I have earlier tested a 6-1/2" Silver Flute in a 15L BR. As expected, needs to be supported with subwoofer. On its own, bass is a bit "short" due to early roll off.
This is an attempt to design a 2-way that has bass extension down to 40-45Hz using the concept of DCR.
Components
=========
Woofer = Silver Flute W17RC38-S-8
Tweeter = Seas 27TDFC
Gross Box Vol = 40L (Internally divided to 2 chambers)
Smaller Chamber = 13L
DCR tuned with 3 PVC Tubes. One in each chamber and the 3rd one between the 2 chambers.
PVC Tube = 2" Dia x 3.1" Length
Chambers lined with R11 Fiberglass wool.
This is an attempt to design a 2-way that has bass extension down to 40-45Hz using the concept of DCR.
Components
=========
Woofer = Silver Flute W17RC38-S-8
Tweeter = Seas 27TDFC
Gross Box Vol = 40L (Internally divided to 2 chambers)
Smaller Chamber = 13L
DCR tuned with 3 PVC Tubes. One in each chamber and the 3rd one between the 2 chambers.
PVC Tube = 2" Dia x 3.1" Length
Chambers lined with R11 Fiberglass wool.
Attachments
Deciding on the Box
T/S measurements done with LMS (20g Delta Mass). As can be seen, they are quite different from the manufacturer's specs.
In the box modeling, the black plot is how the response looks like when tuned to 37Hz in a BR of 40L. In a DCR, there will be a bass boost at F1 (37Hz) and F3 (80Hz). F3 is the port of the smaller chamber. F1 is in the bigger chamber where the woofer and tweeter are located.
There is another port, F2 (42Hz), that connects the 2 chamber within.
More information on Double Chamber Reflex can be found at:
Claudio Negro's home page
T/S measurements done with LMS (20g Delta Mass). As can be seen, they are quite different from the manufacturer's specs.
In the box modeling, the black plot is how the response looks like when tuned to 37Hz in a BR of 40L. In a DCR, there will be a bass boost at F1 (37Hz) and F3 (80Hz). F3 is the port of the smaller chamber. F1 is in the bigger chamber where the woofer and tweeter are located.
There is another port, F2 (42Hz), that connects the 2 chamber within.
More information on Double Chamber Reflex can be found at:
Claudio Negro's home page
Attachments
hi,
first off, kewl project....
i too have a set of the silver flute 6.5" 8 ohm
i have them in a set of Parts Express BR-1S (99 bucks, could not pass it up!!!) i took the shielded woofer out and put in the SF with a new crossover 3500 hz 1st order. the cabinet is like 15 liters, i plugged the port with a sock....
sounds wonderful, at least the woofer does, and yes it needs a sub.
how did you decide on the tweeter?? and what are you planning for a crossover? i am tweeter hunting right now and your input would be appreciated.
i think the silver flutes have potential.
first off, kewl project....
i too have a set of the silver flute 6.5" 8 ohm
i have them in a set of Parts Express BR-1S (99 bucks, could not pass it up!!!) i took the shielded woofer out and put in the SF with a new crossover 3500 hz 1st order. the cabinet is like 15 liters, i plugged the port with a sock....
sounds wonderful, at least the woofer does, and yes it needs a sub.
how did you decide on the tweeter?? and what are you planning for a crossover? i am tweeter hunting right now and your input would be appreciated.
i think the silver flutes have potential.
Hi mdynac.
I've always wanted a 6-1/2" 2-way to go down low without having to resort to a 2.5 or a sub. I believe this DCR can do just that.
The Silver Flute W17RC is a nice woofer. Very reasonably priced and generally good performance. Flat response until some cone breakup which is not too drastic. I don't think there are any other 6-1/2 that has similar response at this kind of price.
When I tested out this woofer earlier, I had it mated with a Vifa D25AG35 in a 15L BR, just to have an idea of its character. I crossed them initially at 2K5. Unfortunately, I picked out some distortion in the D25AG, even with 3rd order HP. I am sure it has something to do with the rear chamber. Doesn't seem to affect others but drive me nuts. Moved the crossover to 3K5. Distortion disappear but lost the mids.
