• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Small 2-way speakers with linear on-axis and power response characteristics (Scan Speak and SB Acoustics drivers). H&V off-axis measurements included

Such a nice design! I'm also from northern Europe and hope to get into DIY speakers. Do you have any advice about retailers for parts? Most chatter I see online order from partsexpress in the US. :/
 
Congratulations on the review. Your hard work certainly paid off.
 
Well done! It's surprising how well of a PIR you've got with just a basic sharp cornered rectangular baffle.
 
@XMechanik, as I said in the EQ post that's a great effort with only off-the-shelf components and simple woodwork,
I noticed that in your simulation the LF is not drooping as much as in the NFS data. Not sure why.
Maybe near field measurement inaccuracy?

The simulation of the 8L + tuning to 45Hz seems to match the NFS data characteristic
If I may, after doing a bit of simulation for the box tuning I would suggest tuning the cabinet to 51 - 53Hz instead.

That would result in a flatter response (up to +2dB above 50Hz) and most probably remove the need for EQ IMO (outside of room integration obviously).
The LF will be attenuated slightly but the power handling higher above 50Hz (about +3dB) so more usable in the useful bandwidth of the speaker.
All in all for a smaller speaker it might be a better compromise especially if listened louder.
 
Maiky76 said:
@XMechanik, as I said in the EQ post that's a great effort with only off-the-shelf components and simple woodwork,
I noticed that in your simulation the LF is not drooping as much as in the NFS data. Not sure why.
Maybe near field measurement inaccuracy?

The simulation of the 8L + tuning to 45Hz seems to match the NFS data characteristic
If I may, after doing a bit of simulation for the box tuning I would suggest tuning the cabinet to 51 - 53Hz instead.

That would result in a flatter response (up to +2dB above 50Hz) and most probably remove the need for EQ IMO (outside of room integration obviously).
The LF will be attenuated slightly but the power handling higher above 50Hz (about +3dB) so more usable in the useful bandwidth of the speaker.
All in all for a smaller speaker it might be a better compromise especially if listened louder.
Click to expand...
Any chance you've computed the port dimensions for such tuning?
 
It’s an off the shelf port so you just bandsaw a few mm off and call it good usually.
 
Incredible. Nice work @XMechanik and thank you again for sharing this design.

My build came in around $300-350 usd for drivers and x-over components. Outstanding performance per $.
 
somebodyelse said:
It's an adjustable telescopic one, so both easier and reversible if you don't like the result.
Click to expand...
Also means you could adjust it for each room. Neat. Way better than a precision port, even.
 
This really makes me want to get into DIY! @XMechanik, I realise I am asking for free stuff here, but if you made a youtube video showing off the speaker you would get a lot of views :)
 
The surface of the port is too small relative to the active surface of the driver (about 11%).
Even at moderate SPL the air velocity in the port exceeds 5% of the speed of sound (343m/s) which is one of the criteria to dimension a port.
The port will be non linear very quickly (as the SPL increases) and restrict the output of the system at LF.
I am not sure what level is used for the NFS but it might already be the case during Amirm'smeasurements.

If you look at the review you'll see that the resonance just above 1k that matches the expected peak at 1100Hz and is only about 10dB down compared to the main output.
Could be better... but I understand the convenience of the solution adopted by @XMechanik

For the following sims I have included 0.77R in series with the driver because of the Xover I was lazy to do the Xover sim and check the actual value....

The sizes
45Hz -> 150mm instead of 145mm (my calculation are 3.5% different from the actual recommended port length not significant IMO)
20240430 Mechano23 default + NFS.png


52Hz -> 106mm, first resonance predicted at 1600Hz
20240430 Mechano23 52Hz.png

The point is: DIY!
So I would recommend a port that provides better linearity (larger diameter) tuned to 51 - 53Hz instead.
 
Last edited:
Maiky76 said:
The surface of the port is too small relative to the active surface of the driver (about 11%).
Even at moderate SPL the air velocity in the port exceeds 5% of the speed of sound (343m/s) which is one of the criteria to dimension a port.
The port will be non linear very quickly (as the SPL increases) and restrict the output of the system at LF.
I am not sure what level is used for the NFS but it might already be the case during Amirm'smeasurements.

If you look at the review you'll see that the resonance just above 1k that matches the expected peak at 1100Hz and is only about 10dB down compared to the main output.
Could be better... but I understand the convenience of the solution adopted by @XMechanik

For the following sims I have included 0.77R in series with the driver because of the Xover I was lazy to do the Xover sim and check the actual value....

The sizes
45Hz -> 150mm instead of 145mm (my calculation are 3.5% different from the actual recommended port length not significant IMO)
View attachment 366821

52Hz -> 106mm, first resonance predicted at 1600Hz
View attachment 366817
The point is: DIY!
So I would recommend a port that provides better linearity (larger diameter) tuned to 51 - 53Hz instead.
Click to expand...
I have seen the standard box and port calculators online where you enter the T/S parameters but none that create port sizes given an input tuning frequency. Can you point me in the right direction please? Thanks.
 
Maiky76 said:
The surface of the port is too small relative to the active surface of the driver (about 11%).
Even at moderate SPL the air velocity in the port exceeds 5% of the speed of sound (343m/s) which is one of the criteria to dimension a port.
The port will be non linear very quickly (as the SPL increases) and restrict the output of the system at LF.
I am not sure what level is used for the NFS but it might already be the case during Amirm'smeasurements.

If you look at the review you'll see that the resonance just above 1k that matches the expected peak at 1100Hz and is only about 10dB down compared to the main output.
Could be better... but I understand the convenience of the solution adopted by @XMechanik

For the following sims I have included 0.77R in series with the driver because of the Xover I was lazy to do the Xover sim and check the actual value....

The sizes
45Hz -> 150mm instead of 145mm (my calculation are 3.5% different from the actual recommended port length not significant IMO)
View attachment 366821

52Hz -> 106mm, first resonance predicted at 1600Hz
View attachment 366817
The point is: DIY!
So I would recommend a port that provides better linearity (larger diameter) tuned to 51 - 53Hz instead.
Click to expand...
...or use a PR and there will be no port resonance, SBAcoustics for example offers PRs with twice the Sd of the woofer - but this increases the price...
 
Assembling and electronic work I am well acquainted with. However, I am new to from-scratch speaker builds. I do not have the facility or accommodations to do woodworking currently. Where would be the best place to get the panels cut?
 
Thanks @XMechanik very nice design!!

If using in a small HT with 3+ SW, would you recommend a sealed alignment?? Any changes on the volume or XO?

Cheers!
 
You must log in or register to reply here.
Back
Top Bottom