TLS Guy

TLS Guy

Seriously, I have no life.
im sorry, i think that was my dyslexia kicking in there on my interior volume. its .56cuft not .65cuft. i believe that i have finally come up with my final dimensions...(ODs)

H: 17.5"
W: 7.5"
D: 11.5"

2.25" round port 6" deep for a port velocity of .08
port will be rear facing right behind the tweeter

tweeter will be changed to Vifa's BC25SG15-04 so i can achieve a smoother frequency band and an easier crossover point of 2khz.

crossover will be 2nd order Butterworth as shown.



C1: 15uF
C2: 6.8uF
L1: .45mH
L2: .09mH

i hope i got it this time...:p
I will check it later. Your crossover looks far too simplistic. I think the woofer slope will likely have to be third order. If a second order works, Linkwitz/Riley constant voltage filters are superior to Butterworth constant power slopes under domestic conditions.
 
A

ameuba10

Enthusiast
3rd order Butterworth


C1: 13.26uF
C2: 39.79uF
C3: 13.26uF
L1: .24mH
L2: .95mH
L3: .32mH
 
TLS Guy

TLS Guy

Seriously, I have no life.
3rd order Butterworth


C1: 13.26uF
C2: 39.79uF
C3: 13.26uF
L1: .24mH
L2: .95mH
L3: .32mH
If only life were that simple, but it isn't. I will model it tomorrow. Be patient there is a lot of work involved.
 
A

ameuba10

Enthusiast
since 2.25" tubes are hard to find ill go with 2.5" diameter 7.5" long getting .06 port velocity...
 
TLS Guy

TLS Guy

Seriously, I have no life.
since 2.25" tubes are hard to find ill go with 2.5" diameter 7.5" long getting .06 port velocity...
With flush ends for that box the length of a 2.5 diameter port is 8.7". If you flare both ends which is optimal then the length is 9.7" Vent velocity is 24 m/sec
 
highfigh

highfigh

Seriously, I have no life.
since 2.25" tubes are hard to find ill go with 2.5" diameter 7.5" long getting .06 port velocity...
If you're using WinISD for this, highlight the port diameter and just enter 2, 3 or 4 as the diameter. The original program was written for Metric, which is why the odd diameters come up. It will give you the correct length for any diameter you enter.
 
highfigh

highfigh

Seriously, I have no life.
I will check it later. Your crossover looks far too simplistic. I think the woofer slope will likely have to be third order. If a second order works, Linkwitz/Riley constant voltage filters are superior to Butterworth constant power slopes under domestic conditions.
Mark- since you already modeled the 830874 in the NASP, it's pretty safe to use that in a dual-woofer design, right? Tuned to ~49Hz, two 2" ports come out to 4.69" in WinISD. If the HP crossover is normal 2nd order, I would think the only thing I should need to do is change the L-Pad for the Vifa BC25SG15-04. Am I correct in my thinking?
 
highfigh

highfigh

Seriously, I have no life.
Mark- since you already modeled the 830874 in the NASP, it's pretty safe to use that in a dual-woofer design, right? Tuned to ~49Hz, two 2" ports come out to 4.69" in WinISD. If the HP crossover is normal 2nd order, I would think the only thing I should need to do is change the L-Pad for the Vifa BC25SG15-04. Am I correct in my thinking?
I really wish we had more time to edit our posts, like the one I'm quoting. I was half-asleep when I wrote it and, well, let's just say that I would never have asked this question if I had been totally awake- it wouldn't work with my dual-woofer design because the woofer's impedance is half of what his crossover design is.


DOH!!!!:eek::eek::eek:
 
TLS Guy

TLS Guy

Seriously, I have no life.
Here is your final design.

The pdf for you final box is attached.

The woofer is Peerless 164 WR 33 102 NWP AL LS 8 ohm.

This woofer has a smooth mid band response and a good off axis response. There is a break up mode peak maximal at 4.5 kHz. It is based between 3 kHz and 5 kHz. So this requires a crossover at in the region of 2 kHz.

In the specified enclosure the F3 is 49 Hz. The tuning peaks of impedance are 33 Ohms at 28 Hz and 34 ohms 80 Hz, which gives a null of impedance of 8 ohms at the F3 of 49 Hz, showing optimal tuning of the design.

Vb is 0.56 cu.ft. The port is flared at both ends and 2.5" in diameter and 9.738" long. Vent velocity at max power is 24 m/sec.

