Classic monitor designs?

[system7]

After an extended foray into a seemingly perfect speaker, I confess nothing here has changed my mind. LOL.

ALL loudspeakers are a trade-off. You trade little midbass drivers having to make huge non-linear excursions, and hence into distortion, against big ones which run into cone breakup at a much lower frequency.

It's easy enough to reduce the bass output from a speaker, so it distorts less, but then less bass!

Here's an interesting diversion. Joachim Gerhard designed a goodish speaker called the Sonics Arkadia. He used the SEAS Ca22RNX, which has an awkward resonance around 3.5kHz. Joachim notched it.

Troels Gravesen did this as a 3 way, using the same SEAS 27TFFC tweeter. I have no doubt Troels' design sounds better. Especially going loud.

 

[Ro808]

Troels' Classic 3 Way - in its respective iterations - may well be the epitome
of this thread, i.e. (near) full range in a reasonably sized enclosure and without major technical flaws.

However, many of us - Troels included - like to push the boundaries to gain some things at the expense of other things.
I guess this is what makes loudspeaker technology such a fascinating subject.

 

[Ro808]

Joachim Gerhard designed a goodish speaker called the Sonics Arkadia. He used the SEAS Ca22RNX, which has an awkward resonance around 3.5kHz. Joachim notched it.

That resonance is pretty 'spikey', but at 3500Hz high enough not to cause major issues. Besides the cone is paper, sooo...
I wonder how that spike would look like with an uncoated / natural paper cone of otherwise similar properties.

 

[Ro808]

If you take a look at the specs and the T&S in particular, set against the statements of Joachim Gerhard in the interview referred to earlier in the thread, some questions may arise:

- What is the amount of energy storage due this drivers' suspension parts?
- Why is net sensitivity only 86dB (compensated for BSC)?
- What is the use of Fs = 29Hz, if the Linear and Maximum Coil Travel is only
12 and 26 mm respectively, considering Qts = 0.41 and VAS = 97L?
- How does the Force Factor (BL) relate to this?
- What are the mechanical properties of the surround and how does it affect response?

 

[Ro808]

Correction:
Coil travel in itself is very good, but my fundamental question concerning this driver is:
What is it?
- A Subwoofer?
- A mid/low woofer?
- All/none of the above?

This question often comes to mind, when I read spec sheets of contemporary hifi woofers.

By the way, this first sentence must be an attempt at humor by SEAS' marketing dept:

"Classical handcoated paper cone and matching natural rubber surround produce a
well behaved roll off characteristic and reduce potential resonance problems"

 

[Ro808]

To conclude the elaboration of the Wilson WATT, here are the final 2 posts, on....
the DRIVERS (how convenient)!

(Including quotes from Thorsten Loesch)

The tweeter in every WATT has always been a Focal inverted dome driver. In Series I and II this was an OEM derivative from the Focal T101 / 120 / 121 K-series, 30mm Kevlar dome tweeter with a Latex treated foam suspension. The Wilson unit - contrary to widespread assumption - was fitted with a Glassfiber dome and (eventually) labeled T-123W.

Differences between the original T120K and the Wilson OEM tweeter are:
*The large vent in back of the magnet
*An orange colored dome rather than the usual yellow dome (the Wilson dome is thicker and made from Glassfiber instead of Kevlar)
*The back steel plate is much thicker and of a higher grade steel.


"The original fiberglass focal was somewhat spitty, snifty sounding, due to the low breakup at 16KHz, the later Ti ones just shifted this breakup high enough (23KHz) to to reduce this tendency low enough to no longer be bothersome. It is possible to re-work earlier Watt's with Ti repair sets, of course then they are no longer original, but the spittiness is reduced."

Some people glorify these kevlar Focal tweeters, but I am certainly not a member that group. Independent of measurements, electrical and mechanical properties etc, I have never heard one loudspeaker (mostly Focal/JMLab in this case) with these tweeters that didn't shriek.

With regards to 'long term reliability': the foam surround/suspension - dependent on exposure to UV radiation - is due to decomposition after some years. This will have catastrophic effects on performance, which is pretty disastrous anyway (see image no. 4).

BUT! I have to admit, as Thorsten stated, the later Focal titanium tweeters (TC120T, TC120Tdx2) are much better. I remember listening to FOCAL 800 series loudspeakers and was pleasantly surprised by the well behaved upper mid and high frequencies.


So much for the tweeter (no offense intended :- ).

 

[Ro808]

Just for fun...

