3 Way bookshelf the easy way

[modulardesign]

This is the same poster from over on DIYaudio that used traced files and AI to come up with a design. The thread ended up being locked when nothing more positive came from it. The OP called his design faultless, and was not open to potential improvements.

What changes would you like to see

 

[modulardesign]

Here are off the shelf parts

 

[modulardesign]

looking at this thread here, we might not get kicked out of a party, we are looking pretty good, https://www.audiosciencereview.com/...way-bookshelf-the-easy-way.61859/post-2268563

 

[modulardesign]

here are the drivers and 1 foot 3inch diameter port, so total budget for a self assembled pair is in the $500 ball park, thats a $2000 speaker set right there

 

[Salt]

here are the drivers and 1 foot 3inch diameter port, so total budget for a self assembled pair is in the $500 ball park, thats a $2000 speaker set right there
View attachment 444290

Simply BMP:
- Build
- Measure
- Publish then

 

[a4eaudio]

What changes would you like to see

A picture of the actual speaker cabinets built, with drivers installed, and actual measurements.

 

[modulardesign]

Well said

 

[modulardesign]

Here's design 2

 

[Heinrich]

Here's design 2

You're experiencing an interaction between the drivers' impedance w/ the crossover. The mid ist mostly affected. A large coil, or a simple resistor in parallel to the driver may help.

 

[modulardesign]

This probably has better driver integration

 

[Wolf]

Nope, still not a good sim. You build cabs yet?

 

[modulardesign]

Nope, still not a good sim. You build cabs yet?

Seems your onto something that I may have overlooked, can you distill the sweeping statement into actionable specifics that would really help

 

[modulardesign]

The 8SW4 HE is still a very attractive something to try for cheap

 

[Azrael]

I see still the biggest problem of him importing probably just the on-axis response from the part-express dataset which makes everything else kind of "luxury problems"...

He uses Boxsim, which as far as I know, unlike VituixCAD, for example, can only import one frequency response per driver. This would mean that the additional frequency responses required for the pretty colorful polar plots are simulated at different angles based on the specifications of the enclosure.

 

[thewas]

He uses Boxsim, which as far as I know, unlike VituixCAD, for example, can only import one frequency response per driver. This would mean that the additional frequency responses required for the pretty colorful polar plots are simulated at different angles based on the specifications of the enclosure.

I use Boxsim myself and you have both options of either importing the measured angle responses in the specific enclosure or having Boxsim compute them based on the enclosure and the nominal diameter of the driver which of course matches reality less as such doesn't consider shape and break up of the driver.

 

[Azrael]

I use Boxsim myself and you have both options of either importing the measured angle responses in the specific enclosure or having Boxsim compute them based on the enclosure and the nominal diameter of the driver which of course matches reality less as such doesn't consider shape and break up of the driver.

Okay, then I might have to take a look at Boxsim after all, although it's very unlikely that I'll return to Boxsim (it's also been a very long time since I last used it).

I suspect that the OP used the second method.

 

[Azrael]

At the moment, it appears that simulating with measured angular frequency responses in Boxsim only works with the supplied driver data, which comes exclusively from Visaton drivers. These are probably angular frequency responses that have been measured in Visaton's (nearly)* anechoic chamber.

I still haven't found a place in Boxsim where you can import your self-measured multiple frequency responses per driver.

* "(nearly)" means that I once read somewhere that the anechoic chamber is anechoic down to 70 Hz.

 

[modulardesign]

The software does its best to simplify loudspeaker design by abstracting quite a lot from baffle step response, infinite baffle data, 200litre data, etc, spl behavior from user angle, driver depth from baffle, that means that even with optimistic manufacturer data things don't get out of hand. You will notice the almost text book perfect phase response from the last two designs, it even beats most two ways. Yes I acknowledge that the portrayed tweeter behavior off center may not perfectly track data however the off axis behavior is known, what's most critical is to be great up to 10kHz

 

[Wolf]

If it does not model baffle position of drivers, then these 2 models are seen as coincident and therefore easier to get phase alignment.
I don't know the program.
However, there are other things your sim does not do correctly or account for.

 

[Salt]

If it does not model baffle position of drivers, then these 2 models are seen as coincident and therefore easier to get phase alignment.
I don't know the program.
However, there are other things your sim does not do correctly or account for.

... besides that little hole in the lower corner on the right side seen from the front ... is it a mousetrap?