@perceval . , weird that they would block...
They are only 3/8 inch thick, but I think @xrk971 has used 1/2 inch foam boards with paper skins before to good effect.
I wouldn't use it for a long term build, but a quick prototype?
And now you got me thinking... what if you laminate 2 together to get 3/4 inch board that a "constraining layer" of cork? That might be a stiff and non-resonant board...
Interested in any thoughts...
They are only 3/8 inch thick, but I think @xrk971 has used 1/2 inch foam boards with paper skins before to good effect.
I wouldn't use it for a long term build, but a quick prototype?
And now you got me thinking... what if you laminate 2 together to get 3/4 inch board that a "constraining layer" of cork? That might be a stiff and non-resonant board...
Interested in any thoughts...
He used the gator boards for very small and curvy designs, like the Nautilus shaped one, and the Karlsonators, with smaller 3" drivers.
For bigger projects involving woofers, at least 1" of XPS foam. Plus, it will be a lot cheaper than that board, limited at 20"x30"
XPS I got here was 36"x 72"x 1" for about $6.
For bigger projects involving woofers, at least 1" of XPS foam. Plus, it will be a lot cheaper than that board, limited at 20"x30"
XPS I got here was 36"x 72"x 1" for about $6.
Looks like cork board used for making bulletin boards to pin stuff. Not sure if stiff enough (if there is paper under the cork then it should be). But combine this with paper faced foam core and it would look quite nice and have damping abilities.
Screenshots for those unable to see link.
Screenshots for those unable to see link.
@xrk971 ,
yeah - that was my initial guess. I'm gonna try to do a quick build with it. I am also interested in gluing two together to see if it acts as constraining layer dampening with significant improvement.
I'd kinda like to to have the cork on the outside for appearance and handling.
wrt cost: yeah, I agree with @perceval , big sheets of XPS are cheaper. But if it's a nicer build of a small speaker, might be worth the extra cost.
yeah - that was my initial guess. I'm gonna try to do a quick build with it. I am also interested in gluing two together to see if it acts as constraining layer dampening with significant improvement.
I'd kinda like to to have the cork on the outside for appearance and handling.
wrt cost: yeah, I agree with @perceval , big sheets of XPS are cheaper. But if it's a nicer build of a small speaker, might be worth the extra cost.
Sometimes I've read where the jokes on you and those protos sound better than the final box material. God's humor; just because one wouldnt think so.
Oh, I did a cork laminate once on a pair of OB panels. They still ring some, but not as bad as plywood.
one more thing wrt cost:
- the image had a the single board price. I bought a 10 pack for $44 = ~ $1 per sq ft
- compare to 2" Owing FOAMULAR 4 x 8 ft sheet at Home Depot is $54 = ~ $1.69 per sq ft
I installed the latest Winisd and entered my driver parameters and started my first project.. For a accidental MLTL. The driver is a 6 inch Fullrange, made in India. In the vented option we have many like EBS6, EBS3 etc.. Which one to choose?
I chose EBS6 and did simulate a graph... Now I get Box volume of 79.09 litrs, Tuning freq - 39.67 HZ and vent area 81.7 sq. cm. Now in the advanced option ( below vents, filters, and signal) , if we click, we get an option to check "use transmission line model for port simulation.. This changes the response graph... To check it Or not? I simulated with transmission line model ticked on and I get the green graph as in the image... Can someone point me am I in the right direction? here are some screen shots showing response with and without transmission line model
I chose EBS6 and did simulate a graph... Now I get Box volume of 79.09 litrs, Tuning freq - 39.67 HZ and vent area 81.7 sq. cm. Now in the advanced option ( below vents, filters, and signal) , if we click, we get an option to check "use transmission line model for port simulation.. This changes the response graph... To check it Or not? I simulated with transmission line model ticked on and I get the green graph as in the image... Can someone point me am I in the right direction? here are some screen shots showing response with and without transmission line model
Last edited:
I didn’t know WinISD has a TL mode. I am not sure if what it shows matters because you just need to follow the recipe here and assume bass reflex. It will work out if driver has moderate Qts.
With the transmission line model selected, no end correction is added to the port tube length. This makes the effective length shorter (it becomes the same as the physical length) which raises the Helmholtz resonance frequency as shown in the chart comparisons.
If you want to design a true bass reflex loudspeaker then the dimensions of the port tube need to be such that the air inside acts as an acoustic mass. This means that the transmission line model should not be used for the simulations.
EDIT 1 - I just realised that you wanted to design a MLTL rather than a BR 🙂.
EDIT 2 - Note that there is an error in the WinISD transmission line model. The post linked below refers:
Last edited:
So, I should uncheck the Transmission line model and take the values of volume and Port dimension.. Another question- Do I need to deduct port dimension and create a Box volume for wood panel dimension?
Thanks
Thanks
Hi, Thanks. The driver had qts 0.69,Fs 77hz and Vas 18.2 litres.. So I should take the Box volume of 80 litres, convert into a transmission line of say 36 inch length and size the CSA. The port area should be excluded from the volume?
Here are some updates.
The winisd with default optimum works well with round port. if it is converted to AMLTL as described here, the vent should be in a round port.
How did i arrive at this? i built a tall cabinet 48 inches, keeping the volume same as modeled,but making the port rectangular with 35cm wide, and 10cm height, depth 20cm ( 35 cm because wanted a slot loaded port that is same as front baffle width)port tuned to 43hz.
I built it and i found the bass lacking with shouty upper mids.which is not similar to the flat line in simulation.
I came back to winisd and using the same volume, with rectangular port as above, when switched to use transmission line model, i noticed the image as below - peaks at upper mids, which confirmed my listening experience
I now adjusted the port dimensions,keepin the 35cm same as baffle width, but reducing the height slowly from 10cm to 1.8cm,at which point i got a flat and extended response. the depth is now 7.1cm. It also tunes the vent to 30 hz.
I rebuilt the cabinet altering the port dimensions
with stuffing of upper 1/3 the peaks in the upper harmonics reduces, and i now got a good sounding MLTL style box. so in short the TL MODEL works, by tuning the vent lower than regular bass reflex, but the port should be designed as a slot loaded one, similar to WIBAQ,TABQ, and the other two dimensions derived to achieve flat response
Last edited:
A 35cm x 10cm wide vent is huge - from experience as we never see a bass reflex vent that big. A vent that big is a true TL where the tuning is based on the length of the line. It makes sense that vent works when reduced to 35cm x 1.8cm cross section.
Typically, a bass reflex vent cross sectional area is only ~15% of the cone Sd. AMLTL vents will mimic a bass reflex vent in approximate size.
Typically, a bass reflex vent cross sectional area is only ~15% of the cone Sd. AMLTL vents will mimic a bass reflex vent in approximate size.
With regular bass reflex designs, when we design the port to such small size as 1.8cm, we get - negative length Or length which is much less than than the plywood thickness itself! Check the box for use TL model and the vent depth now becomes 7.4 cm in my case.... I tried this multiple times and got the same response.. In addition, we also notice the critical dimension is very narrow.. The slot loading works only between 1.8 to 2.4cm.. Nearing 3cm, the frequency response suddenly drops, Just an incidental finding and I think it works very well
It can also be stated this way.. With the default box by winisd, vent is tuned to 43hz.. Keeping the volume same and if the vent is tuned to 30hz, the frequency response dips much below the Fs. But same volume and vent tuning Gives extended flat response when TL model is checked...Only thing the vent size to be adjusted to get flat response...