New Scan-Speak D3004/660000 Distortion Tests

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

Same method, 125mm etc. But instead of changing levels, we change frequency - 100dBSPL was chosen.

First 1KHz- the lower the frequency, the greater the demand on the tweeter.

D30 :

D30_IKHz_100dB_1.325.gif


That looks like high THD - 1.325% - but it is dominated by the 2nd.

HDS:

HDS_IKHz_100dB_0.340.gif


Yes, the THD is lower, but look closely. In D30 the 2nd is greater than 3rd whereas HDS 3rd is greater than 2nd. This makes the Total Harmonic Distortion figures misleading.

Let's go onto 2KHz.

D30:

D30_2KHz_100dB_0.619.gif


Still dominated by 2nd, but 3rd is just noticeable and higher suppressed.

HDS:

HDS_2KHz_100dB_0.401.gif


Lower 2nd but higher 3rd.

Now 4KHz.

D30:

D30_4KHz_100dB_0.382.gif


Nicely cascading - a good sign a la Jean Hiraga.

HDS:

HDS_4KHz_100dB_0.306.gif


Pretty much a draw here.

OK, where does this get us? Note that in every set of plot HDS measures lower? Yes, THD is misleading, only by breaking the figure into individual harmonic components can we see the real picture.

The winner. Curiously that every THD favours HDS, yet as above D30 is a narrow winner.

What do you think?

Joe R.

Next: IMD Tri-Tone Tests.
 
IMD Tri-Tone Tests.

This is the last 'chapter.'

The Tri-Tones will be centered around 1KHz, 2KHz and 4KHz. Two out of three adjusted to 100dBSPL @ 125mm.

1KHz

D30:

D30_1KHz_3T_0.956.gif


HDS:

HDS_1KHz_3T_0.365.gif


2KHz

D30:


D30_2KHz_3T_0.659.gif


HDS:

HDS_2KHz_3T_0.390.gif


4KHz

D30:

D30_4KHz_3T_0.367.gif


HDS:

HDS_4KHz_3T_0.350.gif


The winner? This time you tell me. The calculated IMD figure overwhelmingly favours HDS - what do you think?

So there you have it. Spend some time analysing the graphs, maybe you noticed something I didn't? I have tried to keep running comments to a minimum, only guiding where I feel appropriate. Where I have discussed a so-called 'winner' it is only to point that summed distortion figures - and hence into a single figure - can be misleading. This alone should be discussed.

Other factors: If you are assembling a 4 Ohm speaker system and don't mind paying the extra money, then you should consider the D30. But if you are aiming at a sensitive speaker system and 8 Ohm, then HDS is for you. If cost is a consideration (Madisound lists a 3.7:1 price ratio), then the HDS is made for you!

The final thing that stands out is in relation to D30: The shelved 2nd order distortion. There seems to have been made a conscious decision to trade off greater 2nd harmonic distortion for a lower level 3rd. This alone strikes me as the single overall most important conclusion regarding its behaviour and maybe why these two samples here measure differently than earlier 2006 samples (these are definitely 2007).

Over to you guys!

Joe R.
 
Joe Rasmussen said:
The winner? This time you tell me. The calculated IMD figure overwhelmingly favours HDS - what do you think?

Thanks for the tests, Joe. It's good to see this level of thoroughness.

The calculated IMD% number, like THD, is heavily weighted by 2nd order products. A good way to compare these is to just drop one on top of the other to compare the whole spectrum. We'll call that Mark K style, since he's the only guy I see doing that. Praxis has it built in, but if I want to do it in SE, I have to use an image editing program.

Your results are actually very close to both mine and Mark's, done on 2006 tweeters. The SPL of the 6600 seems to be the same, and the only real difference seems to be minor variations in the top octave.

Unfortunately, there's minor variations in the top octave for all the 6600's from 2006. I've tested 4, my own pair and another pair, and just about every one had a mildly different top octave. Peak at 20kHz (2), dip at 12kHz with a plateau above (1), and perfectly flat out to 35kHz (1). I don't put a lot of value in the top octave response, as long as it's somewhat there. The important part of the response curve starts at the low end, IMHO.

