• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Vintage Yamaha NS-18 Speaker Measurements

MAB

Major Contributor
Joined
Nov 15, 2021
Messages
2,741
Likes
6,341
Location
Portland, OR, USA
I just bought a pair of used Yamaha NS-18 speakers from a local sale. They are from the early '70s, about 50 years old. Here is a 1973 print ad with the NS-18 and siblings:
1730184429387.png

I used an online translator on the the text that describes the driver technology.:cool:

They feature an ear-shaped woofer made from coated Styrofoam in an open-back cabinet with a horn tweeter. According to the add, these are the most revolutionary speaker advancement since the early 1920's.;)

There were a few print adds featuring ears in the early '70s:
1730184466170.png

Yes of course, design by analogy. :facepalm: If the ear's shape is what we use to hear sound, why not also use that same shape to reproduce sound?!? o_O

This ear-shaped Styrofoam woofer concept was originally featured on the NS-30:
1730184529636.png

In the NS-30 marketing material it said:
The NS type was developed with a hint from the soundboard of a piano. It pursues the split vibration produced by the bending of the board, which is completely different from the piston motion of a cone speaker. The diaphragm is made of a unique plastic foam selected from more than 200 tests. This prevents quality changes due to external humidity. Moreover, since the specific gravity is extremely light, the efficiency of the electrical energy required for vibration is excellent.
Wait, is it based on the shape of the ear or the shape of a piano's soundboard?!?

Regarding 'the split vibration produced by the bending of the board'... These ear-shaped drivers have no surround as found on a typical woofer, the Styrofoam membrane is attached to a neoprene rubber surround. All of the driver's excursion is the distributed flex of the entire foam membrane, with the boundary of the membrane held stationary.

Here are the pair I just bought. I removed the JA-5004 woofer and JA-0506 Tweeter (on top) from one of the speakers.
1730186353514.png

The speakers are open back design. The driver looks more like a potato chip than an ear to me.;)

Here is the back side of the woofer:
1730186428696.png

It's really quite impressive, with an incredible magnet and chassis. It has similar surface area to a 50cm (20 inch) round driver.

The Styrofoam is adhered to the stiff rubberized surround, which is then adhered to the aluminum frame.
1730186509622.png

The Styrofoam has a coating that appears to have been applied by hand with a brush. Everything is in great shape, especially for the age.

Getting a bit ahead with the disassembly... Before I did anything, I measured each speaker's impedance:
1730186863570.png

I bought them without testing, and was relieved to see both speakers had reasonable impedance and were matched, or were identically broken!:p Those are some wild impedance traces!:)

I listened for a few hours. I had to get used to speakers so low to the floor. I don't care for what they do to some voices. Sounds like lots of my favorite singers got nose work done. They don't have deep bass. They are inoffensive, even pleasant, but uneven. I hear no rattling, no odd sounds, nothing to indicate they are 50 years old. I pushed the bass hard enough to distort, they sound very bad very quickly when pushed too hard. They are fairly sensitive. TBH, there are way worse speakers, even with the odd and enigmatic midrange.

I measured the on-axis response of the speaker at 1 meter with each of the different tweeter contours, Soft, Flat, and Clear.
1730189844010.png

I wonder what all of the artifacts are. If we look at the FR and distortion, each of the cancelations have large spikes in all HD components.
1730190580892.png


Not all of the impedance resonances map to far field response artifacts.

Not quite 35-20kHz as claimed. I wonder if they are worn out. Both speakers have nearly identical frequency response, and impedance traces. Even the distortion plots and impedance resonances are matched. I found Yamaha's datasheet on the JA-0506 tweeter. I used an online translator to read it in English:
1730191138200.png

1730191155498.png


I hope this isn't too inaccurate translation.
1730191209969.png

1730191245067.png

That's a great datasheet. It's nice that they have a frequency response at the bottom, although highly smoothed.
I digitized Yamaha's graph and compared their data to the measurements of the tweeters in my pair:
1730191420926.png

I used similar smoothing to Yamaha. I measured with the 12dB/oct passive crossover. Looks like the same response, assuming Yamaha's trace is without crossover. The tweeter appears to have aged really well. Both tweeters are dramatically well matched. The lenses and magnets look and feel like finely crafted high quality pieces, and have survived 50 years with no defect.

