Recently i made some speakers boxes using 2x goldwood 15"s in a 11ft3 ported box, very similar to JBL4350/4351.
Thoses goldwoods have specs that are very close to the 2235h. yeah i know they are not 2235 at all.
one need to start somewhere and this is where i am.
i followed specs of the 4350/4351 in terms of volume/construction of the box itself.
i followed the winisd suggestion which where very close too of the 4351. i didnt cheaped the build of the boxes. solid russian veneer for the sides, 1" mdf for the front and rear, stifners inside out,...was a bit disapointed of the overall thump bass....feels anemic over 40hz.
later i bought DATS V3 and found out that my tuning FS was like 23hz instead of 30 !!
i read everywhere that it is ok to tune below FS, but i find that my speakers feels anemic in the 70-100hz region. i really miss that thump bass that my older projetc with 2225h gaved me..... theses goldwoods can go extremely low thats not problem, maybe too low.
i was wondering how i could bring that up. i tought about adding mass volume inside the box (like bag of sands or logs or my cat,... )
i shaved a bit the port from inside but that was very difficult and time wasting as i went from 23 to 26hz like. i stoped there.
VOLUME IS NET 11FT3 for 2x 15" inside the same volume, 6x ports x 3inch diameter, about 8 inches long. when theses parameters are fullfilled and entered into winisd it gives me 30hz of tuning resonance but DATS measure much lower (wtf)
light stuffing inside, not that much.
thanks for any insights
Thoses goldwoods have specs that are very close to the 2235h. yeah i know they are not 2235 at all.
one need to start somewhere and this is where i am.
i followed specs of the 4350/4351 in terms of volume/construction of the box itself.
i followed the winisd suggestion which where very close too of the 4351. i didnt cheaped the build of the boxes. solid russian veneer for the sides, 1" mdf for the front and rear, stifners inside out,...was a bit disapointed of the overall thump bass....feels anemic over 40hz.
later i bought DATS V3 and found out that my tuning FS was like 23hz instead of 30 !!
i read everywhere that it is ok to tune below FS, but i find that my speakers feels anemic in the 70-100hz region. i really miss that thump bass that my older projetc with 2225h gaved me..... theses goldwoods can go extremely low thats not problem, maybe too low.
i was wondering how i could bring that up. i tought about adding mass volume inside the box (like bag of sands or logs or my cat,... )
i shaved a bit the port from inside but that was very difficult and time wasting as i went from 23 to 26hz like. i stoped there.
VOLUME IS NET 11FT3 for 2x 15" inside the same volume, 6x ports x 3inch diameter, about 8 inches long. when theses parameters are fullfilled and entered into winisd it gives me 30hz of tuning resonance but DATS measure much lower (wtf)
light stuffing inside, not that much.
thanks for any insights
Attachments
WinIsd has leakage or QL set to 10
Ql of 7 more so a better standard
A 11 cubic foot box has leakage closer to 4 or 5
which can dramatically change the impedance curve.
The math in WinIsd is correct for port calculation.
But assumes normal error factor.
You can only change to 3 set numbers. Highest being
around .850 I believe or what it calls " dual flare port"
but the real world and error factors take over.
rather common with multiport and large enclosure
to get much lower tuning.
Winisd wont show new tuning with leakage change in project
window.
But the impedance curve will show a change.
Leakage would be closer to 4 or 5 not default 10
Error factor for multiple ports , mounted in large box
close to walls, or in lower cabinet area.
Make the port appear much much longer.
Winisd allows up to .8 ish, when it is closer to .9 or 1
for error correction.
When building large boxes more than 4 cubic feet.
and multiple ports
I make leakage very high in winisd around 4 or 5 and use dual flare
for highest end factor correction possible. Even if not using " dual flare"
Then look at impedance
curve for Fb not project window.
More realistic because the model math is " correct" just doesnt add to real world
numbers. Even Theile Small knew these and added multiple " correction factors"
to the formula to make it actually match the real world , not theory.
WinIsd having some of the basic ones Qa Ql Qp in advanced features.
The real problem is port error factor not adjustable and highest number is the
" dual flare"
Advanced features show effects on impedance curve chart, but project window
tuning will stay the same.
All in all for higher tuning need to shorten ports.
Depending on what goldwoods these are might not make much a difference.
