Introducing Directiva - An ASR open source platform speaker project

[Rick Sykora]

Decided to capture some of the work I have done with my lesser equipment. My main goal was to do an initial minimal crossover, but use it to compare the VituixCad (VC) model to what results were achieved. The VC model was pretty good but did not project the tweeter response accurately until I set the DSP option to be a miniDSP. Note I did not follow the usual minidsp instructions. Notably, I used minimal eq to get these results...

The main eq applied was to bass boost and boost the tweeter rolloff above 8 kHz. As planned, stuck to the 2.4 kHz LR2 crossover. The red plot is the result of my design. The blue trace shows reverse null when the tweeter polarity is reversed. As before, the measurements where taken at 75db, 1 meter, on-axis with a 3 ms gate (no smoothing).

The bass boost was about 6 dB at 100 Hz and a 30 Hz 2nd order HP filter was also applied (mainly to avoid exceeding the Purifi's excursion limits). In this alignment, the woofer is capable of over 100 dB of output at 40 Hz. My quick listening test revealed bass that was quite solid although my room has some serious modes at the speaker's location. With 2 speakers, I suspect the bass will be very pleasing for most music. I tried to keep port resonances tamed, but damping was not optimized and suspect it could be better. Will take some nearfield measurements and share tomorrow.

Please keep in mind this is all quite early in the Directiva development and expect much more refinement once I have a unit to send to Amir for a spin. I need at least one round with my upgraded equipment before I consider sending to him.

 

[Steve Dallas]

Timely thread and interesting approach to the design. I have recently been pondering doing the same thing and had even targeted the same flat pack for my test mule--especially considering the sale price. I even had the same goals: low distortion and wide, even directivity. I was in the middle of driver selection, when I decided I don't need any more speakers and don't need to spend a bunch of $$ on drivers just to have something to keep me busy for a month or two. The only real "justification" for my project would be that I already have a 2x4HD and 4 channels of amplification sitting around and haven't done a speaker project in several years.

(Who am I kidding? Everyone knows I will pick this up again. Especially since I have some very nice spalted maple veneer begging to be used. Alternatively, I have always wanted to build a pair of speakers finished in Mazda Sparkling Black Metallic.)

I has never occurred to me to boost in the DSP domain. If had your initial measurements, I would have pulled everything down to ~73dB and crossed near the same point you did with those drivers. I don't have a technical reason for why I have never boosted; it just seems "safer," and Class D Watts are quiet and inexpensive.

You have not discussed any damping strategy for the cabinet, and I don't see anything in any photos. What are you doing there?

Anyway, watching with much anticipation!

 

[Rick Sykora]

Timely thread and interesting approach to the design. I have recently been pondering doing the same thing and had even targeted the same flat pack for my test mule--especially considering the sale price. I even had the same goals: low distortion and wide, even directivity. I was in the middle of driver selection, when I decided I don't need any more speakers and don't need to spend a bunch of $$ on drivers just to have something to keep me busy for a month or two. The only real "justification" for my project would be that I already have a 2x4HD and 4 channels of amplification sitting around and haven't done a speaker project in several years.

(Who am I kidding? Everyone knows I will pick this up again. Especially since I have some very nice spalted maple veneer begging to be used. Alternatively, I have always wanted to build a pair of speakers finished in Mazda Sparkling Black Metallic.)

I has never occurred to me to boost in the DSP domain. If had your initial measurements, I would have pulled everything down to ~73dB and crossed near the same point you did with those drivers. I don't have a technical reason for why I have never boosted; it just seems "safer," and Class D Watts are quiet and inexpensive.

You have not discussed any damping strategy for the cabinet, and I don't see anything in any photos. What are you doing there?

Anyway, watching with much anticipation!

Will be sharing more data after the next pass. The current damping is nothing fancy. Am using just enough dacron sheets to cover the sides and ensure am able to closely match the modeled cabinet tuning.

Know what you mean on crossover design and amplifier power. Am not saying I am sure about the approach I took as yet, but thought it might be worth while to use the the gain in the dsp to get better overall sensitivity. Have to admit, the baffle step made this challenging. I had an earlier pass that was about 6 dB more sensitive above 200 Hz. Clearly, the low bass was going to be challenged to keep up, but the extra output was enticing on a with the lesser minidsp 2x4 and a Dayton APA150 at almost full volume.

 

[abdo123]

BTW i hope you provide people the option to use a miniDSP for crossover (like what you're doing now in the design stage) as it is just far superior for people who have the equipment for it.

very interesting progress so far!

 

[notabenem]

$1000 in terms of budget will be for a single cabinet? Do you also plan on creating options for lower budgets? A multichannel setup with this budget per cabinet gets quite expensive and often an overkill for the surrounds.

 

[Rick Sykora]

$1000 in terms of budget will be for a single cabinet? Do you also plan on creating options for lower budgets? A multichannel setup with this budget per cabinet gets quite expensive and often an overkill for the surrounds.

