The Radford S90 was the speaker that really launched me into speaker design… i didn’t even know they made amps till much later.
I have a friend with 3 Garrert Deccas. And triple Quads at the other end (driven by 6 channels of LEAK (based) EL84 amp.
dave
My first serious speaker building exercise was to modify a pair of Decca Corner Horns a friend gave up on building. I turned them upside down and built in a (fairly) accurate Radford T-line - Wharfedale 12", Goodmans Axiette and Kef T27 with a Concerto x/o IIRC - happy days!
Before 2k TLs were only guessed. After 2k you could model them just like (sorr of) a BR.
There is no such thing as a perfect speaker, what is best is a case of what compromises best suit.
dave
There is no such thing as a perfect speaker, what is best is a case of what compromises best suit.
dave
It doesn't surprise me that ribbon would have trouble at 1k. The distortion would likely be rising below about 1.5K and diaphragm motional control at higher volume will be lost. The typical small free swinging ribbon is going to get very busy down there as the volume goes up.
However with the horn helping the low end and playing certain types of music and not too loud, it should still be possible to operate that low and when things are aligned right it can sound very good.
My own experiments with a small ribbon used that low that has better control of distortion and diaphragm motion result in exceptional sound.
Always seemed odd to me that with all the deep soul-searching about sound quality, that anybody would want something as crude as an electrolytic capacitor in your system, esp with something as delicate and nice as a Decca ribbon. They can be 100% off their rating and varying up and down.
I'd certainly go for bi-amping, although with an electrolytic to block turn-on pulses. Of course, I'd always go for bi-amping. Would the little amp would cost any more than the passive XO components?
While some great woofers of yesteryear are comparable to the best today, I've often wondered about all those new ribbons for sale compared to the very old Decca? But damped silk fried-egg drivers are about as perfect a source of tweeting as anybody could want.
I'd certainly go for bi-amping, although with an electrolytic to block turn-on pulses. Of course, I'd always go for bi-amping. Would the little amp would cost any more than the passive XO components?
While some great woofers of yesteryear are comparable to the best today, I've often wondered about all those new ribbons for sale compared to the very old Decca? But damped silk fried-egg drivers are about as perfect a source of tweeting as anybody could want.
"I've often wondered about all those new ribbons for sale compared to the very old Decca?"
Ive built enough ribbons to know that the ribbons presently offered on the market most likely have nothing significant on the Decca
Ive built enough ribbons to know that the ribbons presently offered on the market most likely have nothing significant on the Decca
Ok yea not suprised, was giving Kef benefit of the doubt.
I never laid hands on the B139 , however I have made a ton of protos of my own in similar skinned foam design. No matter what I did with shape size type of foam etc etc They ALL had a nasty around 1 k or so and they never sounded good at 4-500. Always had to cross around 250- 300 max
I think we are going off-topic here. 😱
I've been mulling the filter over...
Dave suggests this as the 1.6kHz high pass:
I've been scratching my head over 1.6kHz for a 15 ohm 12" woofer.
This is an 8 ohm eminence PXB2-2k5CX 2.5kHz jobbie:
Using coaxial products for DIY monitors and home hi-fi applications
So for 1.6kHz and 8 ohms, 3mH and 15uF makes sense to me. Convert to 15 ohms, 6mH and 6.8uF... and negative polarity between the drivers.
Well, I tried modelling it, and I got around 6.8mH and 6.8uF. I expect you might put an ohm or two in the shunt too.
What does everyone think?
I've been mulling the filter over...
Dave suggests this as the 1.6kHz high pass:
I've been scratching my head over 1.6kHz for a 15 ohm 12" woofer.
This is an 8 ohm eminence PXB2-2k5CX 2.5kHz jobbie:
Using coaxial products for DIY monitors and home hi-fi applications
So for 1.6kHz and 8 ohms, 3mH and 15uF makes sense to me. Convert to 15 ohms, 6mH and 6.8uF... and negative polarity between the drivers.
Well, I tried modelling it, and I got around 6.8mH and 6.8uF. I expect you might put an ohm or two in the shunt too.
What does everyone think?
Last edited:
That's what your precise models tells you but a gamble at what you will get with electrolytics (which should be paralleled with some good caps too).
Don't forget, even at 18 dB/8ave, an octave down is still putting lots of power into the ribbon, esp since that is getting into the core band with lots of music power. Another reason to go to DSP: to get steeper slopes without moving the crossover point to high.
