TOPIC:

My project is fully described in this forum thread www.diyaudio.com/forums/multi-way/231353...speaker-project.html



Idea
To make a partially dipole speaker in the spirit of the legendary Gradient 1.3.

System
Versions:
AINOgradient
AINOgradient Neo
AINOgradient NeoB
Dipoles need lots of eq and measurements/adjustments, so a minidsp 4x10HD is the way to control it. The lowest bass is a sealed box, midrange is true dipole 150-4kHz, highs with a ribbon and a backside assisting dome. Power amplifiers are B&O ICEpower 125 and 50 modules per speaker. My AV receiver is the input selector and feed analog signal to minidsp. Too bad that it does not have digital out with volume control. I used UMIK-1 and REW for measurements. All minidsp settings were done without REW-automation.

Measurements

With dipoles, it is crucial to know how directivity really is, 0-90¤ and 90-180¤. Other difficult aspect is distortion at high levels, because of poor acoustic efficiency of open drivers, specially with "nude drivers like in AINOgradient. I have not seen exactly comparable data of other dipoles, but I am rather sure that AINOs perform exceptionally well!

The sound of dipoles depends very much on positioning in the room (wall bounces) and room's reflectivity. The strong backwards sound must not interfere with forward sound, but compliment it By smooth late reflections. These issues can be seen in room response RTA, decay graphics and RT values .

More measurements:

I have done PowerPoint presentations of AINOgradient. If you want to get one, please email me to juha.sirkka(*)fimnet.fi
Summary - 4 pages
Full - 27 pages

Wow.. that's really neat info Juha! Is that a product on sale? Maybe we could put it in the "powered by miniDSP section"?
www.minidsp.com/applications/powered-by-minidsp
Keep us updated, very neat work!

Devteam

Thank You for asking,
but this is purely a personal project, I don't have any commercial interests. This is too far-out and unorthodox composition for marketing, but I am happy if any diy'er wants to use/copy my ideas and documented solutions and experience.

This project can be published in your Project gallery. Please give email address straight to Devteam/you! My email is in the previous messge
Juha

AINOgradient NeoC

minidsp 4x10HD-controlled, 4-way active hybrid sealed-dipole speaker
SEAS L26ROY (sealed monopole), Beyma MW12Nd, BG Neo8PDR, BG Neo3PDR
B&O ICEpower class D amplifiers, 125+50 per speaker

I finally pulled the trigger and purchased a pair of B&G Neo3PDR tweeters (without back cup or faceplate). Got them from Spain by eBay. Unused, in plastic bags. First I put them in normal vertical position in the frame, but measured response was not what I expected. Hopla, turn the thing 90¤ and it gets acoustically narrower because it does not fill the opening in the frame (see pic) Raw measurements, some eq, delay setting and xo set LR4 at 3500Hz - bingo!

Dipole action past 6kHz, excellent! The dip in direct response at 7kHz does not show in backside response, so it might be a "cavity effect" and managed with a new frame. Rear radiation was attenuated in NeoB, now it is equal and room decay measurements look more nice. There is more sparkles in the sound too, which is nice.

I will make new plywood frames for Neos in near future. Tweeter must have open space in the frame to maintain dipole action so high in F. Vertical directivity is not perfect because of too much M-T distance but tolerable, and the xo-range beam can be tilted with delay adjustment, now it is a bit upwards.

AINOgradient Neo83

I ended up to purchasing Neo3PDR tweeters too, it was obvious tha the large horn/magnet structure of Fountek dispturbed backwards radiation and back-to-back tweeters were too much a compromise. Crossover between subwoofer and lower mid is LR2, others LR4.

With planar mids and tweeters and narrow open baffle I was able to push acoustic dipole radiation pattern up to almost 7kHz and almost seamlessly. See polar graphics! My measurements are mostly done indoors and there are lots of reflection below 400Hz.

I attended a Finnish DIY-audio event held in May 2015, and AINOgradients were voted to give "Best Sound in show"!
Pictures from the event in this link

I found some more smoothed polar graps, based on same measurements. Backside measurement has only 4 different angles, frontside has 6.

Updated presentation of the project ready for download here, wellcome!
drive.google.com/file/d/0BzUGcoofhVkAbjR...eFk/view?usp=sharing

Did you have any impedance problems with minidsp vs. icepower, have read somewhere that it could be a problem?

Thank You for asking, but no I haven't recognized any such problems. Spl response, distortion and dB scale are clean and wide. I use 50asx2 and 125asx modules without any added circuits. I have used several other home-hifi integrated and power amplifiers too for testing and measuring, without any changes in measured performance.

I have been listening to AINOgradients for 3½ years now. Minidsp 4x10HD and ICEPower modules have had power on all the time without problems. Looks like I'm almost finished with dsp settings too, and my wheaten terrier Willa likes them too! All xo's are acoustic LR2. For extremly high spl mid and treble xo can be switched to LR4 by changing the preset in minidsp. On-axis response and polars stay the same horizontally, vertically they change a little.

Some 360¤ horizontal polars attached

Merry Christmas and a Happy new Year 2018!
Juhazi

Her Master's Voice

excellent thread. watching with keen interest, now that i’ve just discovered it.
i have the gradient helsinki’s, and am figuring out how how to incorporate corner-loaded
woofers (eq’d and time-aligned) with the helsinki’s midrange/tweeter. will probably
not use its dipole woofer, since the nanodigi’s dsp can do a lot of the work.
stereophile’s measurements are a nice little assist...
www.stereophile.com/content/john-atkinson-november-2010

cheers

Hi,
nanodigi should do it easily, just like 4x10HD but without AD/DA naturally. Be brave, just try different setups, do measurements of acoustic performace and listen to carefully. Indoor measurements are very very tricky, it would be nice if you can do polar measurements also outdoors or in a really huge hall, to get a solid understanding of how they work.

But then again, in a normal (acoustically small) room with Schroeder limit around 4-500Hz, measured and listened sound will be different! Biggest difference is in bass, because in most cases the sound from first reflection is louder than direct sound. This is what Atkinson in his post-script tells about, and makes the Helsinki so critical of positioning to front and side walls. I ended with all other drivers delayed 0.7ms after the bass unit, and bass eq and level must be checked if the speaker is moved more than 20cm!.

Gradient Helsinki is a different animal from the Classic 1.3 which my speaker mimics. Helsinki has dipole bass (sideways), cardioid midrange and monopole tweeter in a waveguide. Gradient 1.x or AINOs have monopole bass and dipole midrange + tweeter. This means that the whole package radiates in a very different pattern, with different transition ranges. I don't have any personal experience with Helsinki. Gradient Revolution is a bit more flexible .

To celebrate Finland 100years, Gradient released a new model in the spirit of 1.x, the 1.4. It has also downfiring monopole bass, crossed at 200Hz to a coaxial midrange-tweeter, which has practically cardioid radiation in it's passband. It should sound nice in almost every setup!

www.gradient.fi/en/content/9-gradient-story

Happy New Year!
Juha