Fill pillow stuffing to the cabinets. The front baffles are removable – I used triangular pieces of wood in the corners of the box to create a place for T-Nuts. The baffles then mount with flat-head furniture bolts I found at home depot. I recessed the heads of those bolts to be flush with the baffle. Note that the woofer and tweeter are also recessed flush. To connect to the outside world, I used PE’s 260-283 round terminal cups.
I simmed out the crossover with Jeff’s wonderful PCD tool, and came up with a LR4 design centered around 2500Hz or so. Doing some smart shopping landed me over on Erse Audio’s site, and I placed an order for a bunch of components, not knowing if the sims would hold up in real life or not. Well, they were close, but not quite right. Armed with my Deal-of-the-Day Omnimic, I set out to find the best combination of crossover point and reverse null.
Sparing you the details of all the combination of parts that I tried, I’ll simply provide you with the crossover values.
The woofer uses a simple second-order parallel topology with a tank circuit to tame the RS150’s breakup. The circuit is able to obtain an LR4 slope with the crossover point still centered right around 2500Hz. The inductor value required for the project is 1.4mH. This is NOT a common value, but it’s easy to obtain if you end up purchasing the same inductor I did. I used the Erse Audio 1.5mH ALQ 20AWG air coil (part number EAC13-20-1500), and I unwound the coil exactly 12 times. This number of un-windings resulted in both of my samples hitting the 1.4mH number. Granted, I was able to double-check my work using an Audio Precision, but the same procedure netted the same result twice. If you purchase a different coil than the one I mention above, your results may vary.
You may wonder why I chose the air-coil over an iron-core, and there are actually two reasons. First is cost. Again, I was staying within the Black Box rules. The second is the higher DCR of the air-coil provided a better integration between the woof and the tweet. The lower DCR of the iron-core (yes, I tried it) allowed the woofer to play louder and I was personally not as happy with how this sounded. Again, your mileage may vary, but this was my result.
I added a 0.22uF capacitor (Erse Pulse-X 630V, PN = MPX63-03-0.22) across the inductor as a “tank” circuit to really clamp down on the breakup of the woofer. I tried several other values of capacitors, but 0.22uF was by far the most effective at taming the breakup in my design. You may be able to get away with 0.20uF, but I know that 0.18uF does not provide the results I desired. The shunt capacitor value is 12uF (Erse PEx 250V, PN = MET25-05-12.0/PB). Other than the tank capacitor, I used mylar caps throughout to save costs. I have not tested with any other cap types, so your mileage may vary. In series with the shunt capacitor, I added a 1-ohm, 10W resistor (Erse EWR10-05-1.0/PB). This provides a resistive path for ultrasonic frequencies, and will help prevent some amplifiers from going unstable due to the tank circuit.
The tweeter circuit uses a third order parallel topology with an attenuation pad to obtain an LR4 slope. The circuit starts off with a 3.6 ohm series resistor (Erse EWR10-05-3.6/PB) to provide uniform attenuation throughout the tweeter band. Following is a series 6.0uF, 250V mylar cap (Erse MET25-05-6.00/PB) and a shunt 0.200 air-coil inductor (Erse EAC15-20-200). The third stage is another 6.0uF series capacitor (Erse MET25-05-6.00/PB). Finally, I placed a 30-ohm 10W resistor (Erse EWR10-05-30.0/PB) across the tweeter terminals to help tame the resonant frequency of the tweeter a bit.
I had to pay very careful attention to how I mounted the crossover components because the port for the box is very long, and basically takes up most of the center of the box, leaving very little room to work with. I mounted the components to a RadioShack perfboard and paid close attention so that I could mount them along the sides of the box’s port. The woofer’s air-coil inductor must stand up vertically, or it wouldn’t fit. Close attention must also be paid to the capacitors. Check out the picture to see how I did it. Note that the wire entering the center of the perfboard is the entry from the terminal cups.
In putting them together, the largest challenge came from determining which post of the tweeter was positive. I was only able to do it by watching the reverse null with the Omnimic.
Again, I hope someone gets around to building these. I’m very pleased by them and I feel that most anyone would be as well.
Thanks!
Tyger23
Cut list for a 14" by 48" sheet of MDF:
The crossover design:
The final measured design, showing a very deep reverse null. Note - I'm not sure I took the measurements right when I tried to sum the woofer and port responses, so I really don't trust anything below 300Hz in this picture. Maybe it's right, but I just don't feel comfortable saying it is without further measurements:
Comment