@Wolf said:
Use posts, and an inner-mounted board across a recess hole.
Yes, I have seen this type of mounting on many of Javads speakers. The binding posts are countersunk somewhat, but not so much as to prevent bare wire or spade lug hook-ups. The posts are not completely recessed, however. If you mount the posts too far into the cabinet, then spade lugs or bare wires can become difficult. What I would like to see is a heavy duty wooden terminal cup with the binding posts mounted at a 45 degree angle. Just like the plastic SD-cup or TD-cup styles, only much thicker and more durable.
Along these lines, I just finished putting together a 1.5" thick wooden terminal cup made out of particle board. The binding posts are mounting at a 45 degree angle so that you can easily use banana jacks, spade lugs, or bare wire. Very solid and cheap. I'll start a new thread and post a drawing and some pictures.
@kenrhodes said:
I have used the SD-cup and the TD-cup from meniscus and liked them both. Dayron also makes some nice binding post.
Javad seems to have started a craze with the intigrated wooden terminal cups that is a nice looking option that keeps the back flush.
I like the SD and TD cups too. My only problem is the thin plastic used. Dayton sells the same ones. I also like Javads wooden terminal cups, but I think his terminals stick out the back just a little bit.
@Wolf said:
Ani, Ken just means like I suggested. It's been around a lot longer than just Javad doing it. Even Joey's RS 8" 2-ways resemble the type.
My pair of classic Large Advent Loudspeakers (circa 1974) also uses completely recessed binding posts mounted on an internal piece of 1/4" thick masonite that is glued inside the back panel.
Looking good! I don't know why but I like this look with the exposed edges - would be kind of fun to build a set of speakers like that and just slap some finish on them
Thanks. I was thinking the same thing. Maybe just sand the raw edges a little with 220 grit, and then give them 7 to 8 coats of high gloss lacquer.
I just started a new thread with drawing & pics for my interpretation of the wooden terminal cup idea. I forgot to take a pic showing clearances with banana plugs and spade lugs installed. Will stop back later with a few more pics.
Last night I started building the port-trap/stands. I had a few problems, so I revised my stand drawing (see attached) to include additional information and a better PVC 90 degree elbow.
Here are the two 24" long sections. The top trap mounting board section is 15.25" long with a 2.75" diameter hole for the trap.
Here is a closeup of the PVC trap. I switched over to 2" ID pipe (2.5" didn't fit well). The traps have been cut to about 10 inches total length which is just a starting point. They will be trimmed shorter when I get everything put together for NF measurements:
Here is the lower tuning section, which will allow sliding type length adjustments.
The following 3 pics show how this tuning will be accomplished. The lower section slides up and down the port, as needed, while the length of the upper section is also adjusted to keep the port exit dimensions constant.
Yesterday I made a template for machining the two top plates for my port/stand. The template is the darker colored panel on the right, which is a composite glue-up from 4 separate boards. This template was necessary to create a precise alignment between the upper and lower sections. Each plate has four 1/8" inner mounting holes, four 1/4" outer mounting holes, and a 2.5 x 2.5" hole to match the port inside dimensions.
I started by precisely aligning one of the top plates to the bottom of the box. I marked the hole with a sharpie.
I then rough cut the hole with my jigsaw, being very careful to stay "inside" of the line.
I then aligned and re-clamped one of the top plates over this hole and drilled the four 1/4-20 mounting holes.
Then I inserted four 1/4-20 machine bolts into these holes from the underside and let them stick up proud by about 1/4"
Then I placed two pieces of double stick carpet tape between the bolts:
Then I placed the top plate onto the bottom, using the 1/4-20 bolts sticking up as alignment dowels. I temporarily clamped this sandwich together under pressure to cause the carpet tape to firmly adhere. Then I flush trimmed the hole to precise dimensions.
The router bit left rounded corners, so I squared them up with a 32 TPI hacksaw blade and a flat file.
After double checking the top plate alignment, everything looks good so far:
This morning I rounded the internal port opening with my smaller router with a 3/8" radius bit to ease the transition. Then I enlarged the four mounting holes from 1/4" to 5/16" and installed some 1/4-20 T-nuts. I used the 7/16" long T-nuts, which seemed to give a very good bit into the wood when I pounded them in. T-nuts can be a problem, coming loose, but if this happens the removable back will allow me to get back inside the box to fix the problem.
Alignment to the top plates was good, but not perfect. Even with using my master template, I could only hold a tolerance of about 1/32nd of an inch. So I enlarged the 1/4" top plate holes to 9/32" and now everything bolts and unbolts smoothly without binding.
