Do you believe?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Some have expressed interest, but to be honest I just want to show off a bit. Further investigation into why the Dynaplane ribbon is capable of lower distortion than typical free swinging ribbons of same size has resulted in a refinement of the design.

Kinda excited about this....

The following distortion tests were all done in same setup within about an hour of each other. All with crossover around 1 Khz and mic at 1 meter, all at 100 db at around the 2Khz reference region.

First chart (for comparison purpose) is the Fountek Neo 2 . Typical ribbon showing 3rd harmonic on top and hovering around a rather high 3%

2nd chart shows Dynaplane ribbon of similar size to the Fountek yet showing 2nd harmonic now on top and at 100 dB its only about 0.7 %

3rd chart shows same Dynaplane ribbon but with wave guide. Again 2nd harmonic on top and now averaging below .5%. An order of magnitude less than is typical for other free swinging ribbons this size.

To be fair the Fountek is not designed to be used to 1Khz and yes it did give up the gost after a few slow sine sweeps at 100 dB. It was used simply as an example of the typically higher distortions of similar ribbon designs.

While the importance of certain distortion profiles and levels will likely be debated, I am simply excited that after many years of work I have come to a design that both reduces small ribbon distortion by an order of magnitude, AND provides a means of excursion control allowing reliable operation to lower frequency's.
 

Attachments

  • neo 2 bout 3 %.jpg
    neo 2 bout 3 %.jpg
    104.3 KB · Views: 524
  • 15mm sides neo35 2 row .34mm cor .7% 100db.jpg
    15mm sides neo35 2 row .34mm cor .7% 100db.jpg
    98.9 KB · Views: 532
  • 15mm sides neo35 2 row .34mm cor .5-.2 horn 100db.jpg
    15mm sides neo35 2 row .34mm cor .5-.2 horn 100db.jpg
    101.7 KB · Views: 526
Last edited:
Looks nice !!!,Wonder how it looks :)

But i am not so sure about neos being 3% typically ? cant imagine.

maybe at 100dB ? the datasheet states a 3 order crossover at 2.5khz so yeah it does not like 1khz ... puting 20dB+ to many in there at that freq when you calculate that back to power... hmm poor ribbon

i was busy with some planar tweeters (dont mind the dip that is gone be my next post since i cant find it :))

but that already plays from 1khz to what i can measure distortion/frequency under 0.24 % total harmonic, while the 2e is under 0.24% and third orde under 0.06% cant imagine the ribbon would do 3%
 

Attachments

  • distortion push pull.jpg
    distortion push pull.jpg
    259 KB · Views: 421
Looks nice !!!,Wonder how it looks :)


I would love to talk about that BUT at the moment a patent is in the works . All I can say is that it deals with a fundamental issue with free swinging ribbons.


But i am not so sure about neos being 3% typically ? cant imagine.


You are right. The Neo2 is much less than 3% at lower volumes. This comparison was done to show what happens to that low distortion figure in typical "free swing ribbons" when the volume goes up, And to provide some sort of reference for the overall test. As well the 100db test still shows this 3% distortion level well past 3 khz so despite the perf below that being outside recommended use the stuff above 3 k is still useful

maybe at 100dB ? the datasheet states a 3 order crossover at 2.5khz so yeah it does not like 1khz ... puting 20dB+ to many in there at that freq when you calculate that back to power... hmm poor ribbon


Yes it was done at 100db.Almost any ribbon will show nice distortion at lower volume.


i was busy with some planar tweeters (dont mind the dip that is gone be my next post since i cant find it :)) but that already plays from 1khz to what i can measure distortion/frequency under 0.24 % total harmonic, while the 2e is under 0.24% and third orde under 0.06% cant imagine the ribbon would do 3%

Yes that's sound right. All planers I have measured show lower distortion than typical free swinging ribbons ( of similar size) when the volume goes up. The Planers do not suffer the same issue as a free swinging ribbon.

Its also important to notice that in the typical free swing ribbon as shown in the Neo 2 will have the 3rd harmonic ABOVE the 2nd harmonic at these higher volumes. With the Dynaplane ribbon that is switch around to what you typically see with planers and domes. You probably wont see other ribbon makers posting this fact

The problem I see with planers is to get a low enough resonant freq to play well to below 1khz in a SMALLER dimensioned diaphragm . The free swinging ribbon does this easily as it is not clamped on all sides. The Dynaplane ribbon in these tests is only 14 mm wide and 3 inches long, and has a smooth roll off with no resonant bump all the way down to 300 hz. Makes crossover at 1K almost too easy.