Hi and welcome to ASR!

The primary target at the ~$1000 budget is a stereo music offering.

As planned to be extensible, agree that some other price points will need to be considered eventually. For surrounds, wireless connectivity would certainly be welcome in my home.

 

[sgoldwin]

Gee Rick, maybe I'm just easily impressed, but these seem like excellent initial results with little overall effort. What could be the possible advantage of a passive crossover here? What are you giving up by using the miniDSP?

 

[buz]

Gee Rick, maybe I'm just easily impressed, but these seem like excellent initial results with little overall effort. What could be the possible advantage of a passive crossover here? What are you giving up by using the miniDSP?

Would be easier to use with existing amplifiers (and save a bit on electronics) but I am not too sure it would be worth it.

 

[notabenem]

As an option, we can put a whole Raspberry with a good DAC shield into every single cabinet and then do DSP and everything with central management.

 

[Rick Sykora]

Gee Rick, maybe I'm just easily impressed, but these seem like excellent initial results with little overall effort. What could be the possible advantage of a passive crossover here? What are you giving up by using the miniDSP?

This is a just the start and is limited to on-axis at this point. @ctrl has done a lot of work simulating and I have done modeling in Bassbox and VituixCAD. Have also done quite a bit of measuring too. Have not posted, but will share more of the process shortly. Software and other tech advances have made the speaker design task easier, but there is a lot of experience driving it too. The minidsp is really a huge advantage - overall the main concession is the cost. A passive crossover would be cheaper.

Will consider some off-axis checks, but will get more complex as we go into off-axis and other enclosure options. Ultimately, Amir will do a spinoroma and will make more adjustments.

 

[buz]

As an option, we can put a whole Raspberry with a good DAC shield into every single cabinet and then do DSP and everything with central management.

Good luck getting that all in sync over IP. It's not impossible but far from trivial.

 

[notabenem]

Good luck getting that all in sync over IP. It's not impossible but far from trivial.

True. Signal distribution via IP is not the best idea, probably, even though HEOS AFAIK does it(?). I specifically meant signal processing only, but then again, there may be more purpose built and more efficient solutions for that (though not sure if also SW and REMOTE controllable or not). Anyway, let's get back to the original topic.

 

[somebodyelse]

True. Signal distribution via IP is not the best idea, probably, even though HEOS AFAIK does it(?). I specifically meant signal processing only, but then again, there may be more purpose built and more efficient solutions for that (though not sure if also SW and REMOTE controllable or not). Anyway, let's get back to the original topic.

Something similar has come up in the streamer thread too - I've proposed it as a possible side-project in the new parent thread.

 

[Rick Sykora]

After doing some nearfield measurements, despite the Bassbox sim indicating a major reduction, the port resonance reared its ugly head and so reduced the port to 2x4 inch one. F3 bumped up from around 44 Hz to 52 Hz. Have not listened to as yet, but here is the pic...

I can say that with the back firing port, I saw no evidence that the port resonance was affecting the overall response. However, this may be difficult to determine as is in the range of baffle step transition and may be masked by it.

Will investigate after I complete my equipment upgrades tomorrow...

 

[McFly]

I don't know why more people haven't explored this (plumbers delight) method for eliminating port resonances - I mean other than the obvious space constraints - could be hidden within a floor stander easy enough.

 

[tktran303]

Hi Rick,

Yes I agree that the audibility of port ‘problems’ lower than the visibility in the measured trace.

I strongly suspect this has got to do the masking effects. In the same range where the port is delivering spurious output, the midwoofer is playing about 15-30dB louder, and the tweeter contributing sound as well, except around 250Hz to 1KHz is around 12-24dB down.

I’ve ALWAYS used rear firing or downfiring ports since my days of speaker building (15+ years). Because the other advantage is aesthetics-
a clean look from the front or side.

For those concerned that the port is too long a down firing port is always an option.

 

[Rick Sykora]

Am using the bottom firing approach on one of my other designs. Not using it for any of the early version of Directiva due to shipping costs. Also, the Purifi woofer does well in smaller box with passive radiators.

As there is an alternative, will be exploring the PR options more with the improved directivity with the reduced baffle cabinet.

 

[tktran303]

I’m getting good models with 12-13L vented for Fb: 43Hz, which gives similar response to my current 19L Fb 39Hz... but in a lot smaller cabinet

My PR modeling shows good responses around
9-10L with the SB acoustics oval PR.

Obviously room size, speaker placement, bass preferences and crossover design all affects what is “best” but ultimately personal preference dictates a lot of what is preferred in terms of bass taste.

I like to exceed expectations with big bass in small spaces, and trying to beat what was previously possible under 1/2 cu foot...

 

[Wolf]

Any chance you could share the in-box impedances for the passive-xover inclined? I know the DSP means you don't need them, but I'm curious...

 

[McFly]

Did you need to set any delay on the drivers or are the acoustic centres close enough