Given that my speaker rebuild fantasy is gradually taking shape, I'm
starting to think about what to do about crossovers. The plan is to bi-amp,
one amp to the B139 drivers, the other to the B110/T27. I'd rather ditch the
orginal KEF DN12 units and make new ones for the mid/top, probably keeping
to the spec crossover frequency. So, I have some questions about
components...

Am I right in thinking that air-cored inductors are better than those with
some sort of ferrite(?) core? For the same value, would an air-cored version
be physically larger? If so, is that because it requires more turns?

What is the best type of capacitor to use wthin sensible cost limits?


--
Wally
www.artbywally.com
www.wally.myby.co.uk/music

In article , Wally
wrote:
Given that my speaker rebuild fantasy is gradually taking shape, I'm
starting to think about what to do about crossovers. The plan is to
bi-amp, one amp to the B139 drivers, the other to the B110/T27. I'd
rather ditch the orginal KEF DN12 units and make new ones for the
mid/top, probably keeping to the spec crossover frequency. So, I have
some questions about components...

I don't know the details of the crossover networks in question. However if
you are going to bi-amp the system it may make sense to remove some/all the
networks between power amps and speakers and, instead, use networks at the
inputs to the power amps. Done correctly, this may well give much better
results that networks that have to work at high powers in the complex
loads.

Am I right in thinking that air-cored inductors are better than those
with some sort of ferrite(?) core?

Depends upon your definition of 'best' and the circumstances.

Ferrite cores mean you can reduce the length of wire required. This can
mean a physically smaller coil and/or a lower coil resistance and/or lower
resistance due to internal impedance, etc. Hence the ferrite may improve
things on the conductor side. However it may, itself, increase magnetic
'singing' of the coil and/or introduce some non-linearity. (That said, it
may well be that the speaker units introduce more distortion than the use
of a suitable ferrite.) Sizes and shapes also affect the amount of coupling
from/to a coil and surrounding components. Ferrite non-linearity or
saturation may not matter much in some circumstances, but make ferrites a
poor choice in other circumstances.

So, "It depends"... :-)

For the same value, would an air-cored version be physically larger?

Probably.

If so, is that because it requires more turns?

and/or a larger cross-sectional area.

What is the best type of capacitor to use wthin sensible cost limits?

Answer similar to the above. :-)

Are you wishing to compare bipolar electrolytics with other types, or are
you excluding them?

The advantage of using crossover/equalising networks between pre and power
amps is that you can avoid a lot of the above problems as the required
voltage and current levels in the components will be much lower. This means
you can use smaller, cheaper, and better perfoming coils/caps.

The disadvantage is that producing the correct network may involve a fair
bit of work as it won't be a simple 'copy' of the speaker arrangement.

Slainte,

Jim

--
Electronics http://www.st-and.ac.uk/~www_pa/Scot...o/electron.htm
Audio Misc http://www.st-and.demon.co.uk/AudioMisc/index.html
Armstrong Audio http://www.st-and.demon.co.uk/Audio/armstrong.html
Barbirolli Soc. http://www.st-and.demon.co.uk/JBSoc/JBSoc.html

In article ,
Wally wrote:
Given that my speaker rebuild fantasy is gradually taking shape, I'm
starting to think about what to do about crossovers. The plan is to
bi-amp, one amp to the B139 drivers, the other to the B110/T27. I'd
rather ditch the orginal KEF DN12 units and make new ones for the
mid/top, probably keeping to the spec crossover frequency. So, I have
some questions about components...

Am I right in thinking that air-cored inductors are better than those
with some sort of ferrite(?) core? For the same value, would an
air-cored version be physically larger? If so, is that because it
requires more turns?

What is the best type of capacitor to use wthin sensible cost limits?

If you are going to bi-amp, it makes much more sense to design the
crossover to go before the amplifiers - it's much easier to produce
filters using ICs than large inductors and capacitors.

--
*Just give me chocolate and nobody gets hurt

Dave Plowman London SW 12
RIP Acorn

In message , Wally
writes
Given that my speaker rebuild fantasy is gradually taking shape, I'm
starting to think about what to do about crossovers. The plan is to bi-amp,
one amp to the B139 drivers, the other to the B110/T27. I'd rather ditch the
orginal KEF DN12 units and make new ones for the mid/top, probably keeping
to the spec crossover frequency. So, I have some questions about
components...