Compare this response of a similar sized pro driver to the curve of the CA-22RNX from above.

It's a 4 Pi measurement in a vented box of ca. 30L.
No, it doesn't have a Fs of 29 Hz, but with proper tuning it will play down into the 40's, probably even lower with some room gain.

Flatten everything above 200 Hz and see what's left.

 

[Ro808]

For the SEAS 3-Way Classic mkII Troels exchanged the SEAS CA22RNX for the CA22RNY.
This seems an altogether more sensible woofer, during development of which SEAS appears to have relinquished the 'all in one' approach.

This driver would (definitely) be a more suitable choice for Joachim Gerhard's Sonics Arkadia.

 

[Ro808]

The 6.5" woofer in the first Wilson WATT series is an OEM version of the renowned SEAS CA17RCY.

It is not so much this specific driver, but the legacy of vintage woofer design it inherits that makes it interesting.
These virtues have been lauded by Joachim Gerhard, Troels and many others.
Unfortunately, this specific CA17RCY suffered one design flaw, its coated textile dustcap caused some resonance, which was notched in the original WATT and resulted in a severe impedance drop to nearly 1 Ohm around 1000Hz.


Quote from Thorsten Loesch:

"Analysing the WATT/WAP System a number of details emerged. First of all, the Midrange. The more recent Versions (from 5.0 onwards) of the WATT use a Scan Speak 6.5" Driver, notably the 18W8545. This is a Driver with a Paper-Cone to which Carbon Fibres and damping materials have been applied. Previous versions used a now discontinued 6.5" Driver from Seas, the CA17RCX.
I have heard a range of Speakers using the Scan-Speak 18W8545 in the Midrange (including the Watt/Puppy 5). I have yet to be convinced of the Value of this Driver in this application (Bass-Mid - the 18W8545 is a killer Bass Unit). Artificial, cold and generally not very pleasant sounding (to my ears) I really did not like what I heard from the 18W8545. In comparison, the Bass/Midrange Driver used in the earlier Watt sounds very natural, though by no means laid back or romantic.
I believe that the reason for this sound are a pair of not well damped resonances in the 3kHz region. This means that a very low X-Over Frequency is needed and non of the Speakers I have heard does that. One could also implement complex notch filters.
The other option is to simply ignore the resonances as they do not show up in a frequency response plot (they are well visible in waterfall though) and to hope no-one notices. It seems the designers of all Speakers using the 18W8545 that I have heard to date took that approach.
I decided to have a long hard look at the SEAS Catalogue and discovered the CB17RCY which seems to be the direct replacement of the CA17RCX. Indeed, it appears that this driver is used by Wilson Audio for the WITT and CUB speakers. It has few audiophile credentials, being a boring doped paper-cone, but this driver delivers a extremely natural and uncoloured midrange."

This later CB17RCY appears to an upgraded version of the older RCX. It features a nice injection moulded
metal chassis also a paper cone AND most importantly a coated paper dustcap, instead of the textile cap of the RCX.
This driver is also the basis for the expensive and 'SEAS OEM exclusive' Audio Note woofers, as shown earlier by Steve.
Multiple versions both the older A/CA17 and the later CB17/P17 exist; i.e. the SEAS CB17RCY-P included a phase plug.
All A/CA woofers share the same (SEAS H3310) rubber surround. To make identification even more confusing, OEM drivers have different labels: a SEAS 17F-GWB for example, is basically a Seas CA17RC(Y).

For this reason it is difficult to tell what the specific OEM modifications of the Wilson M-5021 labelled driver consist of, apart from the large metal cap that hides the magnet.
After some research, I found a very similar - if not identical - SEAS driver which appears to be an OEM product for Meridian loudspeakers from the same period. This is the SEAS P17 RCY-BS (BS stands for Boothroyd Stuart, No bullsh.....g here).
Evidently data, for this specific driver is unobtainable, but it is highly unlikely that performance is completely different from the original CB17RCY. Images of both the Wilson and Meridian driver are included (images 5-10).

Variations and OEM versions of this vintage SEAS woofer series were used by numerous manufacturers in the 80's. Without prior notice, I remember at least a couple of loudspeakers with these woofers - 2 of which from local manufacturers - to sound very good and: even with low power amplification of average quality.
They all had one thing in common: a natural, open sound with accurate timbre and good dynamic characteristics, not in the least 'shut-in' or 'sterile' like many modern drivers. In order not to get misty-eyed: many tweeters couldn't match the qualities of these woofers and crossovers were often built from crappy components.

