Nath wrote:
Lots of info here about 1124p, it is used for LF EQ'ing..

http://www.snapbug.ws/bfd.htm

Looks interesting - I'll have a better read later, but I'm currently feeling
that the 8024 might be the better (more transparent?) bit of kit, if the
specs are anything to go by.


if you want more advanced PEQ for full-range, also checkout 2496.

http://www.behringer.com/DEQ2496/index.cfm?lang=ENG

Too pricey. :-)


--
Wally
www.artbywally.com/FiatPandaRally/index.htm
www.wally.myby.co.uk

In article , Wally
wrote:
Jim Lesurf wrote:

Also, if you are simply splitting the links I assume the internal LP
and HP filters in the speakers are still in the signal paths. Thus the
'best' setting for your active crossover/filter may be quite different
to that specified for the speaker when used 'normally'.

No links as such - 3-way speakers with the bass drivers disconnected
from the passive crossovers and wired straight to the bass amp. The
remaining two-way pair are still connected to a single amp, going
through the full 3-way crossover (with the LF output open circuit).

IIUC the above means that the HF signals now go through the same network as
before, but with the LF speaker disconnected from that network.

Two comments on that:

1) If the LF speaker was previously connected via that network, it may now
be producing a different frequency response than before.

2) Removing the impedance of the LF speaker from the network may also
change the response of the network/HF speakers.

Hence in addition to other factors, the changes indicated above may change
the frequency response of the system.

I'm thinking of converting to the two-way Kef crossover for splitting
the B110 and T27 (this version has no high-pass element on the B110 cct,
since the B110 is acting as bass and mid in a two-way set up). The B110
is in its own IB sub-enclosure within the main cabinet. Do you think
this is worth pursuing, or would there be some sort of 'conflict' of
B110 roll-off in the sub-enclosure, compared with the 24dB/octave
low-pass that the active crossover applies to the bass driver?

Not sure what might be meant by 'conflict'. However the B110 and its
enclosure presumable have a given frequency response and set of
resonances, etc. Just driving directly from a flat amp with a LP roll-off
may not deal with this. Hard to say as I'm afraid I don't know the
specifics of the actual arrangement. Rolling away the HF will avoid things
like cone breakup resonances. However in the region where both the LF and
HF speakers are active their outputs will combine and this will also have
effects that will be quite specific to the individual arrangement.

The above is a long-winded way of daying "Dunno". :-)

TBH though, personally I would not embark on speaker development/design
without having access to some kit to measure the response, etc. There are
simply too many variables and potential problems which I'd find too hard to
disentagle simply 'by ear'.

If you can make up CDs of test sinewaves, etc, and have access to a
reliable sound pressure meter, then it would be interesting to see to
what extent the differences you hear correlate with any changes in
overall frequency response being produced by the active crossover,
differences in gain of the two power amps, and effects of the output
impedances of the amps.

Urrr..., I can make test CDs of sine waves. :-) (With a bit of software
I had on my old computer - would need to find it and reinstall.) Don't
have a sound pressure meter, I'm afraid. I suppose the Cyrus 2 has more
gain than the valve amp. I think their input sensitivities are roughly
similar (200-300mV for full power). The valve amp (20-ish Watts) is set
to half volume, the Cyrus (50W) at about '3', and the gains/cuts on the
active crossover are all set to 0dB. This seems to produce a balanced
sound. The overall volume is controlled by a preamp.

The problem here is not just that the two power amps may have different
gains. They may also have different output impedances. In particular if the
valve amp has a 'high' output impedance this will produce a different
frequency response to an amp with a 'low' output impedance.

It's all calibrated with the indubitably dubious power of the human
lug-'ole - but it's my lug-'ole, and that's the one that matters. :-)
My impression of the Behringer is that it's essentially flat - without a
set of before and after measurements, I'm assuming that the feeling of
better clarity and improved bass is down to having more power available,
and perhaps splitting bass and mid/top to separate amps.

Alas, I have my doubts that the reason you assume is the correct one.
Unless the amps were seriously current limiting I doubt that the biamping
will have had very much effect on the power levels available. Could you
hear previously clear signs of clipping?

