Interesting aesthetics (not DIY)

[Paul Ebert]

Well, I think these are quite interesting. What do you think? They are the Gradient Helsinki's. I wonder if the back of the woofer is enclosed. Sidefiring dipole strikes me as a dubious design decision.

I'd prefer the tweeter and midrange 'baffles' to be oak or solid black. Seems to me that the tweeter 'baffle' is a shallow waveguide. Interesting...

SWMBO was quite unimpressed. I'm not too surprised.

[Saurav]

Finally found a photo of the other side:

Sidefiring dipole strikes me as a dubious design decision.

Maybe if you crossed really low, to where the room reverberant field diffused the sound, or... I don't know what I'm talking about. And unless that woofer has massive Xmax, I don't see how it could play very low anyway.

I also saw the phrase "cardoid midrange" somewhere, so that disc thing behind the midrange must be loading the rear wave in some specific way.

Edit: Look at how they're placed, they're toed in quite a bit, so the listener wouldn't be right at the dipole null. I was wondering how tall these were - the tweeter is pointing up, and it doesn't look like a very short speaker. And waveguide loaded. Hmm. Definitely a unique design.

[Jonasz]

Here's a review... http://groups.yahoo.com/group/regsau.../message/21574

[Paul Ebert]

After thinking about it a bit, these remind me of Dan N's Duo-S design with a dipole woofer added. In particular, the mid orientation provides a partial omni dispersion (as in the Duo-S).

The band around the mid 'baffle' appears to be perforated (see attached photo from the Gradient website). So, part of the backwave is emitted there. Very interesting...

It is also interesting that they are single-amped and, apparently, passively crossed-over. I wonder how efficient they are.

We expend a lot of effort on driver selection and crossover design, as well we should. But, I wonder if we might make more progress by experimenting with unorthodox approaches such as these.

Food for thought.

Thanks for the link to the review. I'm glad to have REG's website pointed out to me. I always felt that he was one of the better reviewers at TAS. He wasn't afraid to speak his mind and he had at least some scientific backing to his subjective approach (not that I'm against a subjective approach).

[Xander]

First reaction: ugly

Just way too much going on there

I do wish I could hear them though

[Paul Ebert]

Well, that wasn't my first reaction, though I can understand you're feeling so and I agree that there's too much going on. I do think that could be ameliorated without too much trouble, though perhaps not enough for some folks.

I would also really like to hear them. At $8000 and limited distribution, not much chance of it though.

I'd also like to experiment with some of the design decisions. Not a clone, but an 'influenced' design. Hmm... I have those D26s and the RS180s. Pick up some AE IB15s and I'm good to go. Well, there are some hurdles like finding someone with a lathe who could make the tweeter waveguides.

[dawaro]

The band around the mid 'baffle' appears to be perforated (see attached photo from the Gradient website). So, part of the backwave is emitted there. Very interesting...

Could this be a form of an AP vent to compensate for the enclosure?

Speaking of unorthodox I saw this dipole sub on AudioCircle yesterday. It definitely would fall into that catagory.


Edit by moderator to remove forum graphic. That forum doesn't allow direct links to images.

[matt12v]

actually the midrange has the motor structure from a hunter ceiling fan. Very low distortion driver! lol j/k I read some reviews on this speaker, it's very intriguing and earned some awards. One said it was favored over the Orion? Very interesting midrange enclosure... Somebody reverse engineer it! I love outside the box designs.

[jimluu]

can one of you hifi gurus comment on the technical design? The woofer baffle is relatively narrow and side-firing. Wouldn't the listener be sitting in a null area? The way the mid and high are angled, the listener is a good 30' off axis? am I correct to say that technically this thing doesn't make any sense?

[Rudolf]

can one of you hifi gurus comment on the technical design? The woofer baffle is relatively narrow and side-firing. Wouldn't the listener be sitting in a null area?