This morning I drilled & screwed the top plates onto the stand columns. I sealed the junction with a small bead of 100% silicone sealer. This way, I can remove the top plates later if necessary.
Before screwing the top plates in place, I gave the edge a 3/8" radius roundover. This should make it look a little better from a seated position:
Then I made a 1/16" thick foam rubber gasket to seal the junction between the stand and the cabinet. I glued it on with a water based neoprene rubber formula contact cement, which adheres well, but stays soft and does not damage the foam gasket material.
This is the combo bass reflex port/speaker stand. It will be fastened to the bottom of the speakers. When I bolt it all together it will look like a small bookshelf speaker sitting on a normal looking speaker stand. I'm probably taking too many detailed, closeup photos. I'll move back and take a wider view of the entire project,.
@Kornbread said:
Will the stands be larger and cover the trap?
The trap is a smaller 2" ID PVC pipe that sticks out the back of the stand. Here are a couple wider view pictures showing all parts of the project so far. I boosted the resolution on the pics a little so you can drill down and see all the parts a little bit better. The white 2" PVC plumbing pipe is the trap, which will be stuffed with denim and trimmed to length based on NF port measurements. Currently, the trap is 10" long, but this will be reduced to about 7 or 8" or so later.
The stands bolt to the main cabs via four 1/4-20 threaded plastic wingnut type knobs. A gasket makes this junction air tight. I placed the back panels up against the rear wall, just for the photo. They will be installed soon. The two small boards on the basement floor are part of the adjustable port exit and will be cut to fit later. They will slide up and down to adjust the port length when I start doing measurements.
Another day, another screw up. I messed up and drilled the holes for the stand base in the wrong location. The fix, however, will be fairly simple. I made two small triangular braces to cover up the hole!
Up next: Fiberglass damping, installing woofers, installing backs, installing tweeter, and running NF port and woofer cone tests. I will also be running DATS V2 impedance sweeps and checking for air leaks as I go along.
I lined the cab with 1/2" fiberglass. The small red caps are leftover bar clamp pads that I am using to seal the T-nuts and prevent air leaks.
The back panel bolts in place with 12 screws and a gasket:
In the next picture, a single clamp holds a partition in place as I take a NF port measurement. This measurement was for the shortest port length of 17" and a trap length of 4.5" with light Denim stuffing. The internal box trap was the shortest dimension of about 5.5" and was stuffed with 2.9 ounces of denim. The goal will be to adjust both trap lengths and stuffing densities so as to minimize the peaking in the 250Hz to 1KHz region. I will post my progress as I continue testing and adjusting. I made the mistake of bolting the back on with all 12 screws. I will remove all screws and simply clamp the backs temporarily in place. This way I can quickly pop the back off between measurements.
Then I took a NF cone measurement for the same 17" port partition, 4.5" external trap, and internal 5.5" trap lengths. The tweeter in the picture is not connected. I just screwed it in place to seal the hole:
Then I took an impedance sweep for the same partition and trap lengths:
I have been running NF port measurements while OmniMic's continuous sine wave sweep plays. I like to watch the screen as I adjust internal trap lengths in stages (5.5-8.5"). The internal trap seems to have a slight effect on the peaking in the 500Hz to 1kHz band. I found the best trap length to be 8.5 inches, so I stopped there. Also, I found that adding a 3" thick layer of denim to the entire top panel pulled down the 500Hz to 1kHz peaking somewhat. Since this only caused the main port resonance to fall by about 1dB, I decided to stop there.
On to the external port trap. Again, as OmniMic's continuous sine wave played, I swapped out a wide variety of trap lengths ranging from 10 inches down to about 4 inches. I found the sweet spot to be about 7 inches with a light amount of denim stuffing. This length calculates and lines up with a 330Hz trapping resonance, as you can see in the graph. The port length is 20.5 inches, which I initially thought was 17 inches, but I forgot to add the turns at the port ends. 13560 / 20.5 = 661 / 2 = 330Hz. Trap: 13560 / 7 = 1937 / 4 = 484 * .68 fudge factor = 330Hz. For some reason, the trap length always ends up being shorter than the calculation. I have no idea as to why this is the case. For the Plumbers Delight speakers, my fudge factor was a little higher.
Here is a graph comparison. Not much of an improvement, but I can hear a big difference in the sine sweep. Will keep working on it to see if I can get the peaks lower without killing off the main resonance too much.