The vent and driver displacement is 0.0618 cu.ft. This needs to be added to Vb as well as brace volume and crossover volume. The sum of these will be Vt which is the actual size of the box you need to build.

50% of the internal surface should be covered with damping material such as mineral wool. Areas behind the driver should be covered preferentially. The volume of the damping material is allowed for in the design.

The enclosure should be well braced and as narrow as possible. The tweeter should be as close to the top as possible, the woofer below the tweeter as close as possible. The port should optimally be placed below the woofer. I prefer the front, but rear mounting is also satisfactory. Drivers and port flare should be flush mounted.

To minimize diffraction effects, the corners should be rounded like this.



The tweeter is the Vifa BC25SG19-04.

This tweeter has a very smooth response over most of its pass band. There is gradual attenuation above 9 kHz which becomes steep around 18 kHz. This is not a bad fault in a tweeter for a small bookshelf speaker, and actually tends to give a better bass/HF balance. I have made no attempt to pull up the response.

Fs is 1060 Hz. So the tweeter should optimally be down 24 Hz at that frequency.

The crossover is centered at 2 kHz. The low pass filter has an electrical third order slope. The HF filter has a second order electrical slope. The combined acoustic and electrical slopes give fourth order slopes, with the woofer being 24 db down at 4 kHz and the tweeter 24 db down at 1 kHz.

There is also diffraction compensation. Because of power limitations, loss of sensitivity and lowering of impedance, 4.5 db of diffraction loss compensation is provided rather than the calculated 6 db. However room gain will almost certainly correct this. The curves shows a rising low end response, but because of diffraction loss the response will almost certainly be flat to within 3 db from 49 Hz to 9 kHz.

The crossover pdf is also attached. L2, L3 and C2 form the third order low pass filter. The values have been tweaked to provide 4.5 db of diffraction loss compensation. Req and Ce form a zobel to correct the woofer frequency dependent impedance rise so that the low pass filter can work as intended.

C1 and L1 form the second order high pass filter. Rp1 and Rp2 form an L-pad to match the tweeter output to the woofer.

All inductors should be good quality air core. The capacitors should be good quality Polypropylene. Resistors should be wire wound 10 watt minimum.

After diffraction compensation and insertion loss of the crossover, sensitivity will be in the range of 84 to 85 db 2.83 volt 1 meter. This is fairly typical for a small bookshelf speakers.

The crossover models as a superior crossover. In addition to a nice flat mid band response, the impedance curve is very flat indicating an easy to drive speaker. X-over pro assumes a sealed alignment so the tuning region impedance curve will be the one in the enclosure pdl. Impedance is 5 ohm over a good deal of the frequency spectrum. I would give this speaker a 6 ohm nominal impedance rating.

The phase response is superior. Tweeter and woofer are only 45 degrees apart at crossover. This is an excellent result.

The speaker if built correctly should have a an accurate smooth response and wide sound stage. Hopefully they will sound bigger than their size, and be pleasing to listen to even without a subwoofer.

Good luck with your build. Take your time and build it carefully.

Please keep us informed as to the progress of your build and the final result.

One final note. While getting involved with this design, I seem to remember Hi-Fi had asked for a crossover design for these speakers. My apologies for having forgotten. Better late than never!
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
Thanks, but I'm kind of confused- I'm not using the Nomex cone woofers, I'm using the Vifa BC25SG15-04 tweeter and two of the 830874 woofers. I know the NASP is crossed over at 3KHz but I was thinking of 2700- any reason to choose one over the other? The Fs of the BC25SG15-04 is 1000Hz.

The photo shows the driver arrangement.
 
Last edited:
TLS Guy

TLS Guy

Seriously, I have no life.
Thanks, but I'm kind of confused- I'm not using the Nomex cone woofers, I'm using the Vifa BC25SG15-04 tweeter and two of the 830874 woofers. I know the NASP is crossed over at 3KHz but I was thinking of 2700- any reason to choose one over the other? The Fs of the BC25SG15-04 is 1000Hz.

The photo shows the driver arrangement.
Actually NASP is crossed over at 2 kHz. I just put in any crossover frequency to get components on the board. I do all my designs with a lot of long hand figuring and experience. I don't think you can use the values the come out of any program.

So you really need NASP modified for your tweeter and the diffraction correction crossover for your lower driver of the 2.5 way.

In now have the correct idea of what you want. I was not certain before.