Now here's something that will seem contradictory to what I have posted. When it comes to distortion I am in the camp that says its NOT as important as a smooth freq response and a good impulse/ CSD, and good dispersion out past at least 10K. We have been doing blind tests here on this forever. Its really easy to make two ribbons that are basically identical but one with the mod that produces an order of magnitude less distortion and the results show NO favoritism between the two. AND they sound the same and up to about 90 db measure the same.

Sooo why bother?? Two reasons
1- unfortunately the masses Believe that these low distortion numbers are important, and 2- because the mod that produces these low numbers also makes a ribbon that has the needed stability in the magnet gap and can structurally handle the lower freq cross over . It is a design that just happened to kill two birds with one stone. That alone is rare!!



BTW WrineX, as far as the high distortions at 100 dB on the Neo2. You are quite good at makin these things. Its serious easy to make a typical ribbon. I would throw a small one together and test at higher levels. You will see ;)
 
Last edited:
BTW just a note on the title of the thred, "do you believe?"

I was working a bit with a guy, who will remain nameless, who is a writer in some audio publications. The plan was to have him test the original Dynaplane ribbon design , live with it for a while in different systems, write about it, and possably connect me with Parts Express for possible distribution.
However He kept advising me "not to lie". The remark was about my published distortion figures. Apparently He had past experience with a famous ribbon maker in past and he understood that free swing ribbons this size simply do not show these low numbers.

Long story short I have not perused the relationship. The guy at Impulse Audio seems like an straight shooter. I may see if He has interest in the future.
 
I never knew third could be above 2e. never ever ever seen that in any loudspeaker i made not even in ribbons, Rubanoides, Planars normal drivers,domes,EAPs,ESLS. so learn everyday :) probable will for the rest of my life:)

Yes a ribbon in theory is simple. although a perfect one is not i guess, i think Gerrit made a close one on that regard.
Yeah going in business straight of is scary :) and aparantly they had a bad experience :(
The dynaplane how does it look and work, or is that still in the trade secret (i would understand ) about the resonance, i am using this at the moment. not perfect yet in terms of i have to figure something about terminating the ends so it wont create resonances. but this small panel has a main res of 100 or something. (with huge distortion of course @ that freq)

this is still a non push pull system. will ad the backplate today see what happens :) only thing i am annoyed by is the dip at 12khz but the distortion looks on certain points between the clearly visible resonances even lower then the planar i posten an hour ago. (above)

i also tried horn loading it. no expert in horns at all but it does make te response reach lower even with the backside being open.
 

Attachments

  • ribbonlike.jpg
    ribbonlike.jpg
    699.8 KB · Views: 163
Last edited:
Yes a large area planer will move so little that its distortion will be quite low, even non push pull.

As for 3rd harm above 2nd, I see this on every single FREE SWING ribbon I have ever tested at higher volume and I haven't yet seen published distortion at higher volume from other manufactures. I wonder why? ;). As far as I know Its fundamental issue of all tru ribbons I know of except the one I developed.

The original Dynaplane ribbon was a larger ribbon ( 20mm wide by 150mm long) that was good to 500 hz. I was awarded a patent on that one.
Heres the patent number if interested...

US 9668057 B1

You can see it at dynaplane ribbon loudspeakers

For various reasons I am discontinuing it but will continue to supply parts to those who need.

That one has a unique construction that gives more control over over excusion AND gives a higher spring rate along length without added mass allowing use to 500 hz BUT the reason for the low distortion in that design is hidden and at the time of the patent I missed it. The new version levers off from understandings learned from it and is a better ribbon, BUT I cant talk directly about the secret sauce Ha.
 
Last edited:
…notice that in the typical free swing ribbon as shown in the Neo 2 will have the 3rd harmonic ABOVE the 2nd harmonic at these higher volumes. With the Dynaplane ribbon that is switch around to what you typically see with planers and domes.
Very impressive! Thanks for sharing your progress.
Hopefully the patent application will be filed soon so you can talk a bit more freely about what is going on.