Going down the road less traveled, I decided to develop an 8th order passive crossover @ 910Hz. I plan to order parts next week and give it a try. I think the positives outweigh the negatives in this particular case.
Negatives:
1) 910Hz is very close to the tweeter's Fs of approximately 720 and 740Hz. PE recommends a minimum of 1400Hz.
2) Potentially high HD measurements around the crossover region (TBD).
3) Strained or gritty type sound quality from pushing the tweeter too hard (TBD).
4) An 8th order network might have audible phase or timing issues (group delay?).
5) Higher parts count (11) and cost.
Positives:
1) PE recommends a minimum xover of 1400Hz, but this assumes a 4th order high-pass. My golden pair of RST28F-4's use ferrofluid and have impedance peaks of 6.3 & 6.5 ohms at 720 and 740Hz. My modeled 8th order high-pass slope is 25dB down at 700Hz and does not appear to be influenced adversely by the tweeter's resonance.
2) Smoother polars, both vertical and horizontal
3) Allows increased tweeter to woofer cone spacing for the same polar response.
4) Better sub-woofer breakup suppression (-28dB @ 1.2kHz, and -45dB @ 2.2kHz)
5) Acoustical 8th order tweeter high-pass @ 910Hz puts about the same amount of strain on the tweeter as a 4th order high-pass at 1400Hz.
6) Phase tracks well throughout the tight crossover region and has a deep reverse null (8th order).
7) The 8th order filter solves a frequency response glitch created by my isobarik woofer loading. The 4 inch spacing distance between the inner and outer woofers creates a much larger response dip from 1-2kHz compared to the stock woofer (see below FRD graph). The 8th order filter rolls this off before it becomes a problem.
XSim 3 pane on-axis model (seated listening position):
XSim 3 pane 45H model:
XSim 3 pane +15V model (standing listening position). There should be no difference in tonal balance going from a seated to a standing position:
Here are the raw, as measured, FRDs used to develop the models. Compare the 1-2kHz glitch on my isobarik woofer measurement to Wolf's measurements of a single woofer in his Zinger thread and you can see the difference caused by my 4 inch deep isobarik tunnel.
@Wolf said:
Wow, close to -9dB there, and I had -4dB on one and -3.5dB on the other.
Most speaker builders think that the isobarik configuration simply 1) cuts VAS in half and, 2) possibly reduces HD a little bit for the clam-shell arrangement. But this is not true. The two woofers begin to fight each other through the air gap, based on the spacing distances involved. You have to account for this in your xover.
@ani_101 said:
What does that show up as? Not sure I follow when you say the two woofers fight each other.
Ani, Sorry for the long delay in responding. I usually log on at my local library, but they shut the library down again due to Covid.
This shows up as frequency response ripple, somewhere in the 500Hz to 2kHz region, based on the spacing distance between the two drivers.
For instance, in my speaker the spacing distance between the two woofers is 4 inches. Four inches is one half wavelength at 1695Hz (speed of sound 13560/4 inches = 3390 / 2 = 1695 half wavelength). Both speakers received the input signal at the speed of light. But it takes a small amount of time for the sound to travel from the inner driver to the outer driver. Therefore, at one half wavelength, the two drivers are 180 degrees out of phase, which causes a dip in the frequency response. If you analyze this effect over a band of frequencies, what you will see is an up and down ripple in frequency response with the largest dip at 1695Hz.
The isobarik principle (constant pressure) is only applicable at lower frequencies. At higher frequencies, the two woofers fight each other by boosting and decreasing the frequency response through the air gap.
Bill, do you have an measurement of the ripple or can you post a zoomed in image or indicate the ripple. I suppose if you disconnect the second (inside) driver the 1600 hz ripple go away?
I, maybe erroneously, look at the two iso driver’s as how they affect the defined volume through pressure, which absent differences in the driver's outputs is linear push-pull. And within the low frequency domain only (where I’ve played around), so haven’t seen issues with low-energy high-frequency affecting a woofer.
I did play with an open back iso set-up to test the affects of varying the energy of the drivers (simply placing an attenuator on the back driver) > looking at how doing so could be employed to deal with baffle cancellation issues of open back / open baffle woofers (ok, too much time on my hands). In any event, here is a pic. (the gold being non-attenuated iso)
@Kornbread said:
Could delay the further woofer using dsp?
The front woofer and the tweeter would need to be delayed?
Wondering what would happen if you delayed the inside woofer just a smidge and/or use the dsp to smooth out the overall woofer response by adjusting only the inside woofer.