May be tonight I can make a start.
 
highfigh

highfigh

Seriously, I have no life.
Actually NASP is crossed over at 2 kHz. I just put in any crossover frequency to get components on the board. I do all my designs with a lot of long hand figuring and experience. I don't think you can use the values the come out of any program.

So you really need NASP modified for your tweeter and the diffraction correction crossover for your lower driver of the 2.5 way.

In now have the correct idea of what you want. I was not certain before.

May be tonight I can make a start.
You're right (I would think so- you designed it)- the corner frequency is 3KHz.

I'm running it as a 2-way now with the crossover at about 2700Hz and while the TrueRTA graph shows pretty smooth, extended response, I realized after I did the tests that I need to isolate the audio feed from my AV receiver to the garage receiver because I noticed hum when the pink noise stopped (I never hear it when I'm playing music out there) and that skewed the low frequency response. Since this was before I installed the ports, it had to be wrong- these can't go to 40Hz with almost no roll-off. I didn't have my headphones out there, so I couldn't hear when the levels were correct but my results were consistent from speaker to speaker. I tested them outside when it was very calm, the bottom of the cabinets were about 18" from the ground and the mic was at 1m. The speakers were aimed away from any objects and the nearest house was at a 45 degree angle. If I need to, I can go behind the garage, where it's all heavy growth of bushes, grass and other soft material.

Let me know if you want any measured specs- I have TrueRTA and the Dayton WT-3.

I'm only asking now because the guys who have LEAP are swamped at their business and won't be able to do anything like this for a few months. If you're familiar with Alterra Coffee, they own it and have some new locations in the works.

How do you start with crossovers? Do you use a component substitution box with jumpers or a switch, like they did when designing amplifiers? I have been thinking of doing something like that but it seems like it would cost too much to be practical unless I can have several taps on the inductors, then find the closest values so I can fine-tune it with one coil to find the correct value.
 
Last edited:
TLS Guy

TLS Guy

Seriously, I have no life.
You're right (I would think so- you designed it)- the corner frequency is 3KHz.

I'm running it as a 2-way now with the crossover at about 2700Hz and while the TrueRTA graph shows pretty smooth, extended response, I realized after I did the tests that I need to isolate the audio feed from my AV receiver to the garage receiver because I noticed hum when the pink noise stopped (I never hear it when I'm playing music out there) and that skewed the low frequency response. Since this was before I installed the ports, it had to be wrong- these can't go to 40Hz with almost no roll-off. I didn't have my headphones out there, so I couldn't hear when the levels were correct but my results were consistent from speaker to speaker. I tested them outside when it was very calm, the bottom of the cabinets were about 18" from the ground and the mic was at 1m. The speakers were aimed away from any objects and the nearest house was at a 45 degree angle. If I need to, I can go behind the garage, where it's all heavy growth of bushes, grass and other soft material.


Let me know if you want any measured specs- I have TrueRTA and the Dayton WT-3.

I'm only asking now because the guys who have LEAP are swamped at their business and won't be able to do anything like this for a few months. If you're familiar with Alterra Coffee, they own it and have some new locations in the works.

How do you start with crossovers? Do you use a component substitution box with jumpers or a switch, like they did when designing amplifiers? I have been thinking of doing something like that but it seems like it would cost too much to be practical unless I can have several taps on the inductors, then find the closest values so I can fine-tune it with one coil to find the correct value.
I have finished your 2.5 way crossover. The file is too big to upload here, and there is a problem I have to solve with my site as it is not uploading properly. So please PM me your Email, and I will send you the design right away.

Crossovers are tough. Looking at the acoustic data is key. The trick is to doodle until you get a smooth crossover that takes account of the acoustic roll offs. As you start to play with it, it becomes clear where the crossover had to be.

I like to avoid notch filters if I can be designing the slopes correctly. Sometimes you have to notch a peak.

Then you need to see if the woofers impedance needs equalizing, it usually does, but sometimes you can use it to your advantage.

There are other tricks like bypassing resistors with cap to pull up responses.

It is a complex business and it will drive you nuts when you first start.

The best thing is to get a program and play with it.

If you build this crossover I would be very interested in your measurements. We can fine tune things based on that and listening.
 
TLS Guy

TLS Guy

Seriously, I have no life.
Here is an abbreviated version.
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
I have finished your 2.5 way crossover. The file is too big to upload here, and there is a problem I have to solve with my site as it is not uploading properly. So please PM me your Email, and I will send you the design right away.

