Introducing Directiva - An ASR open source platform speaker project

[ctrl]

but decided to share the latest form @ctrl.

There seems to be something wrong with the settings or parameter - attach the file too (but probably it's a problem with different option settings in VCAD).

It should look like this. It's a LR second order crossover. It should sound kind of warm, but you have to adjust the sound pressure level of the tweeter, because I never heard the speaker.

The crossover was adapted for the miniDSP 48k. Who uses another DSP must make adjustments.

The example crossover is only to show the possible potential of the speaker - which is extremely promising for the simple design. When comparing the crossover simulation of version v9d with other ASR measurements of loudspeakers, the potential becomes immediately obvious.

As Rick said, if you want to build the speaker, you have to have measurement equipment and be able to use it - everything else is probably doomed

 

[notabenem]

As Rick said, if you want to build the speaker, you have to have measurement equipment and be able to use it - everything else is probably doomed

Why exactly? Can't i build now, and put in the (passive) crossover later - when the measurements become available?
Or... can I build it now, put in some fusion/minidsp boards, let them collect some dust and upload the filters after AMIR has measured them?

 

[abdo123]

miniDSP 48k.

doesn't the 2*4 HD uses 96KHz processing?

 

[ctrl]

Why exactly? Can't i build now, and put in the (passive) crossover later - when the measurements become available?
Or... can I build it now, put in some fusion/minidsp boards, let them collect some dust and upload the filters after AMIR has measured them?

Anyone who builds a loudspeaker needs measuring equipment - at least for projects that are priced higher.

In many cases, the manufacturing tolerances of the drivers alone are enough to cause the frequency response of the finished speakers to differ by more than +-0.5dB, which should be avoided for the best stereo sound.
To compensate for this, you have to measure.
To find possible errors, you have to measure.

You can start building now, but it is always an advantage if you can measure - see above.

Since the Directiva project is the first attempt to develop a loudspeaker using Amir's NFS measurements and it is also the first loudspeaker design for Rick, there are still some uncertainties about the transfer from VCAD to the real measurements.

Once these initial uncertainties have been clarified, many things become easier, since the crossover simulation in VCAD is usually identical to the measurements of the finished loudspeaker.

 

[ctrl]

doesn't the 2*4 HD uses 96KHz processing?

This is the setting Rick needed.

As already said, the DSP used must be set in VCAD and the crossover adjusted.
Now that anyone can lend a hand, there should soon be crossover versions for all sorts of DSP's.

 

[Rick Sykora]

doesn't the 2*4 HD uses 96KHz processing?

You would have thought so (it is HD after all), but when I did some simple checks, the setting for the 48 kHz matched the actual filters better.

 

[Rick Sykora]

Why exactly? Can't i build now, and put in the (passive) crossover later - when the measurements become available?
Or... can I build it now, put in some fusion/minidsp boards, let them collect some dust and upload the filters after AMIR has measured them?

Would not hesitate to build as an active design, but passive should wait until we take get @ctrl to do and Amir to test.

 

[mdsimon2]

You would have thought so (it is HD after all), but when I did some simple checks, the setting for the 48 kHz matched the actual filters better.

Are you saying it matched better electrically or acoustically?

All electrical measurements I've done with a 2X4HD have exactly matched other proportional Q DSPs (rephase, camilladsp, biquad programming in octave based on the RBJ cookbook, other miniDSPs) so it is a bit surprising to hear it didn't match what you anticipated.

Michael

 

[Rick Sykora]

Are you saying it matched better electrically or acoustically?

All electrical measurements I've done with a 2X4HD have exactly matched other proportional Q DSPs (rephase, camilladsp, biquad programming in octave based on the RBJ cookbook, other miniDSPs) so it is a bit surprising to hear it didn't match what you anticipated.

Michael

To be specific, we were working on correlations and the project uses shelving filters in the crossover. When the project was set to minidsp 96 kHz, the simulated filter shape did not match the one in the minidsp HD. One I set the project dsp to minidsp 48 kHz, the shelving filters matched.