I decided to check the tweeter's electrolytic filter capacitor, it's so popular to cast shade on these components.:confused:
1730194137292.png

1730194233179.png

I'm embarrassed I doubted it, it's also perfectly fine after all these years.

Not much different than a Hovland 2.2 uF fancy-film-cap except for the cap value:
1730194344120.png


I am confident the tweeter and crossover are working as new. I did hear a few instances of static on the tweeter tone control. It was easy to clean and now works perfectly. I soldered the electrolytic cap back in place.:cool:

I measured the woofers. First the impedance and resonant parameters:
1730192619412.png

I am not confident on V(as) and M(ms). I used added mass, and am not convinced that the change in resonance is going to lead to the correct values with this method due to the woofer not working like an ideal piston. In any case, these are odd parameters for sure, much like those Fostex 31" super woofers.

I did some nearfield frequency sweeps of the woofer in the cabinet.
1730193164964.png


More resonances show up at different frequencies than the far-field measurements of the full speaker. Many of the resonances measured nearfield correspond to resonances in the woofer's impedance trace. I have to wonder if the Styrofoam membrane is driven into a large sea of standing waves. Add to that the peaks and valleys caused by cancellations from the front and back of the speaker. Resulting in me recording different modes as I measure from different locations.

I thought I would do some 3D measurements of the JA-5004 woofer alone in the cabinet.
1730195149223.png

It's rough. But fun. I've never measured an open back speaker and am not sure how illuminating this is.
1730195206751.png


What is a bit lost is the woofer's response is smoother off axis. I love these funky woofers.
1730195254044.png


The woofers appear to have aged as well as the tweeter. Both are matched to each other's unique performance. No rubs or buzzes, just extremely uneven frequency response and moderately large doses distortion.:)

The whole speaker exudes quality and attention to detail, plus quirkiness. All of the fasteners are were secured with fine hardware and thread-clocking compound. The solder joints are all immaculate. The wires are all harnessed. The cabinet work is solid. They sound pretty good until the woofer is pushed too hard. I am really satisfied with these and am going to try to get the cabinets cleaned up as much as possible. I'll clean the other tweeter trim pot, but no need to change any capacitors.

Regarding the quirky idea to make this look like and ear, or a piano, or whatever... I have to wonder if at least one engineer on the team was face-palming at the nonsense about the woofer shape. Yamaha abandoned this unique woofer technology after just a few years, moving to more traditional drivers, including focusing on more practical applications of driver design in models like the NS-1000.

I'll stop here and follow up later with some in-room measurements. Comments and suggestions are welcome.
 

Attachments

  • Yamaha_ja-0506_jp.pdf
    829.8 KB · Views: 47
Last edited:
Follow up on the first post. Earlier I established that the pair work, are closely matched to each other even after 50 years, didn't need capacitors replaced, sound strange but fun, and have a weird woofer.
Here are some additional measurements of the speakers.

The individual drivers' nearfield response with the full NS-18 on-axis response for rough comparison:
1730570634953.png

The crossover region is 5kHz. At least electrically it is. The woofer has an large local peak in response which combines with the tweeter's own resonance at 5kHz, produce a large on-axis peak in response seen throughout the measurements. And a much less than ideal rolloff for both drivers. It's hard to capture all of the woofer's resonances, they are extremely sensitive to the position of the mic across the surface, and distance. More like resonant modes of a drumhead, rather than the modes of a piston with suspension like the drivers I am used to.

I set the speakers up on a turntable, the tweeter is offset from the centerline of the speaker so I used that as the axis of rotation for both horizontal and vertical measurements.
1730572376692.png


I made quasi-anechoic SPIN using windowed measurements, across horizontal and vertical axes. The results are similar to CEA-2034 frequency and directivity response:
1730572750011.png

It's super messy. The on-axis peak at 5kHz is clear.
The dramatic trough at 1.1kHz is a center-axis phenomenon; the high-Q trench (circled below) at 1.1kHz extends across +-30 degrees.
1730574391235.png

1730574518164.png

It's an incredibly deep notch in the response, and high Q. There are too many broader resonances to count, and as mentioned earlier, very sensitive to which position you are listening or measuring from.