The problem is that the changes you perceive may simply be due to you
altering the overall frequency response. This seems particularly likely
(and your assumption unlikely) if you hear the benefits at moderate to low
sound levels as well as at peak levels.

If the results sound better to you, that is fair enough as you have a
result you prefer. For that your lug-ole is king.

However it means that if that *is* the reason, then similar results might
be obtained by much cheaper and easier means. This may not matter to you at
this point, but if it were the case, knowing this might save others time
and money. :-)

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

"Jim Lesurf" wrote in message
...
In article , Wally
wrote:
Jim Lesurf wrote:
..
It's all calibrated with the indubitably dubious power of the human
lug-'ole - but it's my lug-'ole, and that's the one that matters. :-)
My impression of the Behringer is that it's essentially flat - without a
set of before and after measurements, I'm assuming that the feeling of
better clarity and improved bass is down to having more power available,
and perhaps splitting bass and mid/top to separate amps.

..
If the results sound better to you, that is fair enough as you have a
result you prefer. For that your lug-ole is king.

However it means that if that *is* the reason, then similar results might
be obtained by much cheaper and easier means. This may not matter to you
at
this point, but if it were the case, knowing this might save others time
and money. :-)

I'm not in a position to give any results yet, but I am experimenting along
the lines of Wally's ideas. For me they were a necessity because of the
need to fill a larger room. I was not able to get the power I needed from
a single twin amp so decided to try out using three. I have a Behringer
3400 XO and so far am very pleased. Its not there acoustically. I'm using
a Cyrus preamp feeding the XO which drives two InterMn500W
amps, one for two-way Mordaunt-Short floorstanders for the mid
and one for the sub. I am about to use two tweeters run off a NAD
amplifier in addition to the two in the MSs. The MS top will be rolled off.
I
don't know what crossover frequency I will use. And yes, I will be
assessing by ear. If I don't get what I want after playing I'll hook
up a sig gen and use a level meter to find out what's wrong with the
response curve.

It'll be a week or so. I'll post anything useful that I discover.

Peter Scott

In article , Peter Scott
wrote:


I'm not in a position to give any results yet, but I am experimenting
along the lines of Wally's ideas. For me they were a necessity because
of the need to fill a larger room. I was not able to get the power I
needed from a single twin amp so decided to try out using three.

I have doubts that bi or even tri amping should usually make a big
difference to the sound levels you can obtain. This assumes, though, that:

1) Your initial amp wasn't current limiting due to the parallel loads of
the speakers.

2) You aren't using the active xover to avoid large losses in the initial
passive xover arrangements in the speakers.

If either of the above were the case, then the following comments may not
apply...

For getting the biggest possible increase in peak power you'd have to get a
situation where the vectors (voltages) of the bandsplit signals were all in
phase and of the same sign and amplitude. In this extreme case you might
get a peak improvement of x9 in power. i.e. less than 10dB, which in terms
of audibility tends to be regarded as impying you would never get more than
an apparent doubling in peak level.

However in practice I suspect you will find that almost all the time the
vectors are neither in phase nor of equal size. Thus the increase in
available power is likely to be much less than the above.

FWIW I did some analysis of the peak levels of band-divided musical signals
from CD when looking at clipping and tweeters (some pages on this on the
'Scots Guide' site). Can't be sure, but I'd be surprised if in general you
got much more than a few dB of sound level increase before occasional
clipping. i.e. not much more than the smallest change that is clearly
noticable.

You may, however, get an apparent increase if the frequency response
changes as that may affect the audibility and the audible impression
conveyed. However as with Wally's experiments, if this is the case, then
similar results might be obtained more easily.

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

Jim Lesurf wrote:

IIUC the above means that the HF signals now go through the same
network as before, but with the LF speaker disconnected from that
network.

Yup. Well, mid and high still go through the network, and bass bypasses it.


Two comments on that:

1) If the LF speaker was previously connected via that network, it
may now be producing a different frequency response than before.

Okay. Maybe that would explain why the bass seemed a bit muddy (and a bit
too loud) when the active crossover was set to 400Hz. Would it be fair to
say that the low-end roll off (70Hz downwards) would still be more
influenced by the cabinet/room?


2) Removing the impedance of the LF speaker from the network may also
change the response of the network/HF speakers.

Hence in addition to other factors, the changes indicated above may
change the frequency response of the system.

