default page

Vintage
Wharfedale Super 8/RS/DD, 10-15 ohms

These vintage Wharfedale Super 8 drivers were bought to possibly serve as mid-drivers in a 3-way open baffle set-up, but before this minor report starts, I have to tell I have a copy of Gilbert Briggs' book Loudspeakers, bought in 1971. This is one of the best books available on loudspeakers and despite several decades old, still tells most of what is important to know in constructing dynamic loudspeakers. Being 19 years old back then, I desperately needed basic information on loudspeakers and this was my speaker bible for many years, and still is today. Who can write a book on loudspeakers spiced with wise philosophical quotes without being corny? Gilbert Briggs can, but dear Mister Briggs in Heaven (I believe you are there), how on Earth could you release a crummy driver like this? I can forgive you the green felt ring and awful blue alu dome fitted to the whizzer, both probably meant to catch the eye, but all in all, this is probably the worst driver I have ever had, and here is why.
Before I continue my story on the Super 8, here's a response from a former Wharfedaler employee:

Read mail exchange with former Wharfedale employee here.

The Super 8/RS/DD is in many ways a modern driver. A highly ventilated chassis houses the straight-sided paper cone attached to a voice coil wound with aluminium wire. Not many did that back then, because soldering the copper litze wire to the alu wire wasn't an easy thing to do and it took an ultrasonic bath to get rid of oxidation before the solder would bite. The driver has an excellent ceramic magnet system ensuring lifelong performance. So does the fabric surround if kept away from mice and termites.... But back to the cone and the surround, because this is where the trouble originates. The straight-sided cone, voll-konus as the Germans would say, has some great properties when it comes to working pistonic - up to a certain point. Beyond that, it breaks up like crazy and rings like a bell. For this driver it happens around 3 kHz and all of a sudden the amplitude rises some 15-20 dB above average level. A gigantic peak is the results and there's not much we can do about it. Well, Briggs added a whizzer cone and maybe that was the best thing he could do, because the whizzer cone makes a terrible mess in the 3-8 kHz range and to some extent smears the nasticies by producing a deep dip from 4 to 7 kHz. To further dampen the peaks, a polyester foam strip was added outside the whizzer cone. This foam ring simply disintegrated by touching, being probably more than 40 years old.
I have a few straight-sided cone drivers (Philips 9710 and some I've made myself from paper-foam-sandwich cones) and they are immediately recognised when attached to my CLIO amp by a faint hizz. By the way, so does all magnesium drivers, but I've never seen a metal driver with a 20 dB peak - and this low in frequency.

Left: The red graph above is the Super 8 as-is. Severe peaks at 3-4 kHz and 7-12 kHz. The blue graph is from removing the whizzer cones. Now the peaks are at least concentrated between 3 and 9 kHz, but what a peak!

Right: I usually don't show nearfield measurements, but here's one. 32 ms window and 1/12 octave smoothing. It's a rare thing to view such a smooth response from 100 Hz all the way to 3 kHz. The blue is minimum phase and display the severe phase-shift at 3 kHz where all of a sudden, things starts getting out of control. The poor thing rings like a bell as can be seen below from the impulse and step response graphs.

14,500 lines. Magnetic lines were the thing of the day in those days,
telling about the magnetic strength in the voice coil gap. Very few
brag about magnetic strength these days as most people know that
high magnetic strength is not necessarily what makes us happy. Or
does it? Some prople think that a huge magnet is a sign of quality.
What do I know about marketing?

Measured TS data. Actually the Re=9.87 is with driver with whizzer cone, the
Re=10.3 with whizzer cone removed. I took an average anyway.
What was really strange about this was the calculated efficiency, 91 dB, much
more than what was measured. Explanation below.

Removing the whizzer cone displays a voice coil wound on a paper. Ensures low
weight and reduced induction.

Left: Impedance of driver with (blue) and without (red) whizzer cone. At 3 kHz we can see the impact on impedance from cone break-up.
Right: Frequency response of drivers again, blue = without whizzer cone, red = with whizzer cone.

Above: This is an interesting graph. Red is the driver without the whizzer cone and a blue is the same driver with a coating applied to the fabric surround. Thinking the surround looked a little transparent, I pulled out the surround and put it to my mouth and much to my surprise, the fabric was as transparent as any similar type of fabric! Hmm.... Really never seen a driver with an acoustically free suspension. I mean, this will severely short-circuit acoustically and was this really what was intended? I know that Briggs was fond of the acoustic vent and often sliced the back of his cabs, covering it with felt material in order to smooth the rise in impedance from a closed box. But a surround as an acoustic vent? I don't believe it. It's gotta be a severe mistake. So, coating the surround gave an immediate increase of 5-6 dB in response. That's something! And all of a sudden the break-up peak is "only" ~12 dB above average level. Now the average SPL is well in accordance with the predicted efficiency from the TS measurements.

Left: I don't think I can find another driver in my files with an impulse response like this.
Right: Step response tells the same story, peak after peak until it dies.

Conclusion:

Hmm... The driver is back on the shelf until I fully recover, but one thing is clear: If we want to work with straight-sided cone drivers, we have to be prepared for 4th order filters like seen below. I've tried to model a two-way system from the non-whizzer driver and a Vifa XT25TG tweeter fitted with a waveguide, allowing a rather low point of crossover, and it looks good. Sensitivity around 89 dB is not bad at all from a small two-way and we cannot ignore that straight-sided cones can make some of the best midrange available. The cone basically works pistonic, i.e. no smearing of details like from a cone having a curvilinear profile distributing the break-ups to all frequences and if done well, can produce a smooth roll-off, allowing simple crossovers. Compromises again...

Super 8/no whizzer with coated surround and Vifa XT25TG + waveguide.

Well, the Super 8 was intended for an open baffle system using low-order filters, but I'll have to source other options.

Return to intro page.

Mail: 09-05-2009:

I was employed as a technician by the Australian import agents for Wharfedale spkeakers;Simon Gray P/L., inthe '60's. I can confirm that the photographs of the referenced speaker shown on your website are indeed that for that model and date i.e. 1/67 as clearly marked on the cone. Later models differed only in that a black Paint was used on the aluminium dome instead of the translucent blue coating and a slightly modified frame. There were many assembly problems with this model in the Rank/Wharfedale era caused by increased production demand. The coil and former required additional coating (doping) to fully bond the coil and to stiffen the paper former.The cloth roll surround was rarely fully sealed and leaked air.The stepped cone was not a good choice as it was intended for the U.S. market which sought a higher power handling capacity i.e. a 1.5'' voice coil version. Often the aluminium dome was not fully bonded to the tweeter cone causing odd distortions if not buzzes. After a few years the black rubber sealant dried out causing the the surround to stiffen. The suspension then became non- linear as the roll rather than the corrugated disc spider controlled the restoring force. Barry M. Arnstein.

Dear Barry, thanks for the additional information. This makes a lot of things fall into place. Best regards, Troels.