To cross at 2K5, I will have to correct the D25AG issues passively or even mod the tweeter. I don't want to go to all the trouble and complicate things. I rather keep this design as simple as possible so that its easy for diy.
The Seas 27TDFC has the properties that I am looking for. Flat response, low Fs and being a newer design, I hope none of the resonance issues with the rear chamber.
The SF W17 will be crossed at 2K5 to the 17TDFC. The LP for the W17 will be 12dB. This will keep the cone breakup farther away from the cross point.
The 17TDFC HP will be 18dB. I like to keep the lows away from the tweeter. They are always more comfortable.
I've always wanted a 6-1/2" 2-way to go down low without having to resort to a 2.5 or a sub. I believe this DCR can do just that.
The Silver Flute W17RC is a nice woofer. Very reasonably priced and generally good performance. Flat response until some cone breakup which is not too drastic. I don't think there are any other 6-1/2 that has similar response at this kind of price.
When I tested out this woofer earlier, I had it mated with a Vifa D25AG35 in a 15L BR, just to have an idea of its character. I crossed them initially at 2K5. Unfortunately, I picked out some distortion in the D25AG, even with 3rd order HP. I am sure it has something to do with the rear chamber. Doesn't seem to affect others but drive me nuts. Moved the crossover to 3K5. Distortion disappear but lost the mids.
To cross at 2K5, I will have to correct the D25AG issues passively or even mod the tweeter. I don't want to go to all the trouble and complicate things. I rather keep this design as simple as possible so that its easy for diy.
The Seas 27TDFC has the properties that I am looking for. Flat response, low Fs and being a newer design, I hope none of the resonance issues with the rear chamber.
The SF W17 will be crossed at 2K5 to the 17TDFC. The LP for the W17 will be 12dB. This will keep the cone breakup farther away from the cross point.
The 17TDFC HP will be 18dB. I like to keep the lows away from the tweeter. They are always more comfortable.
Plot1 = DCR Impedance Sweep
=======================
As predicted, F2=44Hz and F3=85Hz.
Plot2
===
1) Purple = Raw Z of SF W17
2) Yellow = ZOBEL (11 ohms + 12uF) applied. Region of interest is from 1K - 10K.
Plot3
===
Raw Z of Seas 27TDFC. Fs is 550Hz, exactly as stated in the manufacturer's specs.
All plots swept with 2Vrms
=======================
As predicted, F2=44Hz and F3=85Hz.
Plot2
===
1) Purple = Raw Z of SF W17
2) Yellow = ZOBEL (11 ohms + 12uF) applied. Region of interest is from 1K - 10K.
Plot3
===
Raw Z of Seas 27TDFC. Fs is 550Hz, exactly as stated in the manufacturer's specs.
All plots swept with 2Vrms
Attachments
The third system (left to right) in C.Negro's page, is the one that work best with highish Qt, low Vas woofers. And as long as I know, can be modelled with LSPCad. But parameter insertion done manually, the automatic alignement routine will lead you off road. I suspect Leap can do the job as well...
I don't think there's any specifically for DCR. You can use any for BR programs to tune the box for your desired Fb. From the formula, you can calculate the F2 and F3.
Using LMS, I will measure the actual acoustic response of a DCR. It will give us an insight as to how different it is from a normal BR.
Far Field measurements
To referenced to 1 meter, it was determined that 30dB be subtracted from W17 and 38dB for F2 and F3.
W17 is now at 85dB, whereas the peaks of F2 and F3 are at 79dB.
It is interesting to note that while placing our ears at the mouth of the port is louder than at the center of the woofer, the reverse is true when we are 1 meter away. The outputs from the ports are 6dB less than the woofer.
With the plots now referenced to 1 meter, we can proceed with summing the outputs.
To referenced to 1 meter, it was determined that 30dB be subtracted from W17 and 38dB for F2 and F3.
W17 is now at 85dB, whereas the peaks of F2 and F3 are at 79dB.