Righto. Very interesting...




Not sure what might be meant by 'conflict'.

I mean a mismatch in the roll-off of the B110 in its sub-enclosure, compared
with that imposed on the B139 by the active filter (the B110 is set up as a
mid driver and the IB enclosure is quite small - I don't know what the
resonant frequency of the midrange system is, but I suspect it's rather
higher than that of the driver itself, which is 38Hz). Aren't the slopes
supposed to be similar in terms of dB/octave?


However the B110 and its
enclosure presumable have a given frequency response and set of
resonances, etc. Just driving directly from a flat amp with a LP
roll-off may not deal with this. Hard to say as I'm afraid I don't
know the specifics of the actual arrangement. Rolling away the HF
will avoid things like cone breakup resonances. However in the region
where both the LF and HF speakers are active their outputs will
combine and this will also have effects that will be quite specific
to the individual arrangement.

Not sure we're talking about the same thing - the B110 is acting as a
midrange driver. Its LF response is rolled off both by the HP portion of the
passive crossover and the size of its sub-enclosure. The B139 is being
driven by a flat amp with a LP filter before it. The LF signal to the B110
is also being restricted by the bass/mid crossover frequency in the
Behringer. However, switching off the bass amp and tweaking the active xover
freq reveals that it has a little effect on the B110 below about 250-300Hz;
from about 200Hz downward is only very slight. In other words, the
now-redundant bass/mid parts of the passive crossover could be causing the
roll-off on the B110, or its sub-enclosure, or a combnation of both. What
I'm wondering is, would it be a good idea to remove the passive xover
components and have only the sub-enclosure acting to roll off the B110s bass
response? I appreciate that, without knowing the specifics, this might not
be answerable. :-)



TBH though, personally I would not embark on speaker
development/design without having access to some kit to measure the
response, etc. There are simply too many variables and potential
problems which I'd find too hard to disentagle simply 'by ear'.

The way I see it, I can always undo the changes I make. If I don't like a
change, I can revert to the previous configuration.


The problem here is not just that the two power amps may have
different gains. They may also have different output impedances. In
particular if the valve amp has a 'high' output impedance this will
produce a different frequency response to an amp with a 'low' output
impedance.

The spec for the Cyrus 2 quotes 0.08 ohms, can't find a number for the valve
amp. FWIW, I've always felt that both amps have a similar character to the
sound - ie, pretty clean and flat when used within their available headroom.
By contrast, both amps are substantially better than the Arcam Alpha -
they're much closer to each other than either is to the Arcam.


... I'm assuming that
the feeling of better clarity and improved bass is down to having
more power available, and perhaps splitting bass and mid/top to
separate amps.

Alas, I have my doubts that the reason you assume is the correct one.
Unless the amps were seriously current limiting I doubt that the
biamping will have had very much effect on the power levels
available. Could you hear previously clear signs of clipping?

When each was used to amplify the full range, the Cyrus was louder than the
valve amp, and stayed clean for longer. However, they both reached a point
where I felt the sound became unpleasant to listen to. Whether or not that
was clipping, I'm not sure - to me, that term denotes what you get when a
guitar amp is turned up to the point where it sounds like a chainsaw. I
don't let the hifi get anything like that bad, so I'm not sure it's clipping
that I'm hearing, or some other effect. I sometmes wonder if the speakers
themselves were being overloaded, but I don't think this would be likely in
the case of the valve amp - the Cyrus can easily match it in terms of
in-room volume, and stays cleaner as it does so.


The problem is that the changes you perceive may simply be due to you
altering the overall frequency response. This seems particularly
likely (and your assumption unlikely) if you hear the benefits at
moderate to low sound levels as well as at peak levels.

I haven't listened much at low volumes since I added the active crossover
(rather, only had it on as background music and not been paying attention).
Generally, though, I think it still sounds cleaner.


If the results sound better to you, that is fair enough as you have a
result you prefer. For that your lug-ole is king.