One key aspect is that the VituixCAD filter had a fixed Q and the minidsp allows Q to be manipulated. To match, the filter Q for the minidsp had to be set to 0.5. Not sure I saved all the tests, but recall I tested a couple of other filter types and did not have any issues with any other minidsp setting. As was busy sorting out other measurements issues, was happy to find a solution and move on.

Hope this helps clarify. I would expect the same filter to have the same shape regardless of which minidsp is used. I did not scrutinize further but you can see this issue in the posted project. If you only change the dsp choice, you can see the sim’s frequency response change significantly.

 

[Rick Sykora]

As it caught me as off with regard to earlier posts, wanted to point out that the polar map in @ctrl's view is a vertical axis one.

For those who might be wondering what the horizontal one looks like...

and it actually is better than this (see next post)…

 

[ctrl]

As it caught me as off with regard to earlier posts, wanted to point out that the polar map in @ctrl's view in a vertical axis one.
For those who might be wondering what the horizontal one looks like...

Rick, again there is something wrong with your crossover version (options in VCAD?). It actually looks like this - unsmoothed. The -6dB line is the black on yellow one.

 

[ctrl]

So, here's version v9d, the example crossover from above, in a slightly more common representation using VacsViewer.

Color change after 3dB. The -6dB contour is drawn as a black curve.

No smoothing at all.

horizontal +-180

vertical +-180

 

[ctrl]

Here is the sonogram of version v9d normalized to the axis frequency response.

Color change after 3dB. The -6dB contour is drawn as a black curve.
No smoothing at all.

horizontal +-180

vertical +-180

The directivity is not perfect, which is due to the standard enclosure, but better than many 2-way speakers measured by Amir.

 

[Rick Sykora]

Rick, again there is something wrong with your crossover version (options in VCAD?). It actually looks like this - unsmoothed. The -6dB line is the black on yellow one.


View attachment 146667

 

[Rick Sykora]

Thanks, Am using VituixCAD on different computer for multiple projects, but forgot they all share settings via Onedrive.

Will review. Do we need to post for others or will they get your setting from posted project?

 

[ctrl]

Do we need to post for others or will they get your setting from posted project?

Unfortunately VCAD does not remember all settings in the project file.
Therefore here are the used settings in version v9d:

 

[Biblob]

Here is the sonogram of version v9d normalized to the axis frequency response.

Color change after 3dB. The -6dB contour is drawn as a black curve.
No smoothing at all.

horizontal +-180
View attachment 146680

vertical +-180
View attachment 146681

The directivity is not perfect, which is due to the standard enclosure, but better than many 2-way speakers measured by Amir.
View attachment 146683

It's quite an narrow directivity speaker it looks like! Can't wait for the end result. Great job, guys!

 

[Rick Sykora]

Unfortunately VCAD does not remember all settings in the project file.
Therefore here are the used settings in version v9d:

View attachment 146721

Thanks, pretty sure we match up fully now.

Directiva fans please take note of above settings!

 

[ctrl]

It's quite an narrow directivity speaker it looks like!

Of course, everyone has to judge that for themselves. Compared to other speakers, the directivity is not narrow (not wide either).

The -6dB limit if normalized to the axial frequency response is as follows:

If one assumes an average sound pressure of 90dB (it's not sensetivity, only the average in the VCAD Version v9d), the -6dB limit (none normalized) looks like this:

In both cases the radiation is about +-60°.

For example, a 90-60 horn with +-45° hor and +-30° ver has a narrow directivity.

 

[alex-z]

Just a random thought, a Directiva-esque design could be created by converting KEF Q350 to an active crossover.

They are regularly available at $500/pair and the only major weakness is mid-range distortion, likely due to the first order crossover. Brings the total cost to $725 after the miniDSP 2x4HD. Saves money but costs bass extension vs using the Purifi woofer.

miniDSP also has their 2x8 kit, which is $343 after accessories. Could do a poor mans KEF R3 by putting Q150 drive units in a cabinet with a Dayton RS-225. Blows the .5ft cabinet sizing out of the water though.