Comments
Yes, I have seen this type of mounting on many of Javads speakers. The binding posts are countersunk somewhat, but not so much as to prevent bare wire or spade lug hook-ups. The posts are not completely recessed, however. If you mount the posts too far into the cabinet, then spade lugs or bare wires can become difficult. What I would like to see is a heavy duty wooden terminal cup with the binding posts mounted at a 45 degree angle. Just like the plastic SD-cup or TD-cup styles, only much thicker and more durable.
Along these lines, I just finished putting together a 1.5" thick wooden terminal cup made out of particle board. The binding posts are mounting at a 45 degree angle so that you can easily use banana jacks, spade lugs, or bare wire. Very solid and cheap. I'll start a new thread and post a drawing and some pictures.
I like the SD and TD cups too. My only problem is the thin plastic used. Dayton sells the same ones. I also like Javads wooden terminal cups, but I think his terminals stick out the back just a little bit.
Some examples would be Javad's EPQ-3W speakers, the Rivalries, the Stance speakers, etc.
My pair of classic Large Advent Loudspeakers (circa 1974) also uses completely recessed binding posts mounted on an internal piece of 1/4" thick masonite that is glued inside the back panel.
Thanks. I was thinking the same thing. Maybe just sand the raw edges a little with 220 grit, and then give them 7 to 8 coats of high gloss lacquer.
I just started a new thread with drawing & pics for my interpretation of the wooden terminal cup idea. I forgot to take a pic showing clearances with banana plugs and spade lugs installed. Will stop back later with a few more pics.
https://diy.midwestaudio.club/discussion/1527/heavy-duty-wooden-terminal-cup-with-flush-45-degree-angle-binding-posts#latest
Last night I started building the port-trap/stands. I had a few problems, so I revised my stand drawing (see attached) to include additional information and a better PVC 90 degree elbow.
Here are the two 24" long sections. The top trap mounting board section is 15.25" long with a 2.75" diameter hole for the trap.
Here is a closeup of the PVC trap. I switched over to 2" ID pipe (2.5" didn't fit well). The traps have been cut to about 10 inches total length which is just a starting point. They will be trimmed shorter when I get everything put together for NF measurements:
Here is the lower tuning section, which will allow sliding type length adjustments.
The following 3 pics show how this tuning will be accomplished. The lower section slides up and down the port, as needed, while the length of the upper section is also adjusted to keep the port exit dimensions constant.
Yesterday I made a template for machining the two top plates for my port/stand. The template is the darker colored panel on the right, which is a composite glue-up from 4 separate boards. This template was necessary to create a precise alignment between the upper and lower sections. Each plate has four 1/8" inner mounting holes, four 1/4" outer mounting holes, and a 2.5 x 2.5" hole to match the port inside dimensions.
I started by precisely aligning one of the top plates to the bottom of the box. I marked the hole with a sharpie.
I then rough cut the hole with my jigsaw, being very careful to stay "inside" of the line.
I then aligned and re-clamped one of the top plates over this hole and drilled the four 1/4-20 mounting holes.
Then I inserted four 1/4-20 machine bolts into these holes from the underside and let them stick up proud by about 1/4"
Then I placed two pieces of double stick carpet tape between the bolts:
Then I placed the top plate onto the bottom, using the 1/4-20 bolts sticking up as alignment dowels. I temporarily clamped this sandwich together under pressure to cause the carpet tape to firmly adhere. Then I flush trimmed the hole to precise dimensions.
The router bit left rounded corners, so I squared them up with a 32 TPI hacksaw blade and a flat file.
After double checking the top plate alignment, everything looks good so far:
This morning I rounded the internal port opening with my smaller router with a 3/8" radius bit to ease the transition. Then I enlarged the four mounting holes from 1/4" to 5/16" and installed some 1/4-20 T-nuts. I used the 7/16" long T-nuts, which seemed to give a very good bit into the wood when I pounded them in. T-nuts can be a problem, coming loose, but if this happens the removable back will allow me to get back inside the box to fix the problem.
Alignment to the top plates was good, but not perfect. Even with using my master template, I could only hold a tolerance of about 1/32nd of an inch. So I enlarged the 1/4" top plate holes to 9/32" and now everything bolts and unbolts smoothly without binding.
This morning I drilled & screwed the top plates onto the stand columns. I sealed the junction with a small bead of 100% silicone sealer. This way, I can remove the top plates later if necessary.