However it means that if that *is* the reason, then similar results
might be obtained by much cheaper and easier means. This may not
matter to you at this point, but if it were the case, knowing this
might save others time and money. :-)

Maybe that depends on what is causing the change in frequency response. The
overall aim of this is to add headroom to be sure that the amplification
isn't a factor at high volumes, and to possibly improve the fidelity by
removing the passive crossovers. It started with the plan to double up the
bass drivers by converting them to a pair of isobaric enclosures - it seemed
that the easiest way to account for the change in impedance at the bass end,
but not at the mid or top, was to split the amplification up. I then thought
that, if I'm going to bin the passive crossovers at the bass end, then it
might not hurt to bin them altogether.

It would be fair to say that my curiosity has been piqued, and I'm quite
happy to the follow the tri-amping route just to find out what it sounds
like.


--
Wally
www.artbywally.com/FiatPandaRally/index.htm
www.wally.myby.co.uk

In article , Wally
wrote:
Jim Lesurf wrote:
[snip]


Two comments on that:

1) If the LF speaker was previously connected via that network, it may
now be producing a different frequency response than before.

Okay. Maybe that would explain why the bass seemed a bit muddy (and a
bit too loud) when the active crossover was set to 400Hz. Would it be
fair to say that the low-end roll off (70Hz downwards) would still be
more influenced by the cabinet/room?

Hard to be sure. However the above seems plausible. In general the LF
roll-away at frequencies below 50-100 Hz region does tend to be dominated
by how easily the speaker can produce pressure variations in the room.
Ideally, an infinite baffle speaker in an airtight room with no leaks and
firm walls would allow a response down to dc.! :-) In practice, though, the
LF response will depend on how much air pressure change the room will
experience for a given cone movement once you are below the lowest room
resonance.

FWIW my experience here is limited, but it is that there are two problems
which may combine if you are unlucky. One is the basic room pressure
resonance which will probably be somewhere below 100Hz. The other is the
'mechanical' resonance of the speaker cone mass/compliance which will also
be somewhere below 100Hz. If the two are at a similar frequency they may
produce a particularly severe 'boom', and the bass may fall away rapidly
below this. Which is 'more significant' will depend on the details of your
situation.

With the sub I use in my living room I can measure a change in the peak
frequency and level if I open/close the room door.

IIUC commercial subwoofers employ an amp with a correction network to try
and iron out some of the speaker effects and give a more extended response.
You'd need to experiment to determine what was required for a given system.


Not sure what might be meant by 'conflict'.

I mean a mismatch in the roll-off of the B110 in its sub-enclosure,
compared with that imposed on the B139 by the active filter (the B110 is
set up as a mid driver and the IB enclosure is quite small - I don't
know what the resonant frequency of the midrange system is, but I
suspect it's rather higher than that of the driver itself, which is
38Hz). Aren't the slopes supposed to be similar in terms of dB/octave?

Not sure. The amount of output (sensitivity) will vary in a different way
to the actual drive levels. With HF speakers you need to avoid LF as the
excursions might cause problems. With LF speakers you avoid HF as it can
excite cone resonances. So the degree of attenuation required depends on
different factors as well. However I'm not a speaker designer so don't know
much about the details of practical designs.


However the B110 and its enclosure presumable have a given frequency
response and set of resonances, etc. Just driving directly from a flat
amp with a LP roll-off may not deal with this. Hard to say as I'm
afraid I don't know the specifics of the actual arrangement. Rolling
away the HF will avoid things like cone breakup resonances. However in
the region where both the LF and HF speakers are active their outputs
will combine and this will also have effects that will be quite
specific to the individual arrangement.

Not sure we're talking about the same thing - the B110 is acting as a
midrange driver. Its LF response is rolled off both by the HP portion of
the passive crossover and the size of its sub-enclosure. The B139 is
being driven by a flat amp with a LP filter before it. The LF signal to
the B110 is also being restricted by the bass/mid crossover frequency in
the Behringer. However, switching off the bass amp and tweaking the
active xover freq reveals that it has a little effect on the B110 below
about 250-300Hz; from about 200Hz downward is only very slight. In other
words, the now-redundant bass/mid parts of the passive crossover could
be causing the roll-off on the B110, or its sub-enclosure, or a
combnation of both. What I'm wondering is, would it be a good idea to
remove the passive xover components and have only the sub-enclosure
acting to roll off the B110s bass response? I appreciate that, without
knowing the specifics, this might not be answerable. :-)

Afraid I have to go along with your final comments. :-) I don't know
enough to say for sure.