No guru, but:
The speakers are toed-in 45° WRT the front wall. This seems to be the recommended placement. This should lead to some 20° listening angle in a typical listening triangle - enough to be well outside the null.

The way the mid and high are angled, the listener is a good 30' off axis? am I correct to say that technically this thing doesn't make any sense?

It has been shown that equalizing the frequency response to be flat at 20-30° will give the best compromise between direct radiation and the summed radiated power into the room.
Did you notice that this way the first reflection of the mids willl be from the ceiling and not from the floor? This should almost double the time difference between direct mids and their first reflection, which is to be appreciated.

[Hank]

Somebody reverse engineer it!

Well, there just might be the two gurus of RE resident here :W

Rudolf, you make a couple of interesting points.

[Paul Ebert]

Actually, a very rough proportional measurement of the woofer baffle from the pictures puts it at about 25 inches, assuming the woofer is a 15 inch. That strikes me as about as wide as can usually be accommodated. This includes the acrylic part (I like that idea, by the way). Also, notice that the woofer is right at the floor. REG mentions this in his review (or was it the Gradient Revolution review?) as a way of dealing with the floor diffraction (and helping to compensate for dipole loss?).

The 45 degree orientation clarifies the sidefiring dipole issue for me. I wonder if they recommend the listening position to be closer to or farther away from the point where the on-axis lines meet to avoid the null. I suppose either would work.

I'm most curious about the 'hunter fan' midrange mounting (it does have more than a passing resemblance). The mid and tweeter are supposed to have cardioid response. Does the opening on the side of the mid baffle create this for the mid? Similarly, does the waveguide for the tweeter create it for the tweeter? This might be tricky to reverse engineer.

What does a cardioid response buy you, anyway? Aren't regular boxes cardioid?

Does the angle of the mid create a combination of cardioid and omni response? If so, what are the implications of that from the listening position?

The more I think about this design the more inclined I am to perceive it as very clever. Perhaps even ingenious, if the reviews are correct.

[DS-21]

I'd be shocked if it didn't sound great. The Gradient Revolution is one of very, very few off-the-peg speakers that I've heard in the last decade that sounded good to me.

And I think they look good, too. They'd fit well into a room with a modernist aesthetic.

[its_bacon12]

I like them. Very different style to the flat baffle look. I wonder whether or not that piece of glass on the back of the speaker interferes with the woofer's response curve.

[Saurav]

If you look at the edge of the glass, you'll notice that it somewhat mirrors the front edge of the wooden baffle. Thus, with the glass, the speaker has a more symmetrical baffle, with an offset woofer that's moved towards the front.

I'd guess the decision to offset the woofer wasn't for aesthetics either. I know an offset woofer will smooth response on-axis, but I also remember reading something about it rotating the main lobe in one direction or another. I wouldn't be surprised if moving the woofer towards the front makes the 'on axis' lobe turn towards the front as well, which would also reduce the effect of sitting close to the dipole null.

[Dennis H]

I don't know if they are designed that way but that in-room shot has them pretty close to the side walls so you'd get pretty good bass reinforcement with the way the woofers are aimed. Aiming the null into the corner should help a bit with room modes as well.

[Paul Ebert]

I imagine that the glass (I bet it's acrylic - if not, that's what I would use if I cloned them) actually helps the response since it widens the baffle, thereby lowering the dipole cancellation cut-off frequency.

[Rudolf]

I know an offset woofer will smooth response on-axis, but I also remember reading something about it rotating the main lobe in one direction or another. I wouldn't be surprised if moving the woofer towards the front makes the 'on axis' lobe turn towards the front as well, which would also reduce the effect of sitting close to the dipole null.

Unfortunately the "forward" position of the woofer should widen the dipole null at the "short distance" (front) baffle edge and narrow it at the back side. But the linearity of the frequency response is better along the short edge.