It is interesting to note that while placing our ears at the mouth of the port is louder than at the center of the woofer, the reverse is true when we are 1 meter away. The outputs from the ports are 6dB less than the woofer.
With the plots now referenced to 1 meter, we can proceed with summing the outputs.
Attachments
Summing W17 with F3
From the plots, W17 avg spl is 85dB.
At 80Hz:
W17 is at 82dB (-3dB)
F3 is at 79dB (-6dB)
Looking at the smaller chamber response again, we can conclude that it is essentially a bandpass with a 3dB peak centered at 80Hz. Unlike a conventional bandpass sub where we have a woofer firing into a chamber, it is now replaced by a port instead.
Let's see how well they integrate.
From the plots, W17 avg spl is 85dB.
At 80Hz:
W17 is at 82dB (-3dB)
F3 is at 79dB (-6dB)
Looking at the smaller chamber response again, we can conclude that it is essentially a bandpass with a 3dB peak centered at 80Hz. Unlike a conventional bandpass sub where we have a woofer firing into a chamber, it is now replaced by a port instead.
Let's see how well they integrate.
Attachments
Minimum Phase + Summation
To sum the two plots, minimum phase must first be generated. LMS then calculates the new response when the 2 plots are combined to one.
The green plot is the new acoustic output at 1W1M (anechoic). We can now evaluate the contributions of the the smaller chamber.
There is quite a fair bit of reinforcement below 100Hz. The bandpass added about 3dB to W17. It also extended the F3 from 70Hz to 52Hz. And at 80Hz, it gave W17 a slight bass boost of 2dB.
All in, very impressive. No extra cost for another woofer. No complications with additional crossovers.
To sum the two plots, minimum phase must first be generated. LMS then calculates the new response when the 2 plots are combined to one.
The green plot is the new acoustic output at 1W1M (anechoic). We can now evaluate the contributions of the the smaller chamber.
There is quite a fair bit of reinforcement below 100Hz. The bandpass added about 3dB to W17. It also extended the F3 from 70Hz to 52Hz. And at 80Hz, it gave W17 a slight bass boost of 2dB.
All in, very impressive. No extra cost for another woofer. No complications with additional crossovers.
Attachments
DCR Final Response - Summing F2
Using the same procedures, F2 is integrated to W17+F3.
The yellow plot is the final response of this DCR. It can be seen that F2 extends the -3dB point from 52Hzto 43Hz.
In my opinion, this DCR works well for the SF W17. With -3dB all the way down to 43Hz, bass sounds more natural. There's also an added benefit of a slight bass boost (2dB) at 80Hz. Bass will have a bit more punch.
Now with the DCR sorted out, we can turn our attention to the crossover.
Using the same procedures, F2 is integrated to W17+F3.
The yellow plot is the final response of this DCR. It can be seen that F2 extends the -3dB point from 52Hzto 43Hz.
In my opinion, this DCR works well for the SF W17. With -3dB all the way down to 43Hz, bass sounds more natural. There's also an added benefit of a slight bass boost (2dB) at 80Hz. Bass will have a bit more punch.
Now with the DCR sorted out, we can turn our attention to the crossover.
Attachments
Evaluating the SF W17
In our previous Near Field measurement (left), we can see the plot is accurate up to 500Hz. The region of interest is from 100Hz to 500Hz. It is heartening to see its flat. This can be a very troublesome region. Too much and vocals will sound nasal.
Beyond 500Hz, the plot rolls off. This is not due to the woofer but rather the limitations of Near Field measurements.
Fortunately, Gated Measurements (right) allow us to see accurately the response from 500Hz onwards.
From 500Hz to 1K, the W17 maintains its flatness, followed by a gentle roll off and finally cone breakup.
Bearing in mind the tweeter is a Seas 27TDFC, the lowest crossover point that it would be comfortable is about 2K5. This is only one octave away from 5K, the center of the cone breakup. It would be nice to cross at 1K5 but its asking too much of a dome. I will have to resort to compression drivers at that region.