TBH though, personally I would not embark on speaker
development/design without having access to some kit to measure the
response, etc. There are simply too many variables and potential
problems which I'd find too hard to disentagle simply 'by ear'.

The way I see it, I can always undo the changes I make. If I don't like
a change, I can revert to the previous configuration.

Yes. The snag, though, is the phenomenon I tend to call "lost in space".
This is where there are many variables you can fiddle with, or which alter
out of your control. This means as you change some things you sometimes
don't know what is really happening, or if some other slight change would
be desirable, etc. This is worse when working entirely 'by ear' as your
reactions will vary according to how you feel and your 'recent'
experiences, and affected by the unreliability of memory.

The problem here is not just that the two power amps may have
different gains. They may also have different output impedances. In
particular if the valve amp has a 'high' output impedance this will
produce a different frequency response to an amp with a 'low' output
impedance.

The spec for the Cyrus 2 quotes 0.08 ohms, can't find a number for the
valve amp.

Afraid I don't know any figures for the valve amp you mentioned. However if
you look in reviews, values when quoted are often of the order of 0.5 Ohms
or more. i.e. somewhat larger than the value for your Cyrus,


... I'm assuming that the feeling of better clarity and improved bass
is down to having more power available, and perhaps splitting bass
and mid/top to separate amps.

Alas, I have my doubts that the reason you assume is the correct one.
Unless the amps were seriously current limiting I doubt that the
biamping will have had very much effect on the power levels available.
Could you hear previously clear signs of clipping?

When each was used to amplify the full range, the Cyrus was louder than
the valve amp, and stayed clean for longer. However, they both reached a
point where I felt the sound became unpleasant to listen to. Whether or
not that was clipping, I'm not sure - to me, that term denotes what you
get when a guitar amp is turned up to the point where it sounds like a
chainsaw. I don't let the hifi get anything like that bad, so I'm not
sure it's clipping that I'm hearing, or some other effect. I sometmes
wonder if the speakers themselves were being overloaded, but I don't
think this would be likely in the case of the valve amp - the Cyrus can
easily match it in terms of in-room volume, and stays cleaner as it does
so.

Hard to say, but if the sound becomes unpleasant at high levels then I'd
expect that to be due to some mix of:

1) Amp clipping (voltage) or current limiting

2) Speaker distortions

3) Amp distortion levels rising with output levels

With 'traditional' valve designs (1) and (3) may be factors, but can't
really say without more details.

[snip]

If the results sound better to you, that is fair enough as you have a
result you prefer. For that your lug-ole is king.

However it means that if that *is* the reason, then similar results
might be obtained by much cheaper and easier means. This may not
matter to you at this point, but if it were the case, knowing this
might save others time and money. :-)

Maybe that depends on what is causing the change in frequency response.

Yes. Indeed, there are some good reasons for feeling that an active
crossover and 'direct drive' of the speaker units can be optimum. But this
then may well require the crossover to also do some other alterations to
cater for any inherent variations in sensitivity with frequency of the
speaker units. However it is difficult to say without more specific
measured data.

My concern here is that people may buy relatively expensive active
crossovers and bi- or tru-amp, but then get result which could have been
obtained more easily and cheaply with a slightly better amp and a tweaked
speaker or room acoustic. Hence I am inclined to be wary of this without
suitable measurements, etc.

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

"Jim Lesurf" wrote in message
...
In article , Peter Scott
wrote:


FWIW I did some analysis of the peak levels of band-divided musical
signals
from CD when looking at clipping and tweeters (some pages on this on the
'Scots Guide' site). Can't be sure, but I'd be surprised if in general you
got much more than a few dB of sound level increase before occasional
clipping. i.e. not much more than the smallest change that is clearly
noticable.

You may, however, get an apparent increase if the frequency response
changes as that may affect the audibility and the audible impression
conveyed. However as with Wally's experiments, if this is the case, then
similar results might be obtained more easily.


I don't know if this is relevant, but the Behringer allows you to delay each
frequency band with respect to the others by up to 2 ms. This is intended
to pull speakers back into phase when one is further back than another.
Would this help with the potential problem you described?

I'm waiting for Studio Spares to deliver the connectors before I
can hook up the amp for the tweeters. Gnash, gnash!