I have included a response curve from a Seas 17F-GWB, an OEM derivative of the Seas CA17RC(Y) from one of these vintage 2-Way monitor (type of) speakers. Not a bad performance, after 30 years of (ab)use and only glueing a fresh (H3310) rubber surround (see 4th image).

These vintage drivers are still held in high regard and some are still available from SEAS in limited production runs.

 

[Bigun]

not in the least 'shut-in' or 'sterile' like many modern drivers.

I like the idea that my speakers have low distortion and don't colour the sound, or the enclosures for that matter. I particularly dislike cone resonances, or similar, adding to the sound. But it does put me off to hear that a driver is not 'sterile' because this implies it's coloured.

 

[Ro808]

I like the idea that my speakers have low distortion and don't colour the sound, or the enclosures for that matter. I particularly dislike cone resonances, or similar, adding to the sound. But it does put me off to hear that a driver is not 'sterile' because this implies it's coloured.

Resonances from cone, dust cap, cavities, enclosures and the like are - without doubt - highly undesirable. As I have stated earlier 'the sound signature' from ANY driver (although some people believe otherwise) depends on many
interdependent factors, mechanical, electrical and material.

It's a simple fact of physics + chemistry in the analog world.

The closest you may get to a distortion free, absolutely neutral driver - 'sound source' might be more appropriate here - is to virtually define one in the digital domain. What would that sound like?
Pretty sterile, I guess.

 

[Bigun]

If the music is recorded well, I should be listening to all the harmonics and 'distortions' of the instruments, be they wooden violin's or other instruments. I can add distortion in my preamp perhaps it can be switched in or out. But if I have colour in my speakers it's baked-in to my system and will forever colour my listening window.

The more I read about multi-way speakers though, the more I am seeing the complexities associated with their design and the more I appreciate the work that has to go into making good decisions on tradeoffs. And I'm starting to fall out of love with FR drivers because so many of them have such atrocious FR.

 

[system7]

I've been hovering for a while while you do this interesting rap on 6" plus 1" tweeter speakers, Ro808. LOL.

TBH, all the SEAS 8" paper and polycone woofers look very doable to me.
WOOFERS

The CA22RNY is the big magnet version of the CA22RNX. The woven polycone U22REX/P-SL resembles the CA22RNX but for an extra copper damping ring, a shorter voicecoil and a phase plug. All quite large voicecoils and aluminium formers, which aren't the last word in extended frequency response.

But all nice. All with well-behaved built in rolloff. They'd probably all work well with a 7.5R/15uF Zobel to correct impedance.

People here are probably sick of seeing the WLM La Scala, using a Visaton W200S 4 ohm, and a TW70 tweeter, which inspired me. But my ideas have evolved. I haven't finished fiddling around with this, but it's all working tremendously well with impedance correction. 4th. order tweeter crossovers sound excellent, and I have had success with these before. The circuit below was using something like 1mH, 8.2R and 4.7uF bass shunt IIRC.

My latest idea is a goodish LR4 design with less of the omega impedance. Impedance and phase alignment to die for in an an LR4 style. Of course, above and below axis you get the usual hole, which I try and avoid with near BW3 designs.

On the subject of 6" bass, I always find terrible time-alignment problems. I have heard the scruffy looking Scan 18W-8545 in a ribbon MTM, and it's the best and most flat of the scan 6" drivers. Probably crossed low.

 

[Ro808]

Agreed on full range drivers, some may sound 'good' or 'nice', they certainly do have tradeoffs like any loudspeaker concept, multi- or single driver.

We have roughly 2 tools to evaluate 'soundquality' from a given source
- Ears (subjective)
- Measurements (towards objective)

In order to illustrate that 'measurements don't tell the whole story' an 'in house' comparison:

Harman Group - or should I say: Samsung - offers 2 high end products, which are both considered to be excellent studio monitors. In fact, both are being used in a number of well known studios.
They are even 'only a few grand' apart.

1. Revel Ultima Studio 2
2. JBL M2 Master Reference Monitor

Measurements of both loudspeakers are available from numerous institutes, magazines, forums etc.
Apart from a few very rare exceptions, most people would agree on the excellence of 'measured performance' of these 2 speakers.
Yet, you only have to look - not even listen - to know they WILL SOUND rather, if not completely different.
And they do.
There's a nice thread on audioheritage about it and I have crossed threads on other websites in which these loudspeakers are compared.