We can deduce from the 2 plots that the W17 is virtually flat from 100Hz to 1K with no lift beyond. BSC may not be necessary.
In our previous Near Field measurement (left), we can see the plot is accurate up to 500Hz. The region of interest is from 100Hz to 500Hz. It is heartening to see its flat. This can be a very troublesome region. Too much and vocals will sound nasal.
Beyond 500Hz, the plot rolls off. This is not due to the woofer but rather the limitations of Near Field measurements.
Fortunately, Gated Measurements (right) allow us to see accurately the response from 500Hz onwards.
From 500Hz to 1K, the W17 maintains its flatness, followed by a gentle roll off and finally cone breakup.
Bearing in mind the tweeter is a Seas 27TDFC, the lowest crossover point that it would be comfortable is about 2K5. This is only one octave away from 5K, the center of the cone breakup. It would be nice to cross at 1K5 but its asking too much of a dome. I will have to resort to compression drivers at that region.
We can deduce from the 2 plots that the W17 is virtually flat from 100Hz to 1K with no lift beyond. BSC may not be necessary.
Attachments
Crossing the W17
With the W17 impedance equalized by the Zobel, we can apply textbook formulas for the LP. I intend to keep the crossover network simple, yet effective. To avoid major phasing issues, it will be asymmetric - 12dB for woofer; 18dB for tweeter.
The Red plot is with a 12dB Butterworth at 2K5 (0.7mH + 6uF)
The peak of the cone breakup is now attenuated by 11dB. On first glace, it looks a lot but taking 85dB as the avg spl, it is only -8dB. Would be nice to have it down to -12dB from 85dB because this is one of the causes of harshness. Hopefully, it will be benign at -8dB.
With the W17 impedance equalized by the Zobel, we can apply textbook formulas for the LP. I intend to keep the crossover network simple, yet effective. To avoid major phasing issues, it will be asymmetric - 12dB for woofer; 18dB for tweeter.
The Red plot is with a 12dB Butterworth at 2K5 (0.7mH + 6uF)
The peak of the cone breakup is now attenuated by 11dB. On first glace, it looks a lot but taking 85dB as the avg spl, it is only -8dB. Would be nice to have it down to -12dB from 85dB because this is one of the causes of harshness. Hopefully, it will be benign at -8dB.
Attachments
Asymmetric Crossover
In the first display, the green plot is the raw gated response of the 27TDFC. The purple plot is with 18dB/oct Butterworth applied. This filter was tweaked to remove some peaking at the "knee". Hence, the 100uF.
In the middle display, we can see the true acoustic summing by the crossover. The red plot is with both LP and HP connected.
What is of prime importance, apart from the targeted crossover frequency, is the region before and after. This network sums nicely as there are no cancellation or peaking on either side of 2K6.
The full crossover is shown in the display on the right. During listening test, I decided to apply a very light BSC (-1dB) to the woofer. This made the bass and vocals more distinct.
This crossover is not exclusive to a DCR. It can be used for a BR, Sealed Box and others.
In the first display, the green plot is the raw gated response of the 27TDFC. The purple plot is with 18dB/oct Butterworth applied. This filter was tweaked to remove some peaking at the "knee". Hence, the 100uF.
In the middle display, we can see the true acoustic summing by the crossover. The red plot is with both LP and HP connected.
What is of prime importance, apart from the targeted crossover frequency, is the region before and after. This network sums nicely as there are no cancellation or peaking on either side of 2K6.
The full crossover is shown in the display on the right. During listening test, I decided to apply a very light BSC (-1dB) to the woofer. This made the bass and vocals more distinct.
This crossover is not exclusive to a DCR. It can be used for a BR, Sealed Box and others.
Attachments
Thank you all for the recommendations on modelling the DCR. I am in a bit of a hurry, but I will pursue all the recommendations.
Incidentally, as a side note, does anyone have any idea of a simplified lumped parameter schematic of this enclosure? Just as a lark.
Incidentally, as a side note, does anyone have any idea of a simplified lumped parameter schematic of this enclosure? Just as a lark.
- Status
- Not open for further replies.