Am I right in thinking that air-cored inductors are better than those with
some sort of ferrite(?) core? For the same value, would an air-cored version
be physically larger? If so, is that because it requires more turns?

What is the best type of capacitor to use wthin sensible cost limits?


--
Wally
www.artbywally.com
www.wally.myby.co.uk/music




What frequency are you intending to crossover at between the B139 and
the B110? (I guess its around 600Hz or so).

At this frequency the capacitors will be around 20 to 25uF, and the
inductors around 2 to 3mH.

I detest bipolar electrolytics, and its not too difficult to get good
quality film capacitors in 10uF and 4.7uf values to parallel up to get
the value you need. Polypropylene dielectric will be the best (and the
dearest.)

No point going to working voltages over 100V on the caps though.

You could wind (or buy from Wilmslow) air-cored inductors rather than
open-bobbin ferrite ones, but remember: an air-cored inductor will have
a higher resistance than a ferrite-cored one (more turns as the
permeability of air is a lot less than typical ferrites), and the
external field will extend further. People often forget that the
inductors on the crossover will interact via their mutual-inductance if
they are aligned the same way and close enough. Try if possible to have
the axes of inductors that are close oriented orthogonally to each other
to reduce the transformer action.


--
Chris Morriss

Dave Plowman wrote:

If you are going to bi-amp, it makes much more sense to design the
crossover to go before the amplifiers - it's much easier to produce
filters using ICs than large inductors and capacitors.

3-way speaker, but two amps. I'm considering a cabinet which will roll off
the mid naturally, so no low-end cut, just a mid/top crossover. The bass
would have some kind of filter before the amp.


--
Wally
www.artbywally.com
www.wally.myby.co.uk/music

In message , Dave Plowman
writes
In article ,
Wally wrote:
Given that my speaker rebuild fantasy is gradually taking shape, I'm
starting to think about what to do about crossovers. The plan is to
bi-amp, one amp to the B139 drivers, the other to the B110/T27. I'd
rather ditch the orginal KEF DN12 units and make new ones for the
mid/top, probably keeping to the spec crossover frequency. So, I have
some questions about components...

Am I right in thinking that air-cored inductors are better than those
with some sort of ferrite(?) core? For the same value, would an
air-cored version be physically larger? If so, is that because it
requires more turns?

What is the best type of capacitor to use wthin sensible cost limits?

If you are going to bi-amp, it makes much more sense to design the
crossover to go before the amplifiers - it's much easier to produce
filters using ICs than large inductors and capacitors.


Oh Oh! I've just seen that he intends to bi-amp. In which case, ignore
my advice about coils and capacitors in my other post, and follow Dave's
advice. Active filter crossovers before the amps are MUCH more
sensible.

You could even design some 'constant voltage' subtraction type active
crossovers to keep a flat summed amplitude and phase response. (in
theory at least. I fear your venerable drive units will impose a sonic
signature whatever the crossover. Both the B139 and the B110 had some
nasty resonances.)
--
Chris Morriss

Chris Morriss wrote:

What frequency are you intending to crossover at between the B139 and
the B110? (I guess its around 600Hz or so).

It depends on what I do with the bass cabs. If I build one-piece 3-way
units, then the xover would probably stay the same as the spec 400Hz. I'm
considering separating the bass cabs from the mid/top so's I can stash the
big boxes out of the way and just have small stand-mounted speakers that are
easier to move and don't dominate the room. I want to keep two bass cabs,
and they'll probably be positioned either side of the listening position,
wider than the mid/top. There's a good chance that they won't be
symmetrical.

Having no direct experience of it, I don't know how much the directional
bass frequencies would suffer with the speakers spread around, so I'm
tending to the view that a separate cab arrangement would entail lowering
the B139/B110 crossover point, to put more of the directional bass into the
easily-positioned B110s. I have a plus in that there will be two bass cabs,
with left/right separation, but a potential minus in how lower bass I can
get out of the B110s in their proposed enclosures. Ideally, I'd like to have
the B110s roll off without any LC, and have some kind of (adjustable?)
filter before the bass amp to fit the low end to the mid roll off. I need to
learn more about the effect of the mid enclosures before I can come up with
a probable number for a bass/mid crossover frequency.