 

[Ro808]

Steve, is that from the same Monitor Audio cab you've (ab)used before?
I guess, with these La Scala Visaton cones + turbo tweaks these things have never sounded better.
That response looks really nice, that little cone is quite amazing isn't it? And these cost next to nothing. It's a bit difficult to tell, but you did add some BSC didn't you?

Lately I found 2 empty Monitor Audio Gold / Studio SE, or ProAc Tablette cabs in the storage.
Might try something similar....
But these are too small for 6" or 8" unfortunately.

There will be more rap, tomorrow

 

[system7]

I do all sorts of things with 8" bass, reflex and closed box.

FWIW, the Visaton W200S is a high inductance reflex woofer. This is best judged from the impedance curve, not Le. At 10KHz, the Visaton is up round 40 ohms. That tells me that it has an inductance around 2.3mH as Visaton say. Not ideal for closed box at all, which would be flat at around 12L if I recall the simple calculation correctly.

The SEAS woofers are around 20 ohms at that point. So less capacitative impedance correction needed, IMO.

High inductance comes with bafflestep built in. You scarcely need apply it in a real room with all it's room gain below 200Hz. Just put the speakers next to a wall to get some bass.

My own trick with the severe time-alignment problems of an 8" bass is just flip polarity for a 3kHz crossover. 5 cms is half a wavelength, so it works. If you want an even lower crossover, recess the bass further like Harbeth do.

6" is a near hopeless case, IMO. Unless time aligned.

 

[Ro808]

Lately I have have been reading Ron Elliot's, Linkwitz pages and similar literature on XO's and your experiments seem to confirm theory.
I guess John Devore has chosen a similar path with his Orangutans.

In contrast to the monitor speakers I have discussed so far, I am not at all a fan of 6" or smaller + tweeter. In this respect, I am working my way up from down below
My favorite concepts are from the 'bigger is better' category.

Recessed drivers + a waveguide are useful in counteracting phase issues.

 

[Ro808]

Inadvertently, I referred to the Audio Note 8" drivers being based on the SEAS CB17/P17 series
These are of course 8" instead of 6.5" woofers and as such based on the larger CA, P and L21 series.
Probably the most famous of these 8" drivers among DIY enthusiasts is the Seas CA21REX.

However, these are members of the same 'SEAS family' from that period, just like their smaller brothers, the legendary 4.5 Inch woofers which, despite a coated fabric dust cap , were used in a huge number of now classic loudspeakers.
Just to name a few: Keesonic Kub, Monitor Audio ONE Mini, Sonus Faber Minima and Sonus Faber Parva, many ProAC Tablets, some Naim speakers, as mid driver in the Thiel CS3.0 and last but not least the 3A Master 5.

In SEAS lingo:
Seas 11F-M: 4.5" doped paper cone, diecast magnesium chassis, otherwise known as the H143 and Seas 11F-GX (H149)


Here is an excellent guide for those who like to completely rebuild their 11F drivers.


Fortunately, these are reissued by SEAS, along with with a number of other 'SEAS Classics'


Examples:

Seas CA11RCY H0149-08 - NEW 2017 MADE

SEAS CA11RCY H0149-15 - NEW 2017 MADE PROAC MINI TOWER EBT REPLACEMENT (15 Ohm Midwoofer)

Seas CA17 RC 2 H0843-08 Woofer - The updated Version 2 of the classic CA17 RC

Seas P17RE/P H0528-08 Woofer - High power Seas Woofer with a 6.5" polypropylene cone, Phase Plug and High Loss Rubber Surround

Seas P21RE/P H0942-08 Woofer - H0313 2015 remake.

 

[system7]

SEAS have been ploughing the same furrow for many years now.

I was looking at a prototype John DeVore Orangutan. He started with a recessed baffle and what looks like a H1305-08 CA26RFX. Morel waveguide tweeter for efficiency. Maybe flipped polarity, BBC Rogers style. IDK.

And then got something like this built with a phase plug: H1411-08 A26RE4

The final omega impedance curve looks like regular stuff. Can't be a million miles away from Peter Comeau's efforts for World Audio. A bass crossover that doesn't do much! LOL

 

[Ro808]

That A26RE4 is a nice driver with a rare natural rollof. Unfortunately it's fairly low sensitivity.
But it will likely do better in that large cab, compared with SEAS' own A26 kit

Another one: I like the concept, but not the woofer.

For those interested, you may download all data incl. Magazine Article (in German) here.