Thanks,
John
 
Zaph said:

Thanks for the tests, Joe. It's good to see this level of thoroughness.


I'll take that as praise. :)


The calculated IMD% number, like THD, is heavily weighted by 2nd order products. A good way to compare these is to just drop one on top of the other to compare the whole spectrum. We'll call that Mark K style, since he's the only guy I see doing that. Praxis has it built in, but if I want to do it in SE, I have to use an image editing program.


I have SE V13, but I don't, have not, used the data acquisition part of it (I've had various versions down to V1 - that virtually made me a beta-tester). I supposed you have pestered Bohdan about it. He only lives down the road from me, the road is called the Hume Highway. OK, it's 1000K but it's still down the road. :)

But is this what you are looking for:

D30_100dB_THD_Graph.gif


The maroon line tracks the Total Harmonic Distortion, but is totally dominated by red, the 2nd harmonic. This proves your point.

You are saying that Praxis does this and SE not? This is ClioWin 7.03 (very latest) and is a dodle.


Your results are actually very close to both mine and Mark's, done on 2006 tweeters. The SPL of the 6600 seems to be the same, and the only real difference seems to be minor variations in the top octave.

I don't put a lot of value in the top octave response, as long as it's somewhat there...


In these two samples, the top averages out flat in the top 10-20KHz octave off axis. This is a plus with me and an advantage over both XT25 and HDS. Yet to hear D30 but I suspect it will have some 'sparkle' and I mean that in a nice sense. But I am very familiar with the sound of HDS - it's a terrific performer. What is of concern is D30's sample variations from a manufacturing point of view and as the company I am associated with is considering using this in a design.



The important part of the response curve starts at the low end, MHO.



I didn't quite nail that in my FR plots, my baffle not large enough. But the depression is common in all three FRs and you can see past it on that basis. My baffle needs to be closer to IEC size.

Re bandwidth: The HDS is further down @ 1KHz but can be pushed hard to bring it up to matching measurement level of D30. To match this level is emphasizing 3rd a bit in some of the plots? As I have used the HDS in a system, I have used 1st order down to 1500Hz approx (from 3KHz) but engineered a response 4-5 order below that. It's about a controlled low end of the bandwidth to get lowest distortion. This may well help cancel out some of the higher 3rd, compared to D30?

John, did you ever look at a tweeter response when not flush mounted. YIKES!

Joe R.
 
Member
Joined 2005
Paid Member
Hi,

For Windoze users out, the quick way I like to do overlays, is in MS Paint.

Start -> Accessories -> Paint

Paste your first picture in. (use Paste or Edit->Paste From...)

On the left panel, click on the Image Overlay icon. (see attached image #1)

Then paste your 2nd image in.
(see attached image #2- 1.3KHz comparison-HDS blue, D30 red)

Then quickly tap <Ctrl>-<Z> (Undo) <Ctrl>-<Y> (Redo) to get a feel for the differences.

And, yes, I'm sure there's an easier way to do it in OS-X, but meh.
 

Attachments

  • overlay icon.png
    overlay icon.png
    50.9 KB · Views: 646
Yup, that's a similar thing that I have to do with SE to show one IMD on top of another like this. Otherwise I can only show one driver at a time. Praxis does a lot of things that SE can't but for the price I can get by with SE. Hell, I could probably get by with the Arta demo and SW for $0.

I've got a pair of 2007 6600's on the way soon to compare to the 2006 version, but based on Joe's results I'm not expecting major differences.
 
To throw in some more measurements

I've got the 6600 from Mark for testing. I've done all of the measurements and tried a small tweak, of limited to no use, but haven't put up a page yet.

raw_6600_spl.gif


One of the benefits of the 6600 is supposedly to have the increased output down low (which it does) while having the dispersion of a 3/4" tweeter (which it doesn't). I finally set up and measured a bunch of drivers at 90 degrees recently. The 6600 is 3db lower than the OW1 at 7K and continues to lose ground above that until about 12K, at which point it does exceed the OW1. It's too late at that point in my book.