At this frequency the capacitors will be around 20 to 25uF, and the
inductors around 2 to 3mH.

Yup, in the ballpark for the DN12.


I detest bipolar electrolytics, and its not too difficult to get good
quality film capacitors in 10uF and 4.7uf values to parallel up to get
the value you need. Polypropylene dielectric will be the best (and
the dearest.)

No point going to working voltages over 100V on the caps though.

Okay, I'll troff for polypropylene caps.


You could wind (or buy from Wilmslow)

I'm up for winding them if there's a sensible cost saving (I'll have a look
at Wilmslow's prices).


air-cored inductors rather than
open-bobbin ferrite ones, but remember: an air-cored inductor will
have a higher resistance than a ferrite-cored one (more turns as the
permeability of air is a lot less than typical ferrites), and the
external field will extend further. ...

Can I get around the resistance thing by using thicker wire?


... People often forget that the
inductors on the crossover will interact via their mutual-inductance
if they are aligned the same way and close enough. Try if possible
to have the axes of inductors that are close oriented orthogonally to
each other to reduce the transformer action.

Duly noted. :-)


--
Wally
www.artbywally.com
www.wally.myby.co.uk/music

Chris Morriss wrote:

Oh Oh! I've just seen that he intends to bi-amp. In which case,
ignore my advice about coils and capacitors in my other post, and
follow Dave's advice. Active filter crossovers before the amps are
MUCH more sensible.

As I said to Dave, the B110 and T27 will be fed from one amp, so I still
need a crossover for those.


You could even design some 'constant voltage' subtraction type active
crossovers to keep a flat summed amplitude and phase response. (in
theory at least.

Without really understanding what you're saying, this sounds like the sort
of thing I'm considering for the bass amp, especially if I can adjust it.


I fear your venerable drive units will impose a
sonic signature whatever the crossover. Both the B139 and the B110
had some nasty resonances.)

I don't doubt it for a minute, but replacing them is outwith the budget. :-)
I've had them for years and like the overall tonal character - I've never
felt that they were super-clean-fi, but I do find them to be a smooth
listen. Mid range could come out more - some voices are a bit quiet (maybe a
new crossover can fix that...).


--
Wally
www.artbywally.com
www.wally.myby.co.uk/music

Wally wrote:

Dave Plowman wrote:

If you are going to bi-amp, it makes much more sense to design the
crossover to go before the amplifiers - it's much easier to produce
filters using ICs than large inductors and capacitors.

3-way speaker, but two amps. I'm considering a cabinet which will roll off
the mid naturally, so no low-end cut, just a mid/top crossover. The bass
would have some kind of filter before the amp.

That doesn't sound like a very good solution.
Better to use an active crossover (4th order Linkwitz-Riley) before
the poweramps and make the B110 have a flat response down to
an octave or so below the crossover frequency.

I have a scan of the R50 crossovers from my old B139/B110/T27
project if you are interested.

Why not make replicas of the LS3/5a crossover for your new boxes.
I believe they enjoyed a certain reputation once!

--
Roger.

In message , Wally
writes

Having no direct experience of it, I don't know how much the directional
bass frequencies would suffer with the speakers spread around, so I'm
tending to the view that a separate cab arrangement would entail lowering
the B139/B110 crossover point, to put more of the directional bass into the
easily-positioned B110s. I have a plus in that there will be two bass cabs,
with left/right separation, but a potential minus in how lower bass I can
get out of the B110s in their proposed enclosures. Ideally, I'd like to have
the B110s roll off without any LC, and have some kind of (adjustable?)
filter before the bass amp to fit the low end to the mid roll off. I need to
learn more about the effect of the mid enclosures before I can come up with
a probable number for a bass/mid crossover frequency.



I've got 'WinSpeakerz' on the pc, so if you let me know the box volume I
could give you a reasonable idea of the LF response.

--
Chris Morriss