- Do you normally see 3rd harmonic stay so constant with frequency like shown in Post#1? The few ribbon I have measure generally had it falling off fairly quickly with increasing frequency. I guess that might also have to do with the transformer, if one is used.
- Can you comment on how distortion for your new and old construction techniques compare between monopole and dipole operation?
- Can you share any details on the waveguide profile used in the Post#1 measurements?

***
Just noticed your waveguide related thread in the Multi-Way Forum:
ribbons and wave guides
 
Last edited:
Very impressive! Thanks for sharing your progress.
Hopefully the patent application will be filed soon so you can talk a bit more freely about what is going on.


Hi Bolserst,
well unlike the first patent this one has been a serious struggle. The problem is every wording we come up with to describe the construction could potentially be describing the constructions of some stuff that's been done in the past. Not that what's been done in past is exactly the same and certainly not for same reason, BUT the darned wording needed to "exclude others" always seems to be words that could at least in part describe something that's been done. If that makes any sense Ha.

At the moment I am on the fence about it and if I don't find better wording to exclude, combined with the fact that these will likely never be a mass market thing, I may simply build and sell. My hope is to release by spring at the latest.


- Do you normally see 3rd harmonic stay so constant with frequency like shown in Post#1? The few ribbon I have measure generally had it falling off fairly quickly with increasing frequency. I guess that might also have to do with the transformer, if one is used.


No, usually as you noted it starts taking off below about 2khz. The more it moves , the worse it gets. Much of my work has been to combat this to gain respectability in taking a small ribbon lower. Its not in the transformer, its in the diaphragm

- Can you comment on how distortion for your new and old construction techniques compare between monopole and dipole operation?

It measures basically the same mono or dipole, BUT I used a large baffle rather than strong correction

- Can you share any details on the waveguide profile used in the Post#1 measurements?


Ha You would laugh out loud at my "waveguides". No math, no knowledge, just hot gluing up different ones out of thin wood changing angles and size till I get a nice response. The WG used in these test are about 100 deg horiz and 70 deg vert with a mouth opening of 12 inches horiz and 9 inches vert, and depth about 4 inches. In another WG I get a bit smoother response and it goes lower BUT is quite a bit larger and again just straight sides


In all honesty I really don't know what Im doing in the wave guide dept. BUT damn, I get a nice flat response with these crude wave guides And about a 5 db increase in sens. The darned guide totally evens out the response. No trap or any filter needed!

I guess a good set of polars will be necessary but that will have to wait for now.

***
Just noticed your waveguide related thread in the Multi-Way Forum:
ribbons and wave guides
 
Last edited:
I never knew third could be above 2e. never ever ever seen that in any loudspeaker i made not even in ribbons, Rubanoides, Planars normal drivers,domes,EAPs,ESLS. so learn everyday :) probable will for the rest of my life:)

Yes a ribbon in theory is simple. although a perfect one is not i guess, i think Gerrit made a close one on that regard.
Yeah going in business straight of is scary :) and aparantly they had a bad experience :(
The dynaplane how does it look and work, or is that still in the trade secret (i would understand ) about the resonance, i am using this at the moment. not perfect yet in terms of i have to figure something about terminating the ends so it wont create resonances. but this small panel has a main res of 100 or something. (with huge distortion of course @ that freq)

this is still a non push pull system. will ad the backplate today see what happens :) only thing i am annoyed by is the dip at 12khz but the distortion looks on certain points between the clearly visible resonances even lower then the planar i posten an hour ago. (above)

i also tried horn loading it. no expert in horns at all but it does make te response reach lower even with the backside being open.

Interesting construction. Yea that same size planer would have a higher and more difficult main resonance if it were a simple flat diaphragm and clamped at all sides.

The 12k wiggle I would say is likely the size of your corrugations. If you look at the CSD that wiggle will likely show a strong "tail" at 12k. Be carful here as moving the mic can make it magically disappear ha


A way to tell if its the corrugation size is to run the FR sweep while pulling on the diaphragm so that the corrugations change size. My guess is if you pull it a bit flatter you will see the wiggle move down in freq.
 
Last edited:
Interesting construction. Yea that same size planer would have a higher and more difficult main resonance if it were a simple flat diaphragm and clamped at all sides.