Curiously, at 90 degrees the RS28 is about an even match in absolute terms. My plan is to set up a bunch of graphs to show the 90 degree response normalized to the drivers own on-axis response as something of a dispersion indicator.
 
Waveguide

Hi Guys

Just as an afterthought, come back to the earlier posted posted graph that shows the family of OFF axis of Scan-Speak D30 (RED), new Peerless HDS (BLUE) and Vifa XT25 (GREEN):

OFF_Family.gif


It's the XT25 that is of interest here. Now note below the same tweeter that I have measured with a square waveguide used in my Elsinore Project speakers:

Tweeter_Off.gif


Both are calibrated dBSPL measurements and 25 degrees OFF axis.

Ignore the rise below 3KHz (that is caused by the waveguide). It's higher up that interests me. Compare the XT25 being -20dB at 20KHz and the results when using a waveguide. The OFF axis response does not seem to drop off anywhere near as when it is flush mounted to a largish flat rectangular baffle.

Any comments on this? Has anyone else observed something similar? I suppose, what I am asking, does a waveguide bolster the OFF axis, in this case 25 degrees (the wave guide is much shallower than that).

There was a review of the Lipinski L707 speakers in December 2005 Stereophile that also used a waveguide (staggered foam) with XT25 tweeter. The technical side of the review showed surprise that the off axis was that good since it was using THAT tweeter. "Unusually for a ring radiator tweeter... doesn't get as directional at ultra-sonic frequencies as other designs I have measured."

Any thoughts?

Joe R.
 
I appreciate all these measurements very much... and may be you could help me, since I am not an expert on designing drivers.

Comparing the on and off axis of a number of "recent tweeters" (like 6600 and HDS) I somehow get the suspicion, that the "better 3/4''-like dispersion" (which is basically non-sense to me) is achieved by designing a resonance somewhere in the 15kHz region (you can suspect that also in zaph's CSD for both 6600 and HDS I think).

Would you agree? Wouldn't that be counter-productive in the end for designing a neutral system?

I know that the relevance of the top octave is debatable. However, from my own experience in recording orchestras and from an extensive small diaphragm cardioid microphone test, I came to the conclusion, that resonances or unlinearities in the top octave can be quite crucial for the last bit of "realism" in orchestra recordings, I would say more crucial than harmonic distortion.

Thanks for answers.

Regards,
Leif
 
leifislive said:

Comparing the on and off axis of a number of "recent tweeters" (like 6600 and HDS) I somehow get the suspicion, that the "better 3/4''-like dispersion" (which is basically non-sense to me) is achieved by designing a resonance somewhere in the 15kHz region (you can suspect that also in zaph's CSD for both 6600 and HDS I think).

Would you agree? Wouldn't that be counter-productive in the end for designing a neutral system?

Leif

I can say with some conviction that the better Tweeters around right now, and the HDS is that at a lower price than the others, they DO sound like they have lower distortion AND they measure that way.

I am inclined to believe that it is the lower end (bandwidth wise) of these Tweeters, that is the key to lower distortion overall, but not exclusively. The better Tweeters have good low end bandwidth with lower distortion.

I like the idea of the Vifa (now V-Line) XT25 being a ring radiator. The problem lies that in the top octave even a dome tweeter has to behave like a ring radiator, the centre of the dome should be stationary and the area nearest the voice coil should radiate. It's not hard to visualise that this is then a 'ring' radiator. But will the centre of the dome now behave properly? Especially as it does not have a natural point of termination? I remember the early 19mm STC Coles hard dome tweeters used in BBC monitors, some would cut out the centre of the dome.

But this 'stationary' trick that allows a dome to behave like a ring radiator, the better materials used in the best Tweeters, well, the audible results speak for themselves.