The only box i heard with a 1kHz XO did not struggle as volumes went up. Active, 4th order should work fine.
dave
interesting..
what X over was used? passive active order ?
do you/anyone know the approximate physical dimensions of that diaphragm?
length width ?
It was so long ago i can’t recall, given the day likely 2 or 3 order.
Also been a long while since i replaced a ribbon, something like 15mm x 70 mm or so. Easy to change.
dave
Also been a long while since i replaced a ribbon, something like 15mm x 70 mm or so. Easy to change.
dave
system7; said:
Thanks Steve. I am hoping that, as the Decca x/o under discussion is the one designed for the London and the Kelly DK1 (which I have), the rest of that original schematic should be the business.
But how critical are the inductor values? A search yields components pretty close in value but not exactly the same.
I have been wondering about that discussion about electrolytic vs other caps for x/os these days. Any recommendations? Presumably same values and wkg V and not audiophile/snake oil territory - these are 50/60 year old drivers after all.
Think of electrolytics like you'd think about old carbon-zinc flashlight batteries.
The cap is made by pouring a damp paste put between the sheets of aluminum foil and it dries out over time. Very hard to be precise about the manufacture in the first place. If you look on the cap, it will have a very wide tolerance like +50 to -20 or something (I'm not near my workbench). Sound like the kind of garbage you wouldn't want in the path?
As far as one plastic or ceramic cap versus a different manufacturer with the same specs, my guess it that it is nothing but snake oil and deception.... appeals to people who think they can hear things others can't.
The closer I look at passive crossovers, the more illogical they seem to me. Makes little tech sense except as a crude tool. At the very least, you need to be able to measure your components (which very few of us can) or measure the electrical results (which is easy if you know how to wire REW) or the acoustic results (which is easy with REW).
Last edited:
This is getting very off-topic. I would assume that unless money is very tight we are talking about modern polypropylene film capacitors. 250V types are compact enough and easy to work with. But, TBH, 50-100V electrolytics will probably work well enough.
Let's reiterate that for the 8 ohm Decca London ribbon, this is a workable 1.6kHz circuit:
I think you could use 3X 10uF there. Put two 10uF in parallel to get 20uF. The coil could be a convenient 0.56mH or 0.68mH aircoil.
The bass circuit will wire up like this in the bottom section:
But because we think the DK1 bass driver is a 15 ohm 12" unit, I have suggested 6.8mH coil and 6.8uF capacitor. You could use a ferrite coil, or an aircoil.
We think the horn should be negative polarity reference the bass. It certainly won't be worse than the sound from 50 years ago. And nothing should break if you are careful with the volume control.
Let's reiterate that for the 8 ohm Decca London ribbon, this is a workable 1.6kHz circuit:
I think you could use 3X 10uF there. Put two 10uF in parallel to get 20uF. The coil could be a convenient 0.56mH or 0.68mH aircoil.
The bass circuit will wire up like this in the bottom section:
But because we think the DK1 bass driver is a 15 ohm 12" unit, I have suggested 6.8mH coil and 6.8uF capacitor. You could use a ferrite coil, or an aircoil.
We think the horn should be negative polarity reference the bass. It certainly won't be worse than the sound from 50 years ago. And nothing should break if you are careful with the volume control.
Many thanks. While I am quite happy carrying out straightforward repairs on valve amps - I have never thought about the characteristics of x/o components, so please humour me when I double check I am buying the right bits!!
From the Impact Audio site - parts 5011, 5031, 3810, 5229
and reverse the polarity at the tweeter?
Cheers - Roger
From the Impact Audio site - parts 5011, 5031, 3810, 5229
and reverse the polarity at the tweeter?
Cheers - Roger
One x-over made and fitted. Took me back - it was probably in the 60s I last re-jigged a circuit diagram for stripboard! I have been repairing old valve amps for years, ancient enough to be wired point to point, so it was interesting. Anyway, the results are very promising - vocals have come forward just that important little bit. In fact Sheryl Crow climbed out the speaker and whispered in my ear - that was nice.
Question - are these rectangular capacitors a design statement of some kind - or is there an electronic justification for them? I have seen them in pics, but until now, only worked with cylindrical ones.
Many thanks for all the advice - specially Steve - good circuit!
Question - are these rectangular capacitors a design statement of some kind - or is there an electronic justification for them? I have seen them in pics, but until now, only worked with cylindrical ones.
Many thanks for all the advice - specially Steve - good circuit!