Before screwing the top plates in place, I gave the edge a 3/8" radius roundover. This should make it look a little better from a seated position:
Then I made a 1/16" thick foam rubber gasket to seal the junction between the stand and the cabinet. I glued it on with a water based neoprene rubber formula contact cement, which adheres well, but stays soft and does not damage the foam gasket material.
I'm trying to follow along ... I kinda lost.
This is the combo bass reflex port/speaker stand. It will be fastened to the bottom of the speakers. When I bolt it all together it will look like a small bookshelf speaker sitting on a normal looking speaker stand. I'm probably taking too many detailed, closeup photos. I'll move back and take a wider view of the entire project,.
Will the stands be larger and cover the trap?
The trap is a smaller 2" ID PVC pipe that sticks out the back of the stand. Here are a couple wider view pictures showing all parts of the project so far. I boosted the resolution on the pics a little so you can drill down and see all the parts a little bit better. The white 2" PVC plumbing pipe is the trap, which will be stuffed with denim and trimmed to length based on NF port measurements. Currently, the trap is 10" long, but this will be reduced to about 7 or 8" or so later.
The stands bolt to the main cabs via four 1/4-20 threaded plastic wingnut type knobs. A gasket makes this junction air tight. I placed the back panels up against the rear wall, just for the photo. They will be installed soon. The two small boards on the basement floor are part of the adjustable port exit and will be cut to fit later. They will slide up and down to adjust the port length when I start doing measurements.
Another day, another screw up. I messed up and drilled the holes for the stand base in the wrong location. The fix, however, will be fairly simple. I made two small triangular braces to cover up the hole!
Up next: Fiberglass damping, installing woofers, installing backs, installing tweeter, and running NF port and woofer cone tests. I will also be running DATS V2 impedance sweeps and checking for air leaks as I go along.
Seeing them assembled gives you a much better overall picture of how they will work - looks good👍👍
A little progress this past week.
I lined the cab with 1/2" fiberglass. The small red caps are leftover bar clamp pads that I am using to seal the T-nuts and prevent air leaks.
The back panel bolts in place with 12 screws and a gasket:
In the next picture, a single clamp holds a partition in place as I take a NF port measurement. This measurement was for the shortest port length of 17" and a trap length of 4.5" with light Denim stuffing. The internal box trap was the shortest dimension of about 5.5" and was stuffed with 2.9 ounces of denim. The goal will be to adjust both trap lengths and stuffing densities so as to minimize the peaking in the 250Hz to 1KHz region. I will post my progress as I continue testing and adjusting. I made the mistake of bolting the back on with all 12 screws.
I will remove all screws and simply clamp the backs temporarily in place. This way I can quickly pop the back off between measurements.
Then I took a NF cone measurement for the same 17" port partition, 4.5" external trap, and internal 5.5" trap lengths. The tweeter in the picture is not connected. I just screwed it in place to seal the hole:
Then I took an impedance sweep for the same partition and trap lengths:
Thanks, squamishdroc.
I have been running NF port measurements while OmniMic's continuous sine wave sweep plays. I like to watch the screen as I adjust internal trap lengths in stages (5.5-8.5"). The internal trap seems to have a slight effect on the peaking in the 500Hz to 1kHz band. I found the best trap length to be 8.5 inches, so I stopped there. Also, I found that adding a 3" thick layer of denim to the entire top panel pulled down the 500Hz to 1kHz peaking somewhat. Since this only caused the main port resonance to fall by about 1dB, I decided to stop there.
On to the external port trap. Again, as OmniMic's continuous sine wave played, I swapped out a wide variety of trap lengths ranging from 10 inches down to about 4 inches. I found the sweet spot to be about 7 inches with a light amount of denim stuffing. This length calculates and lines up with a 330Hz trapping resonance, as you can see in the graph. The port length is 20.5 inches, which I initially thought was 17 inches, but I forgot to add the turns at the port ends. 13560 / 20.5 = 661 / 2 = 330Hz. Trap: 13560 / 7 = 1937 / 4 = 484 * .68 fudge factor = 330Hz. For some reason, the trap length always ends up being shorter than the calculation. I have no idea as to why this is the case. For the Plumbers Delight speakers, my fudge factor was a little higher.
Here is a graph comparison. Not much of an improvement, but I can hear a big difference in the sine sweep. Will keep working on it to see if I can get the peaks lower without killing off the main resonance too much.
Photo of the 8.5" internal trap:
Going down the road less traveled, I decided to develop an 8th order passive crossover @ 910Hz. I plan to order parts next week and give it a try. I think the positives outweigh the negatives in this particular case.