The resulting in-room response is choppy.
1730575273240.png


I played Cocteau Twins on them last night and they sounded like they had become Quadruplets.;) They are strange sounding speakers, and sound stranger to me now that I have them for a few days. Small changes in head position lead to unexpected changes in the sound. And they are sensitive to changes in room position too. I may be chasing my tail, and not getting good sound no matter the setup.

I took some in-room measurements using Moving Mic Method (MMM). I set them up about 25cm off the floor, toed them into the listening position, measured both speakers together and individually. All with the tweeter control set to FLAT.

1730589700302.png

With MMM I compared the speakers toed in using, to pointing straight out from the back-wall, to Toed Out (pointing away from listening position):
1730590351738.png

The sound better pointing straight out than toed in. I can still hear that loud peak at 5kHz. Toed out away from the listening position actually helps, even if it looks silly. I see this in the SPIN data, and can tell I like the sound better as I move off-axis, even if the treble goes away. Nothing else gets fixed with positioning.:mad:

In summary, I'm glad I got these. They are odd sounding speakers, and have strange measurements. They are very high quality pieces. The Styrofoam woofers are as good as the day Yamaha shipped it. The cabinets and chassis are first class and all business. Clearly a ton of thought and technology was put into the woofer, but it doesn't work well. They go on about the woofer in their marketing materials, built their first NS line of speakers around it, and quickly abandoned the project for more traditional drivers, although made with Beryllium...:)

Here are some scans and text from Yamaha's print ad for the second iteration of the NS line:
1730591608387.png



The Yamaha Natural Sound is a revolutionary concept in audio sound projection, the first real speaker news in almost 50 years. Never before have the obvious defects of symmetrical design been so fully overcome. Never before have audio components benefitted from the rich beauty of musical instrument lore.

And never before have the ear’s physiological characteristics been given their just due, balancing the mechanical flat frequency response so long considered the ideal in speaker design. The distinct shape of the NS Speaker was inspired by the piano soundboard, but it’s constriction is based on today’s most advanced sound technology. First of all, it’s made of a completely new material: foam polystyrene, assuring optimal resonance (thickness tailored to speaker size) and durability. It’s also lighter in weight and impervious to dampness.

Second, it’s unique shape and free-edge design produces a bending motion which is reflected back from the edge, resulting in incredible sound clarity over a wider frequency range, with no distortion caused by the standing waves cone speakers produce. Third, NS speaker boasts completely redesigned drive magnet, the columnar-crystal type YM55. Produced in the Yamaha foundry by special zone melting method, it features perfectly aligned crystal axes which provide superior energy for higher efficiency and reduced distortion.

It's interesting they describe the woofer's edge-reflection. I would say if Yamaha avoided a traditional cone-speaker's harmonic distortion modes, they seem to have traded it for different distortion, more like a drumhead responding to an impulse.
1730591835324.png

1730591853619.png


Thanks for reading.
 
It's interesting they describe the woofer's edge-reflection. I would say if Yamaha avoided a traditional cone-speaker's harmonic distortion modes, they seem to have traded it for different distortion, more like a drumhead responding to an impulse.
Yeah, looks like the funky styrofoam woofer has all the breakup modes. Almost like a DML. Super interesting and clearly quite efficient, but I think it's obvious why they went to "normal" loudspeakers later. ;)
 
Thanks for the extra information.

I believe the strangeness of the woofer might have more to do with the material choice. It must be a lot less rigid than traditional cone materials and flex in odd ways due to the shape of it.
 
  • Like
Reactions: MAB
I wonder how pianos sound on it and what it would sound like with a bit of DSP. Sometimes really irregular FR’s are masked to a greater degree than you expect with stereo playback in room.
 
Yeah, looks like the funky styrofoam woofer has all the breakup modes. Almost like a DML. Super interesting and clearly quite efficient, but I think it's obvious why they went to "normal" loudspeakers later. ;)
Yeah, it really does have everything. While efficient, if you push it past it's limit it distorts really suddenly and alarmingly. When it does, it sounds really bad, like the Styrofoam is tearing.:eek:
 
Thanks for the extra information.