But coming back to the XT25, being a true ring radiator it ought to have an advantage over the HDS in the top octave. But it does NOT sound wise. But XT25 has rising distortion below2-3KHz where the HDS is much better behaved. Don't forget that 2KHz 5th order harmonic distortion turns up at 10KHz, 3rd order 5KHz is 15KHz, etc. I think it can be linked to the fact that Tweeters do not like amplitude (the XT25 is worse than HDS), they don't like to behave like pistons a la bass drivers. Use the crossover to limit amplitude at its Fs using LCR Traps or Null Filters. The distortion throughout the Tweeter's range is then audibly reduced right up to the top octave.

Joe R.

PS: Re Condenser mics, I am no expert but my understanding is that some resonance is built in to get a slice of the top octave and then they fall off. This explains why ribbon mics sounds less toppy and yet have far more extended response. They are just cleaner. Same goes for the better Tweeter's. They sound kinda less 'trebly' - if you get my drift.
 
Thanks for your reply.

Do I understand your answer correctly, that the unlinearities off axis result from interferences due to the "pseudo-ring-radiator" like behavior of the diaphragm rather then resonances

Have you listened to the different tweeters within the direct field (is this term existing in English language?)?

Have you also done pink noise listening tests?
 
I am not aware of any ribbon that can beat a DPA 4006, but that is comparing figure 8 to omni... ;)

To make it short, I have not completely understood how a mic is build (I am not an electro engineer and I do not have the time to read a lot in this field presently...).
I think the problems vary among the different types of capsules (pressure vs. gradient transducers, small to large diaphragm) and are specific for each. Actually, resonances are a part of building a condenser or a dynamic mic, as you stated, I remember that from an article from one of the Schoeps engineers. But it depends on the polar pattern for each capsule, if I remember correctly, where the resonances in the spectrum are placed.

I have only evaluated cardioids so far in detail (searching for perfect cheap spot mics - something that does not exist... :( ) and developed the hypothesis, that resonances may show up in cardioid mics also as an unlinearity in the polar pattern (especially at 180°). But I have to deal with that in detail again. Maybe someone more experienced can comment on this.
I do not exactly know about omni or figre 8 mics; those mics might have other problems based only on the specific polar pattern. So, ribbons could potentially work around these problems a little, since they are only working in figure 8, according to my knowledge.

I do not know, whether comparing a SD condenser to a tweeter is valid at all (I always thought mics may "shut-off" the outer membran part rather than the inner as you stated for tweeters).

I welcome any comment to clarify things :)
 
leifislive said:
Thanks for your reply.

Do I understand your answer correctly, that the unlinearities off axis result from interferences due to the "pseudo-ring-radiator" like behavior of the diaphragm rather then resonances

Have you listened to the different tweeters within the direct field (is this term existing in English language?)?

Have you also done pink noise listening tests?

I think I understand what you mean by 'direct field' is ON axis?

No, I don't think there is anything deliberate going on. The off axis is going to suffer from beaming effects unless we reduce the radiating area. The same applies to cone drivers. The radiating area reduces (at least in theory) above the 'knee' which is determined by the diameter. For example, a 100mm diameter the knee will be around 2KHz. If the cone is infinitely stiff the response would fall 6dB/Octave. But because the cone flexes, the response can be maintained above the knee, but the off axis will not be as successfully extended as the on axis. Dome Tweeters also reduce radiating area towards the voice coil to achieve something similar. Except the radiating area now become 'ring' like and the centre of the dome is now not terminated and thus its behaviour becomes less predictable.

The knee of a dome Tweeter is around 8KHz. The material which the dome is made of, needs inherent damping (some add doping) to control the centre of the dome. I don't think that any manufacturer enhances off axis response by using above 8KHz resonances. I can't see it work like that.

I do believe that Tweeters should have some reasonable off axis response but it is expected that it will not match the on axis. The off axis does help give a better sense of 'air.'

But I reiterate, I think distortion well below 10KHz can show up in the top octave as harmonic distortion - creating sizzle and excess energy. The best Tweeters don't sound like Tweeters.

Joe R.
 
Thanks for your answer. Direct-field was meant to be within the reverb radius (I googled the term and think is the translation for hall-radius). It is, where the reverb of the room has less energy than the direct beam of the speaker, I hope this is more understandable...