The 12k wiggle I would say is likely the size of your corrugations. If you look at the CSD that wiggle will likely show a strong "tail" at 12k. Be carful here as moving the mic can make it magically disappear ha


A way to tell if its the corrugation size is to run the FR sweep while pulling on the diaphragm so that the corrugations change size. My guess is if you pull it a bit flatter you will see the wiggle move down in freq.

the wiggle is also in the planars (flat ones) so i guess magnet structure :( but when used single ended in my tiny horn (a regular planar) it looks like this.

the crossover is till active i see at 800. but since i should get resonance a little lower since it is at 500-600 right now. thats why it drops like a brick under that haha
 

Attachments

  • in minihorn to 650 hz.jpg
    in minihorn to 650 hz.jpg
    148 KB · Views: 208
Last edited:
Ah I see, interesting. Well that looks like a quite usable response anyway.

I built a large curved full range planer once years ago. It was a bunch of 3 inch wide segments about 40 inches long, and it was single ended. If I remember correctly I ended up spreading magnets wayyy out so that the sensitivity was crap but the response was quite good with no big wiggles
 
More data...

1st graph is ribbon freq response with trap, NO wave guide, and NO smoothing
2nd is a close mic ( approx 1/2 inch) CSD direct drive , no trap, no Wave guide
3rd is impulse close mic, no wave guide
4th is impulse, at 1/2 meter, No wave guide
 

Attachments

  • 15mm rib FR w trap.jpg
    15mm rib FR w trap.jpg
    95.9 KB · Views: 141
  • 15mm ribCSD.jpg
    15mm ribCSD.jpg
    237.4 KB · Views: 146
  • 15mm rib pulse close mic.jpg
    15mm rib pulse close mic.jpg
    29.2 KB · Views: 139
  • 15mm rib pulse .5 meter.jpg
    15mm rib pulse .5 meter.jpg
    29.8 KB · Views: 142
yeah the less they are in teh way the better :) obstrutions are not your friend :) reason why i dont use a back metal pole piece at the moment to achieve rather Ok response at the cost of efficiency

Looking back through old notes I see I tried a "shaded array" once years ago.
Basically made a planer about 5 inches wide X 40 inches long, BUT both the magnet row spacing AND the distance from magnets to diaphragm was manipulated.

the magnet rows in center were close together and close to diaphragm. Then as you go outward from there the rows each get farther apart and farther from the diaphragm surface.

I did not see the resonant peak from slots. I assume it was simply spread out enough to hide it better.

The idea was to get better dispersion from a wider planer. In the end it really didn't give enough increase in directivity to make it worthwhile AND sensitivity took a serious hit.

Also noticed that dynamics were not as good. However the mid range was to die for! It was easily noticeable. It was soooo smooth. Human voice was magic.
It appears that when you vary the drive force across the diaphragm you get usefully less resonant behavior.

It might be possible that if you only varied the magnet row spacing and not the distance of mags from diaphragm, you may get the advantages of a spread out slot resonance and keep more of the sensitivity??

Just a guess.
 
Looking back through old notes I see I tried a "shaded array" once years ago.
Basically made a planer about 5 inches wide X 40 inches long, BUT both the magnet row spacing AND the distance from magnets to diaphragm was manipulated.

the magnet rows in center were close together and close to diaphragm. Then as you go outward from there the rows each get farther apart and farther from the diaphragm surface.

I did not see the resonant peak from slots. I assume it was simply spread out enough to hide it better.

The idea was to get better dispersion from a wider planer. In the end it really didn't give enough increase in directivity to make it worthwhile AND sensitivity took a serious hit.

Also noticed that dynamics were not as good. However the mid range was to die for! It was easily noticeable. It was soooo smooth. Human voice was magic.
It appears that when you vary the drive force across the diaphragm you get usefully less resonant behavior.

It might be possible that if you only varied the magnet row spacing and not the distance of mags from diaphragm, you may get the advantages of a spread out slot resonance and keep more of the sensitivity??

Just a guess.


well i actually thought about the exact same thing. but to prove it what the distance between magnets did i made 2 with different spacings but the dip did not move much even if the ratio would be rather big from 5 mm gap to 3 mm.

i would expect it to move down if it acted like a resonator
 
As an ESL partisan, can't help but be very impressed by this data.

With ESLs, we have a clean amp and a clean driver. But in-between is a transformer that can't work right without some circuitry at the primary and even then, the output will have odd bumps and dips and the amp will never be completely happy driving the load. So maybe the better routes are direct drive ESLs or great ribbons like in this thread... if only they could reach down to 300 Hz like a Martin Logan or lower.

B.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.