I'm most curious about the 'hunter fan' midrange mounting (it does have more than a passing resemblance). The mid and tweeter are supposed to have cardioid response. Does the opening on the side of the mid baffle create this for the mid? Similarly, does the waveguide for the tweeter create it for the tweeter? This might be tricky to reverse engineer.

I can´t find the discussion on some scandinavian forum any longer, where these aspects had been discussed. What I remember: The midrange is cardioid and the tweeter should be constant directivity (CD). So the main goal of this construction is to have CD radiation throughout the complete frequency range. The "off-axis" orientation of all drivers is meant to raise the indirect fraction of the total radiated energy and make the reflections of this fraction reach the ears as late as possible.

What does a cardioid response buy you, anyway? Aren't regular boxes cardioid?

Taking the expressions "dipolar", "omnipolar" and "cardioid" seriously - which none of us really does - they define radiaton patterns in the first place. And those radiation patterns don´t change with frequency.
So the radiation pattern of a regular box resembles a "cardioid" at one point in its frequency range, but below that point it becames omnipolar and above that point it tends to beam ever more.

Does the angle of the mid create a combination of cardioid and omni response? If so, what are the implications of that from the listening position?

If for you "omni" implies "half space", then it is almost the same as cardioid. By tilting up the angle you control the ratio of sound radiated directly to you WRT to the fraction radiated to the back of the speaker.

[Paul Ebert]

The midrange is cardioid and the tweeter should be constant directivity (CD). So the main goal of this construction is to have CD radiation throughout the complete frequency range. The "off-axis" orientation of all drivers is meant to raise the indirect fraction of the total radiated energy and make the reflections of this fraction reach the ears as late as possible.
Taking the expressions "dipolar", "omnipolar" and "cardioid" seriously - which none of us really does - they define radiaton patterns in the first place. And those radiation patterns don´t change with frequency.
So the radiation pattern of a regular box resembles a "cardioid" at one point in its frequency range, but below that point it becames omnipolar and above that point it tends to beam ever more.

If for you "omni" implies "half space", then it is almost the same as cardioid. By tilting up the angle you control the ratio of sound radiated directly to you WRT to the fraction radiated to the back of the speaker.

This makes a lot of sense. Thanks! So, is the purpose of tilting the midrange to achieve CD for the midrange or to get as close as possible to CD for the midrange? And, does the fact that some of the midrange backwave is emitted out of the side of the midrange 'baffle' help it achieve CD (through destructive interference) as well? I'm not sure I understand how cardioid can become CD.

If this is what is going on with this design (CD across the mid and highs), I'm even more impressed. I would think it would take a lot of experimentation to come up with the right mix of mid angling, backwave control and tweeter waveguide design to achieve such a thing. Or, is it somehow easier than I think?

[Paul Ebert]

Anyone care to conjecture on what drivers are used in this thing? Perhaps the mid is the Seas MCA15RCY. Previously I thought the woofer was a 15 incher, but now I'm thinking 12 inches (given that it is 3 ft. tall).

Wish I knew someone who has a pair.

[Rudolf]

Anyone care to conjecture on what drivers are used in this thing? Perhaps the mid is the Seas MCA15RCY. Previously I thought the woofer was a 15 incher, but now I'm thinking 12 inches (given that it is 3 ft. tall).

There is absolutely no doubt that the woofer is (a derivative of?) a Peerless SLS 12 (or 15):

Image not available

For more info regarding positioning look at the manual: helsinki_15.pdf

Basis_of_stereo_Gradient_design.pdf

[Saurav]

Unfortunately the "forward" position of the woofer should widen the dipole null at the "short distance" (front) baffle edge and narrow it at the back side. But the linearity of the frequency response is better along the short edge.

Thanks for setting that straight. I couldn't remember which way it went.

[Paul Ebert]

There is absolutely no doubt that the woofer is (a derivative of?) a Peerless SLS 12 (or 15)...