Crossovers are tough. Looking at the acoustic data is key. The trick is to doodle until you get a smooth crossover that takes account of the acoustic roll offs. As you start to play with it, it becomes clear where the crossover had to be.

I like to avoid notch filters if I can be designing the slopes correctly. Sometimes you have to notch a peak.

Then you need to see if the woofers impedance needs equalizing, it usually does, but sometimes you can use it to your advantage.

There are other tricks like bypassing resistors with cap to pull up responses.

It is a complex business and it will drive you nuts when you first start.

The best thing is to get a program and play with it.

If you build this crossover I would be very interested in your measurements. We can fine tune things based on that and listening.
The funny thing about this combination is that, even though I KNOW the crossover is wrong, the response from about 200Hz-12KHz is devoid of any major peaks or dips. Pre-installation of the ports, the impedance curve has two strong peaks- the low one I know is the woofer's natural peak and the upper is the combination of the woofer's natural rise being slightly separated from the low end of the tweeter's impedance curve and the crossover transition being in the wrong place. That said, the phase plot is +/- 45 degrees, if that. I'm on my desktop now and the data is in my laptop. With my Denon receiver, I went in and manually EQ'd the front channels and below 1KHz, I leave it alone completely, while I roll the high end off gradually and smoothly at 2K (-2.5dB), 4K (-3dB) and 8K (-2.5dB) being the range where the tweeter's sensitivity makes it too loud for any kind of balanced sound. 16K is -2dB and it's easy to listen to them. HOWEVER, I know it's not right and want them to be correct. The ports can't be far off because the bass is even and smooth. While it's not going to be good for rap (OK by me), it sounds very nice with acoustic bass, has good extension (I don't really need the frequency response to go from DC-channel 5) and piano sounds very natural. Since these will be in a living room with a piano, that's important. He and his daughter are taking violin lessons, too.
 
A

ameuba10

Enthusiast
progress thus far...








the black box is what they were in...big difference eh? i know the holes are far from perfect but i think itll suffice for my first project. until my crossover components come in, this is where i stand. im going to wood puddy the seams and screws for a nice smooth finish and give it a coat of midnight blue. i have a hard wood that i will stain for the front baffle. still contiplating on the design. what do you guys think?
 
annunaki

annunaki

Moderator
I think that it is great that you started a project and are getting into the DIY area! :)

If you would have went with a slot vent you could have avoided any rough cuts around the port openings as you can make easy straight cuts.

Here is an example of what I mean:
http://forums.audioholics.com/forums/showthread.php?t=61012

If I were doing this over again I would have done things a little different but it should give you some idea of what I mean.
 
highfigh

highfigh

Seriously, I have no life.
progress thus far.the black box is what they were in...big difference eh? i know the holes are far from perfect but i think itll suffice for my first project. until my crossover components come in, this is where i stand. im going to wood puddy the seams and screws for a nice smooth finish and give it a coat of midnight blue. i have a hard wood that i will stain for the front baffle. still contiplating on the design. what do you guys think?
Don't be stingy with the screws- The corners each need one and I usually put one every 4". If you tap on the side where you're putting them, you'll hear the note go up as you add screws and this is putting the cabinet resonance above anything that will be excited by the woofer's frequencies.
 
R

ramiro77

Audiophyte
The other issue is he won't gain much over a sealed alignment. His driver is really intended for sealed and the 830874 for vented.

The sealed enclosure has an F3 of 85, but the rolls off 12 db per octave, so actually the sealed has greater low bass output. At 40 Hz the output of vented or sealed is pretty much a wash.
Interesting. With this you are stating that the peerless 830875 works best in a sealed box?
I would like to know more about it. All the enclosures I saw for this woofer were vented. I have a pair of these woofers, and I would like to apply them in a sealed box if they really deliver good bass.

P/D:
Sorry if my English is not good enough.
Merry Christmas to all!
 
TLS Guy

TLS Guy

Seriously, I have no life.
Interesting. With this you are stating that the peerless 830875 works best in a sealed box?
I would like to know more about it. All the enclosures I saw for this woofer were vented. I have a pair of these woofers, and I would like to apply them in a sealed box if they really deliver good bass.

P/D:
Sorry if my English is not good enough.
Merry Christmas to all!
Yes, output at 40 Hz with that driver sealed or vented is the same, but the sealed will have greater output at 20 kHz.
 

Latest posts

newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top