May be lower non-linear distortion numbers point towards a more accurate motor design of the driver; therefore when playing multiple frequencies, IMD redue dramatically. Maybe one could see clearer differences doing extensive (>5 spread over an octave?) multitone testing? I don't know ...

If I sum up would you have said, the Scanspeak ringradiator should theoretically fit your needs best, but in reality, you don't like it, true?

Best,
Leif
 
One other consideration

The off-axis for a dome tweeter is better than that of the ring because of two aspects. The dome occludes increasing large areas the more the off-axis angle increases. Prior to the reduction in radiation due to damping at higher frequencies the dome is radiating in the off-axis with reduced phase deltas due to the blocking of the occluded areas. The non-occluded areas of a dome have smaller phase deltas as well than a ring because the ring of the XT25 has very little occlusion at all.

Add to this the damping factor that a good dome includes and you have additional reduction of phase deltas in the off-axis, not just that of the on-axis. The ring is nearly ideal on-axis, but will always suffer at upper frequencies in the off-axis vs. a good dome.

Dave
 
Re: One other consideration

dlr said:
The non-occluded areas of a dome have smaller phase deltas as well than a ring because the ring of the XT25 has very little occlusion at all.
Dave

Here is a curious fix of sorts:

Due to the poor of axis response of XT25, you'd think that adding a waveguide (shallow horn) would make it, at least on paper, more directional - that is what horns usually do, right? Yet the exact opposite is the case. A waveguide actually builds up the off axis response, very noticeably. If you look back with an eagle eye, I have actually given measured examples of this on this thread.

Post #50 http://www.diyaudio.com/forums/showthread.php?postid=1172978#post1172978

The difference is huge. Kinda fascinating!

It all comes under that curious broad heading of 'diffraction effects.' Right?

Joe R.
 
Re: Re: One other consideration

Joe Rasmussen said:


Here is a curious fix of sorts:

Due to the poor of axis response of XT25, you'd think that adding a waveguide (shallow horn) would make it, at least on paper, more directional - that is what horns usually do, right? Yet the exact opposite is the case. A waveguide actually builds up the off axis response, very noticeably. If you look back with an eagle eye, I have actually given measured examples of this on this thread.

Post #50 http://www.diyaudio.com/forums/showthread.php?postid=1172978#post1172978

The difference is huge. Kinda fascinating!

It all comes under that curious broad heading of 'diffraction effects.' Right?

Joe R.
I saw that post, but didn't reply as I have only nominal experience with waveguides. The results are interesting, but I have little direct experience on which I can make direct comments.

A waveguide is supposed to increase sensitivity in the front hemi-sphere at the expense of the rear hemi-sphere (below step) and increase it in a cone (if a circular opening) defined by the waveguide geometry. As Geddes does it he's trying, as I understand it, to create constant directivity down to the point where diffraction diminishes the directivity and the dispersion starts to increase.

The source for that waveguide has an impact on the effectiveness (flat wavefront compression tweeter vs. non-flat of a dome or an XT25 for example). The XT looks to be a closer match than some domes from some of the posts I've followed. 25 degrees off-axis as in your measurement is likely, I would think, to be within the defined cone of most waveguides, so an increase in the off-axis at this angle seems quite reasonable. The dip around 750 must be some sort of diffraction issue internal to the horn (I''m guessing here), but the high end looks a lot better than I would have expected. Certainly, as you point out, most of the peaks/dips are diffraction from the internal regions of the waveguide.

The more interesting comparison to me would be 90 degrees. How much signal "illuminates" the edges of the baffle with and without the waveguide? IME the XT25 has far less energy reaching the baffle edge than most domes. This will be in the region of distances that yield diffraction signatures from about 4K down, the region where most tweeters show wide dispersion. IME the XT tweeters are more directional to lower frequencies on flat baffles.

Any baffle (or waveguide) that presents a non-flat barrier is mostly about diffraction, I suppose, though some of it must be reflections inside a waveguide, not diffraction. Interesting topic in any case.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.