I believe the strangeness of the woofer might have more to do with the material choice. It must be a lot less rigid than traditional cone materials and flex in odd ways due to the shape of it.
I agree, the material choice is odd. Yamaha keeps telling us about the ear-shape, but all I can think of is what a Styrofoam cooler full of beer on a washboard road sounds like.:facepalm:

That being said, it's got to be the highest quality piece of Styrofoam I have ever seen.:cool:
 
I wonder how pianos sound on it and what it would sound like with a bit of DSP. Sometimes really irregular FR’s are masked to a greater degree than you expect with stereo playback in room.
I'm going to try DSP.
I'm actually thinking of a passive notch on the woofer to damp the 5k resonance. But in reality, I probably won't open them up again.
Piano is a pretty much like everything else, a cacophony. I was listening to Bruce Hornsby and I swear he was playing a harpsichord.;)
 
These loudspeakers have always interested me.

Not surprised they measure poorly, I am surprised they are still working properly after 50 years. Talk about building things to last.

How crowded was the market in 1974? I wonder if that was the impetus to come up with a USP that might capture the imagination of the consumer even if it wasn't actually the breakthrough they claimed? A bit like Tekton nowadays, with their tweeter array.
 
These loudspeakers have always interested me.
Me too!
Not surprised they measure poorly, I am surprised they are still working properly after 50 years. Talk about building things to last.
I was hoping for better, but when I saw the woofer membrane was actually rigidly fastened to the driver chassis my expectations got lowered! The reliability really is exceptional, each driver is still perfectly matched to each other.
How crowded was the market in 1974? I wonder if that was the impetus to come up with a USP that might capture the imagination of the consumer even if it wasn't actually the breakthrough they claimed? A bit like Tekton nowadays, with their tweeter array.
Yeah, I was 10 years old at the time these came out, was only just becoming aware of audio equipment. I recall all of those fantastic console systems, and these definitely fit in with that market. But I never recall seeing these except in print advertisements. The next NS series was the good sounding NS-1000 and the bad sounding but very popular NS-10. Funny that the first version of the NS-10 was with a Styrofoam ear-shaped "Natural-Sound" woofer:
1730651326239.png
 
Wait, the membrane is fixed?
I thought the black was a flexible surround, but looking again I see it is fixed.
They never let it have an honest chance, did they? :p
 
  • Like
Reactions: MAB
Wait, the membrane is fixed?
I thought the black was a flexible surround, but looking again I see it is fixed.
They never let it have an honest chance, did they? :p
Yes. It uses a neoprene rubber gasket glued to the perimeter of the woofer, and to the aluminum frame. It allows deflection, but no excursion!
 
A little Manger'ish. Very interesting. Never did know yamaha had made something like this. Thx a lot!

For the OP. To find and report about a treasure like this!
1730205318279_1.jpg
 
Last edited:
  • Like
Reactions: MAB
I wonder how pianos sound on it

The original design was, I believe for keyboards/organs to fit into the bottom cabinet.

I had a pair (can't remember the model) in some original tweed fabric type cabinet- semi pro stuff Yamaha made in the 1970s.

Efficient and sounded like a PA speaker. Couldn't take any power at low frequencies and I was worried about melting the foam near the voice-coil interface. Sold them to a dude who was into weirder speakers than me...
 
  • Like
Reactions: MAB
I wonder what all of the artifacts are. If we look at the FR and distortion, each of the cancelations have large spikes in all HD components.
View attachment 402442

This shouldn't be surprising. If a speaker is making 50dB of 2nd harmonic during a sweep but then the fundamental level drops dramatically, then the percentage harmonic (relative to the fundamental level) will increase.

I've seen this crop up a few times (particularly when speakers are subject to an ASR review, and LF distortion seems very high, when in fact it's just that the output level is low), and thought it was worth addressing. Try viewing distortion as absolute level.


Chris
 
  • Like
Reactions: MAB
Back
Top Bottom