In music there tends to be a bit more power at low frequencies but not a lot more (varies with genre) and so it won't withstand much higher power but a bit more power. This is why midrange drivers tend to have 100+ W power handling rather than the typical 5-10W of a wideband driver. It is one of the main reasons wideband drivers are not suitable as midrange drivers in 3 ways.
Power handling is not directly related to cone displacement but how well the heat dissipated by the resistance of the coil can be transferred to the air and away. If the ability to transfer heat is poor the temperature of the coil builds up and the motor performance changes. If the OP wants to look it up it is often termed power compression and is starting to be measured and plotted in some home audio speaker reviews (pro audio has always given it attention). With a normal modern high fidelity speaker with typical amp and speaker power ratings of say 150W (that of the new B139) and a speaker efficiency around 90 dB it tends not to be an issue at standard levels in typical rooms though would be at higher party levels. But when the speaker power handling is a lot less and the speaker efficiency less it becomes an issue at levels one would typically want to use for normal listening in the home.
I know all that, but this story has expanded a lot. There are a lot of FR drivers with higher power, not all of them are 5-10W.
For the combination with this particular B139, you don't need a very powerful FR driver. With those small drivers and home listening, the movement of the membrane at low frequencies is more critical, and the power it can withstand at medium and high frequencies is less important. There is not much power going there anyway. Here, the low frequencies will be cut off with a crossover.
For the combination with this particular B139, you don't need a very powerful FR driver. With those small drivers and home listening, the movement of the membrane at low frequencies is more critical, and the power it can withstand at medium and high frequencies is less important. There is not much power going there anyway. Here, the low frequencies will be cut off with a crossover.
I have a Fostex FE103 Sol 16Ω here up for grabs… and one with fried VC i will disect and post pictures…
dave
I think there are a few more lower than that?
I approach system quality as a musician (because that is what I originally trained as at a music conservatoire) and one test I use is a good string quartet recording. Yesterday I connected up the Celestes (with the B139) to remind myself of what they're like. I used the Pavel Hass quartets recording of Schubert's 'Death and the Maiden Quartet' on Supraphon - stellar musicianship and recording.
The cello should provide a rich swell on its lower register and the upper notes on a violin should sing like a Nightingale - but on these Celestes? Urghhhhh. The cello is largely absent and all the other instruments sound harsh and artificial - actually quite painful to listen to. I know the XOs capacitors will need replacing - but this is a still a sows ear - and it has to be the various breakups of the B139s in the crucial registers of 200 to 1000 Hz.
Is this going to affect the Leak Sandwich speaker as well? (referring to the original 15 ohm one) as I believe that is styrofoam sandwiched bewtween Al.
The Leak sandwich and the KEF B139 were designed to be stiff. That stiffness pushes resonances up into the midrange where they become irrelevant for a woofer. But they are a serious problem for a midwoofer. A large midwoofer has to be designed to handle resonances and that leads to a substantial different design to a woofer. To control and damp the resonances it will include things like different materials, corrugations, damping, doping,... which will tend to bring the resonances down in frequency, down in magnitude and often distributed in a more even controllable way.
Not sure I understand planet10s point w.r.t. to foil and styrofoam. Aluminium is a metal and has negligible damping. Styrofoam has some damping but isn't a material one would choose if damping was the primary purpose.
I was unaware the B139 had ever been used as anything other than a woofer and would be curious to see the response in the midrange which cannot be good. No woofer can be expected to have a good midrange response although a midwoofer may have a reasonable low end response albeit down a bit on a woofer.
The leak sandwich might be both stiff and have some damping provided by the separating layer. A quick google has not come up with anything to confirm or deny this w.r.t to midrange resonances. Anyone? The later models were 3 ways removing the need for the woofer to be well damped.
Not sure I understand planet10s point w.r.t. to foil and styrofoam. Aluminium is a metal and has negligible damping. Styrofoam has some damping but isn't a material one would choose if damping was the primary purpose.
I was unaware the B139 had ever been used as anything other than a woofer and would be curious to see the response in the midrange which cannot be good. No woofer can be expected to have a good midrange response although a midwoofer may have a reasonable low end response albeit down a bit on a woofer.
The leak sandwich might be both stiff and have some damping provided by the separating layer. A quick google has not come up with anything to confirm or deny this w.r.t to midrange resonances. Anyone? The later models were 3 ways removing the need for the woofer to be well damped.
The situation is more difficult than I thought. It is possible that the B139 has started to disintegrate and the T15 is no better. Maybe you should replace the electrolytes and renovate the boxes a bit, then keep it as it is for sentimental reasons and devote yourself to making brand new speakers. If it has to be B139B then buy those new copies. At least solutions for TL and other loading are known for them. It can also be a bass reflex.
I think that aluminum foil on the membrane is not damping, but increases the stiffness of styrofoam (like aerolam sandwich floor in airplanes), protects it from evaporation and UV light. Styrofoam has a tendency to evaporate over time and shrink when exposed to air. I've seen a lot of those B139B and B139BD passive radiators with cracked diaphragms.
When KEF designed the original Celeste (I believe circa 1960?) I believe they had no squawker - so used just the T15 and the B139 crossed over at 1,000 Hz. This is the later iteration of the Celeste (Mk2) (aka K2) which uses the Mk2 versions of the drivers - I think this was 1966. By then, I believe, they had a squawker in the form of the B110 but carried on with the original design. Their mid-60s data sheet gives the frequency range as up to 1,000 - yet for a three-way enclosure they recommended a XO of 200 - so they knew that its performance above that was not especially good.
...or that may just be my fussy ears! They are really sensitive to any form of roughness, bad coloration or sibilence in that frequency area and I find a lot of speakers/headphones sometimes just painful.
You can try recording the FR range with a mobile phone. There are several applications for Android, Spectrum Analyzer is called something like that. You also need a laptop and an online white noise generator. You also have an online tone generator, you put it on a sinusoidal signal, change the frequencies and listen to what happens.
You have to take those B139s and T15s out of the boxes, inspect them carefully to make sure that something hasn't come off over time, or cracked, that the coil wire might not have started to separate, that the coil has not been centered, and the like. Then you listen to them outside the box with a tone generator to hear if there are any strange sounds (resonance). T15 should not be tested below 500Hz probably.
You need to determine whether the resonances you hear come from a worn-out box or from the drivers themselves. The 8uF bipolar electrolytic capacitor should definitely be replaced.
https://www.szynalski.com/tone-generator/
You need to determine whether the resonances you hear come from a worn-out box or from the drivers themselves. The 8uF bipolar electrolytic capacitor should definitely be replaced.
https://www.szynalski.com/tone-generator/
Last edited:
Thanks. I've got a tone generator so I'll play around with that during the xmas period. So far I've not done anything with them - still in their cabinets.
Pay particular attention to the condition of the rubber suspension, spider and on all glued joints. Sometimes even the bad solder joint on the PCB can make strange noises.