Negatives:
1) 910Hz is very close to the tweeter's Fs of approximately 720 and 740Hz. PE recommends a minimum of 1400Hz.
2) Potentially high HD measurements around the crossover region (TBD).
3) Strained or gritty type sound quality from pushing the tweeter too hard (TBD).
4) An 8th order network might have audible phase or timing issues (group delay?).
5) Higher parts count (11) and cost.
Positives:
1) PE recommends a minimum xover of 1400Hz, but this assumes a 4th order high-pass. My golden pair of RST28F-4's use ferrofluid and have impedance peaks of 6.3 & 6.5 ohms at 720 and 740Hz. My modeled 8th order high-pass slope is 25dB down at 700Hz and does not appear to be influenced adversely by the tweeter's resonance.
2) Smoother polars, both vertical and horizontal
3) Allows increased tweeter to woofer cone spacing for the same polar response.
4) Better sub-woofer breakup suppression (-28dB @ 1.2kHz, and -45dB @ 2.2kHz)
5) Acoustical 8th order tweeter high-pass @ 910Hz puts about the same amount of strain on the tweeter as a 4th order high-pass at 1400Hz.
6) Phase tracks well throughout the tight crossover region and has a deep reverse null (8th order).
7) The 8th order filter solves a frequency response glitch created by my isobarik woofer loading. The 4 inch spacing distance between the inner and outer woofers creates a much larger response dip from 1-2kHz compared to the stock woofer (see below FRD graph). The 8th order filter rolls this off before it becomes a problem.
XSim 3 pane on-axis model (seated listening position):
XSim 3 pane 45H model:
XSim 3 pane +15V model (standing listening position). There should be no difference in tonal balance going from a seated to a standing position:
Here are the raw, as measured, FRDs used to develop the models. Compare the 1-2kHz glitch on my isobarik woofer measurement to Wolf's measurements of a single woofer in his Zinger thread and you can see the difference caused by my 4 inch deep isobarik tunnel.
Wow, close to -9dB there, and I had -4dB on one and -3.5dB on the other.
InDIYana Event Website
Wolfy's Photobucket
Most speaker builders think that the isobarik configuration simply 1) cuts VAS in half and, 2) possibly reduces HD a little bit for the clam-shell arrangement. But this is not true. The two woofers begin to fight each other through the air gap, based on the spacing distances involved. You have to account for this in your xover.
What does that show up as? Not sure I follow when you say the two woofers fight each other.
Ani, Sorry for the long delay in responding. I usually log on at my local library, but they shut the library down again due to Covid.
This shows up as frequency response ripple, somewhere in the 500Hz to 2kHz region, based on the spacing distance between the two drivers.
For instance, in my speaker the spacing distance between the two woofers is 4 inches. Four inches is one half wavelength at 1695Hz (speed of sound 13560/4 inches = 3390 / 2 = 1695 half wavelength). Both speakers received the input signal at the speed of light. But it takes a small amount of time for the sound to travel from the inner driver to the outer driver. Therefore, at one half wavelength, the two drivers are 180 degrees out of phase, which causes a dip in the frequency response. If you analyze this effect over a band of frequencies, what you will see is an up and down ripple in frequency response with the largest dip at 1695Hz.
The isobarik principle (constant pressure) is only applicable at lower frequencies. At higher frequencies, the two woofers fight each other by boosting and decreasing the frequency response through the air gap.
Bill
Could delay the further woofer using dsp?
The front woofer and the tweeter would need to be delayed?
Would a 2.5 way work for isibarik? The front woofer playing upto the hf, the back woofer only the lows? Would that be weird?
But then mark (from meniscus) had one woofer ported and one sealed playing the same freq...
Bill, do you have an measurement of the ripple or can you post a zoomed in image or indicate the ripple. I suppose if you disconnect the second (inside) driver the 1600 hz ripple go away?
I, maybe erroneously, look at the two iso driver’s as how they affect the defined volume through pressure, which absent differences in the driver's outputs is linear push-pull. And within the low frequency domain only (where I’ve played around), so haven’t seen issues with low-energy high-frequency affecting a woofer.
I did play with an open back iso set-up to test the affects of varying the energy of the drivers (simply placing an attenuator on the back driver) > looking at how doing so could be employed to deal with baffle cancellation issues of open back / open baffle woofers (ok, too much time on my hands). In any event, here is a pic. (the gold being non-attenuated iso)
Wondering what would happen if you delayed the inside woofer just a smidge and/or use the dsp to smooth out the overall woofer response by adjusting only the inside woofer.