5 Hz change not noticeable, Takes 10 to 12 Hz jumps for tuning to be notable.
To high is annoying and smeary anyways.
Ql of 7 more so a better standard
A 11 cubic foot box has leakage closer to 4 or 5
which can dramatically change the impedance curve.
The math in WinIsd is correct for port calculation.
But assumes normal error factor.
You can only change to 3 set numbers. Highest being
around .850 I believe or what it calls " dual flare port"
but the real world and error factors take over.
rather common with multiport and large enclosure
to get much lower tuning.
Winisd wont show new tuning with leakage change in project
window.
But the impedance curve will show a change.
Leakage would be closer to 4 or 5 not default 10
Error factor for multiple ports , mounted in large box
close to walls, or in lower cabinet area.
Make the port appear much much longer.
Winisd allows up to .8 ish, when it is closer to .9 or 1
for error correction.
When building large boxes more than 4 cubic feet.
and multiple ports
I make leakage very high in winisd around 4 or 5 and use dual flare
for highest end factor correction possible. Even if not using " dual flare"
Then look at impedance
curve for Fb not project window.
More realistic because the model math is " correct" just doesnt add to real world
numbers. Even Theile Small knew these and added multiple " correction factors"
to the formula to make it actually match the real world , not theory.
WinIsd having some of the basic ones Qa Ql Qp in advanced features.
The real problem is port error factor not adjustable and highest number is the
" dual flare"
Advanced features show effects on impedance curve chart, but project window
tuning will stay the same.
All in all for higher tuning need to shorten ports.
Depending on what goldwoods these are might not make much a difference.
5 Hz change not noticeable, Takes 10 to 12 Hz jumps for tuning to be notable.
To high is annoying and smeary anyways.
There was a discussion recently in the sub forum where David McBean pointed out the multiple ports, close together, will behave more like one large one. That means they get more effective end correction, and thus tune lower then WinISD will predict. That's probably what has caused your tuning mismatch.
However awkward it is, shortening the vents will be the best way of raising the tuning frequency back up.
Or, would it be easier to add in a 4th vent to each set of 3? That would also allow you to raise Fb without shortening the existing vents.
However awkward it is, shortening the vents will be the best way of raising the tuning frequency back up.
Or, would it be easier to add in a 4th vent to each set of 3? That would also allow you to raise Fb without shortening the existing vents.
Roundover on the ports will both raise tuning frequency and reduce port distortion, so it's a win-win solution. Would look less 70s though.
i want to be very clear and specific that i used published JBL specs and plans for 4350 and kenrik sound 4351 plans
-same amount of port (6)
-same diameter (3 inch)
-same initial lenght ( i did cut them down to 8 inch to no avail it raised the tuning to 28so hz)
-no flared ends
almost same volume of 10ft3 (i am at+1ft3)
box is AIR TIGHT i did test it with several method and stupids ways. it doesnt leak anythere and is very very solid and firm.
lets forget what winisnd said.
my question is really specific about adding internal mass/volume (ie reducing internal volume ft3) to raise the tuning frequency. i did cut the port already and it is way out of a potential solution as it is too much diffuclt to do.
-same amount of port (6)
-same diameter (3 inch)
-same initial lenght ( i did cut them down to 8 inch to no avail it raised the tuning to 28so hz)
-no flared ends
almost same volume of 10ft3 (i am at+1ft3)
box is AIR TIGHT i did test it with several method and stupids ways. it doesnt leak anythere and is very very solid and firm.
lets forget what winisnd said.
my question is really specific about adding internal mass/volume (ie reducing internal volume ft3) to raise the tuning frequency. i did cut the port already and it is way out of a potential solution as it is too much diffuclt to do.
OK, if you remove one or more drivers is the interior open enough to expand a fairly durable (garbage) bag packed with light weight rigid foam/whatever or multiple smaller loosely packed bags stuffed into crannies?
I understand what you're trying to do PatAllen. Running a sim w/WISD it looks like you'd need to take out about 3cu.ft. of volume to get the enclosure tuned to 32hz or so with the port length you have (leaving about 8cu.ft.). Down side is the low end response is going to be shifted higher up in frequency (f3 of 30hz up to f3 of 35hz and a bit shallower cut off slope w/reduced port gain). Not sure that fits in with your original design goal. Like GM suggested ... can't you pull both drivers and get in there with a Sawzall?
Last edited: