You're welcome!
FWIW, I've used this converter since its inception = 23.64655 L, so whatever you're using is correct and while HR shows it rounded to 23.65 L it's correct out to at least several decimal places IIRC from a very dim memory of this subject way back when.
Anyway, 'we' just accounted for the inside volume of the vent = (3.5^2*pi)/4 = 9.62113^2*3 = 28.86338"2/186.215 cm^2 or the way I do it is: 3.5*sqrt(3) = (6.06218^2*pi)/4 = 28.8634 cm^2, but ideally would use the outside diameter (o.d.) and subtract the thickness of the baffle to get its net internal length. Its internal flare is small enough in this case to ignore.
Re SBL, I 'fiddled' with it some awhile back and found 'issues', wrong published driver specs and emailed about some of them with no responses, so didn't remember it had the feature nor would have suggested it.
That said, it reminded me of this (very flexible) calculator for doing on-line BPs I've yet to vet, but might want to try it in lieu of learning HR (at least for now), though at a glance doesn't do any driver optimization.
FWIW, I've used this converter since its inception = 23.64655 L, so whatever you're using is correct and while HR shows it rounded to 23.65 L it's correct out to at least several decimal places IIRC from a very dim memory of this subject way back when.
Anyway, 'we' just accounted for the inside volume of the vent = (3.5^2*pi)/4 = 9.62113^2*3 = 28.86338"2/186.215 cm^2 or the way I do it is: 3.5*sqrt(3) = (6.06218^2*pi)/4 = 28.8634 cm^2, but ideally would use the outside diameter (o.d.) and subtract the thickness of the baffle to get its net internal length. Its internal flare is small enough in this case to ignore.
Re SBL, I 'fiddled' with it some awhile back and found 'issues', wrong published driver specs and emailed about some of them with no responses, so didn't remember it had the feature nor would have suggested it.
That said, it reminded me of this (very flexible) calculator for doing on-line BPs I've yet to vet, but might want to try it in lieu of learning HR (at least for now), though at a glance doesn't do any driver optimization.
good point
The baffle depth or material thickness should be subtracted.
Hate to be the bad guy here ... but
Both you guys had numbers which were near correct, except one decimal point off.
.835 cu ft or 23.6 liters
were posted numbers
it is .0835 cu ft or 2.36 liters
Dont worry, it has happened before.
Same thing will happen when guys use
different enclosure calculators.
With 3 different posters you get 3 different outcomes
it is a combination of human error and program formula differences
far as port displacement
most people are familiar with 2 liter beverage or soda containers.
which is a easy way to imagine physical volume.
if the ports where 23 liters
it would be close to 11 large soda containers
obviously it is easy to imagine or see
they are not that large
one decimal place has a big difference.
the 3 ports combined displace
pretty close to one 2 liter bottle
far as simplicity a radius of 1.75" or diameter of 3.5"
Length is 5 inches but baffle material of .75
makes the length 4.25 for displacement
A single port would be .670 liters or .023 cu ft
then multiply by 3
2.01 liters or .069 cu ft
The baffle depth or material thickness should be subtracted.
Hate to be the bad guy here ... but
Both you guys had numbers which were near correct, except one decimal point off.
.835 cu ft or 23.6 liters
were posted numbers
it is .0835 cu ft or 2.36 liters
Dont worry, it has happened before.
Same thing will happen when guys use
different enclosure calculators.
With 3 different posters you get 3 different outcomes
it is a combination of human error and program formula differences
far as port displacement
most people are familiar with 2 liter beverage or soda containers.
which is a easy way to imagine physical volume.
if the ports where 23 liters
it would be close to 11 large soda containers
obviously it is easy to imagine or see
they are not that large
one decimal place has a big difference.
the 3 ports combined displace
pretty close to one 2 liter bottle
far as simplicity a radius of 1.75" or diameter of 3.5"
Length is 5 inches but baffle material of .75
makes the length 4.25 for displacement
A single port would be .670 liters or .023 cu ft
then multiply by 3
2.01 liters or .069 cu ft
Last edited:
FYI/FWIW, I tried this seemingly flexible calculator's 3 box, zeroing out Vr2 and all ports except Port3, selected the Vr1+Vf1 enclosure variant, but won't calculate. 🙁
Hmm, the math I did was right as far as it went, just in my usual doing the post in bits n' pieces didn't vet Rick's math and forgot to factor in the vent length after pointing out the baffle thickness reduction! 
Anyway, 186.215 cm^2*(4.25*2.54) = 2010.191/1000 = ~2.01 L, so thanks for the correction.
Anyway, 186.215 cm^2*(4.25*2.54) = 2010.191/1000 = ~2.01 L, so thanks for the correction.
so with reasonable agreement on measurements
you basically now know with driver, port, and amp cavity displacement.
We have 2 equal chambers of 36 liters
and a total port cross sectional area of 186.2 cm^2
So far I have tried WinIsd , VirtuixCad and get the same overall transfer function
I could try a 1/2 space response in HornResp.
It will likely be the same. Horn Resp will give much more detailed frequency response
and show port resonance. But the overall response will still be the same.
Most simulators for bandpass will show a " smoother" response
because they dont model port resonance like hornresp will.
WinIsd and many others will tell you what the 1st port resonance frequency is.
But the transfer function model wont show the actual resonance.
I have tried multiple suitable 12" drivers
and basically the outcome will always be the same.
There will be a peak on the upper frequency
Which most people dont do in a simulator because visually it " does not look right"
Important thing to do is apply the crossover filter
then in the simulator it will look correct.
Often people will make the response pretty and flat with no crossover in simulation.
and they dont realize in practice, they actually loose bandwidth.
I typically design with the crossover filter.
So without the filter it usually looks ugly.
or it " wont look right" until you do a complete
realistic model with crossover.
you basically now know with driver, port, and amp cavity displacement.
We have 2 equal chambers of 36 liters
and a total port cross sectional area of 186.2 cm^2
So far I have tried WinIsd , VirtuixCad and get the same overall transfer function
I could try a 1/2 space response in HornResp.
It will likely be the same. Horn Resp will give much more detailed frequency response
and show port resonance. But the overall response will still be the same.
Most simulators for bandpass will show a " smoother" response
because they dont model port resonance like hornresp will.
WinIsd and many others will tell you what the 1st port resonance frequency is.
But the transfer function model wont show the actual resonance.
I have tried multiple suitable 12" drivers
and basically the outcome will always be the same.
There will be a peak on the upper frequency
Which most people dont do in a simulator because visually it " does not look right"
Important thing to do is apply the crossover filter
then in the simulator it will look correct.
Often people will make the response pretty and flat with no crossover in simulation.
and they dont realize in practice, they actually loose bandwidth.
I typically design with the crossover filter.
So without the filter it usually looks ugly.
or it " wont look right" until you do a complete
realistic model with crossover.
Actually, the specs make a flat response in my HR sim and a 150 Hz/4th deals with the single large peak around 62 Hz. Changing the vent length and net Vbs to 36 L only changes the response a fraction of a dB.
The Infinity Kappa requires a 65 Hz/4th with a slightly narrower BW resulting in a ~7 dB lower efficiency.
The Infinity Kappa requires a 65 Hz/4th with a slightly narrower BW resulting in a ~7 dB lower efficiency.
Last edited:
Interesting. good find
WinIsd has The Kappa Perfect 12 in the database
The suggested volume for 4th order BP is very very close
to the OP's box.
Probably one of the better matches I have seen so far.
I see a few versions
Perfect 12 , 12.1 and 12.1d
All seem to work well.
Fine tuning the ports might help.
You could either add length to what is there.
or plug one to lower tuning.
as mentioned before, sometimes people measure the port flare.
the flare might be 3.5" the actual port maybe around 3 to 2.83
either way kappa perfect seems like a good match
WinIsd has The Kappa Perfect 12 in the database
The suggested volume for 4th order BP is very very close
to the OP's box.
Probably one of the better matches I have seen so far.
I see a few versions
Perfect 12 , 12.1 and 12.1d
All seem to work well.
Fine tuning the ports might help.
You could either add length to what is there.
or plug one to lower tuning.
as mentioned before, sometimes people measure the port flare.
the flare might be 3.5" the actual port maybe around 3 to 2.83
either way kappa perfect seems like a good match
My apologies for the misplaced decimal. Your idea about visualizing a 2-liter bottle really helps.
Question: The existing Paradigm amplifier only has either a 130 or 150-watt RMS output. Also, the original Paradigm driver measures 6.5 ohm with my multimeter suggesting it is an 8 ohm driver. Do either of these facts impact the possible fit for the Kappa Perfect 12.1 driver? (I only see 4 ohm versions online)
I read online that someone replaced the original Paradigm driver with the Goldwood driver linked below and claimed to have improved response.
Does it model well in your software?
https://www.parts-express.com/Goldw...p5GeSrQHtTb6QNJH7fs0l8k8RszgZCfUaAsq_EALw_wcB
Question: The existing Paradigm amplifier only has either a 130 or 150-watt RMS output. Also, the original Paradigm driver measures 6.5 ohm with my multimeter suggesting it is an 8 ohm driver. Do either of these facts impact the possible fit for the Kappa Perfect 12.1 driver? (I only see 4 ohm versions online)
I read online that someone replaced the original Paradigm driver with the Goldwood driver linked below and claimed to have improved response.
Does it model well in your software?
https://www.parts-express.com/Goldw...p5GeSrQHtTb6QNJH7fs0l8k8RszgZCfUaAsq_EALw_wcB
Last edited:
Ugh! Didn't notice that when I looked at its specs, i.e. 4 ohm or dual 4/8 ohms is very common these days, so with the Lab12c 4 ohms it may not work either, but at least we know what specs to look for.
Historically, GW drivers don't have accurate enough published specs, i.e. worse performance wise, but FWIW it's way under-damped, though once low, high passed it looks OK, just with low efficiency, low Xmax needs to be at least in 1pi or 0.5pi space where it has a decent output above ~50 Hz, so not really a 'sub'.