I think you are right! Is there a SLS 15? I think it must be a 12. I can't imagine how they get reasonable bass out of it. Now, I've considered in the past that the SLS 12 would make a nice dipole woofer (OK Qts, very flat response well into the midrange), but I was envisioning it being rolled off at about 80 hz and used with a subwoofer.

If I cloned it, I think I'd want to go with a woofer with more air moving capability.

[Saurav]

This may be a point of reference... I'm using the AE OB12 in my dipoles... going from memory here, but I think it has similar Qts and ~twice the Xmax, and it definitely needs a subwoofer. Though I only took my shelving filter down to 80Hz, I could have tried to push it lower, I probably have the Xmax to spare. The SLS12 was next on my list of options.

[Rudolf]

This may be a point of reference... I'm using the AE OB12 in my dipoles... going from memory here, but I think it has similar Qts and ~twice the Xmax, and it definitely needs a subwoofer. Though I only took my shelving filter down to 80Hz, I could have tried to push it lower, I probably have the Xmax to spare. The SLS12 was next on my list of options.

Comments here suggest that the SLS will fall off below 50 Hz in the Helsinki and a subwoofer would be needed for the lowest octave.

Some more sonic impressions

[CraigJ]

Has anyone done a sketch of the Helsinki yet? Is there enough "info" out to clone.....I have six sheets of .25" hdf, a RS390, RS28a, waveguide, dcx for testing, and a small budget for a mid woofer. Might be a fun build, if I can find the Hunter fan parts.

Craig

[Saurav]

Someone on DIYAudio is running some experiments that reminded me of this thread:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=142691

See the polar plots in post 8, and the photo in post 10 showing the setup that achieved those plots.

[fjhuerta]

They look somewhat like a Gallo speaker.

[brianpowers27]

See the following article regarding the angle of the dipole woofer. If I remember correctly sitting directly on axis for dipole bass is not optimal a slight off angle provides more uniform response.

http://www.musicanddesign.com/PowerMatching.html

[MuaDibb]

See the following article regarding the angle of the dipole woofer. If I remember correctly sitting directly on axis for dipole bass is not optimal a slight off angle provides more uniform response.
http://www.musicanddesign.com/PowerMatching.html

Good catch Brian, I read that the other day researching dipoles. Referring to Fig. 4 in the article, it seems like what they are trying to achieve. It was then that I realised my current project somewhat resembles the Gradient, at least as far as the mid goes.





I wonder what they gained by sandwiching the baffle like that? I'm also sketchy as to how one achieves a cardioid response, is it accomplished by baffle orientation, x-over, a combination of things? I would welcome some comments. Tom

[Paul Ebert]

I would guess they get the cardioid response by having the backwave come out the sides.

[brianpowers27]

I think I saw one of AJINFLA's self described Cardoid speakers exhibit the same sandwiching technique.

[AJINFLA]

Nah, I've never shown one of those, you probably confused this with me eating a sandwich while describing one of my speakers.
The cardioid response is from the backwave being delayed in time with respect to the front, in this case by using a flow resistance material behind the driver. Cardioids are described in great detail on both SL's and John K's site.
Both those sites contain a treasure trove of information regarding speaker design, period (they are not just about dipoles).
They should be a must read for all speaker builders, but unfortunately, seemingly rarely are.

cheers,

AJ

[brianpowers27]

Nah, I've never shown one of those, you probably confused this with me eating a sandwich while describing one of my speakers.
The cardioid response is from the backwave being delayed in time with respect to the front, in this case by using a flow resistance material behind the driver. Cardioids are described in great detail on both SL's and John K's site.
Both those sites contain a treasure trove of information regarding speaker design, period (they are not just about dipoles).
They should be a must read for all speaker builders, but unfortunately, seemingly rarely are.

cheers,

AJ

Indeed, these are useful sites. What would one do if they had read these sites, yet are seeking more knowledge? Perhaps there is a knowledge gap or just a general lack of coffee clarity...