Peter Scott

In article , Peter Scott
wrote:

"Jim Lesurf" wrote in message
...
In article , Peter Scott
wrote:


FWIW I did some analysis of the peak levels of band-divided musical
signals
from CD when looking at clipping and tweeters (some pages on this on
the 'Scots Guide' site). Can't be sure, but I'd be surprised if in
general you got much more than a few dB of sound level increase before
occasional clipping. i.e. not much more than the smallest change that
is clearly noticable.

You may, however, get an apparent increase if the frequency response
changes as that may affect the audibility and the audible impression
conveyed. However as with Wally's experiments, if this is the case,
then similar results might be obtained more easily.


I don't know if this is relevant, but the Behringer allows you to delay
each frequency band with respect to the others by up to 2 ms. This is
intended to pull speakers back into phase when one is further back than
another. Would this help with the potential problem you described?

Sorry, I'm not clear which specific problem you are referring to above. Can
you remind me?

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

"Jim Lesurf" wrote in message
...
In article , Peter Scott
wrote:



I don't know if this is relevant, but the Behringer allows you to delay
each frequency band with respect to the others by up to 2 ms. This is
intended to pull speakers back into phase when one is further back than
another. Would this help with the potential problem you described?

Sorry, I'm not clear which specific problem you are referring to above.
Can
you remind me?

OK Wrong word. I meant the fact that you might not get as great an
increase in perceived power as you might expect.

Peter Scott

Jim Lesurf wrote:

Hard to be sure. However the above seems plausible. In general the LF
roll-away at frequencies below 50-100 Hz region does tend to be
dominated by how easily the speaker can produce pressure variations
in the room. Ideally, an infinite baffle speaker in an airtight room
with no leaks and firm walls would allow a response down to dc.! :-)
In practice, though, the LF response will depend on how much air
pressure change the room will experience for a given cone movement
once you are below the lowest room resonance.

Well, I don't have an ideal environment, or speakers, so I think I'm going
to the Behringer digital parametric EQ shop... :-)


FWIW my experience here is limited, but it is that there are two
problems which may combine if you are unlucky. One is the basic room
pressure resonance which will probably be somewhere below 100Hz. The
other is the 'mechanical' resonance of the speaker cone
mass/compliance which will also be somewhere below 100Hz. If the two
are at a similar frequency they may produce a particularly severe
'boom', and the bass may fall away rapidly below this. Which is 'more
significant' will depend on the details of your situation.

The way I see it, the only thing measurements will tell me is whether sound
I like has an even response, or a bumpy one. :-)


With the sub I use in my living room I can measure a change in the
peak frequency and level if I open/close the room door.

That's interesting - I'll bear that in mind for when I start doing bass EQ.


IIUC commercial subwoofers employ an amp with a correction network to
try and iron out some of the speaker effects and give a more extended
response. You'd need to experiment to determine what was required for
a given system.

That's pretty-much the plan. Now that the bass is separate from the rest, I
think my next step is to get the parametric gizmo and try doing some EQ on
the low end. I'd like to see how much improvement can be had from the
present set up, before I look into building new bass cabinets. (Before
someone tells me that I might not need it by then, I would like to mention
that I am already aware of this possibility.)


Not sure. The amount of output (sensitivity) will vary in a different
way to the actual drive levels. With HF speakers you need to avoid LF
as the excursions might cause problems. With LF speakers you avoid HF
as it can excite cone resonances. So the degree of attenuation
required depends on different factors as well. However I'm not a
speaker designer so don't know much about the details of practical
designs.

Okay, Fairy Nuff. :-)


Yes. The snag, though, is the phenomenon I tend to call "lost in
space". This is where there are many variables you can fiddle with,
or which alter out of your control. This means as you change some
things you sometimes don't know what is really happening, or if some
other slight change would be desirable, etc. This is worse when
working entirely 'by ear' as your reactions will vary according to
how you feel and your 'recent' experiences, and affected by the
unreliability of memory.

I appreaciate the somewhat non-scientific approach you're talking about, and
I admit that that's kinda what I'm doing. However, it's not so much about
the small incremental changes, it's the overall change that results from the
whole tri-amping thing. Using the active crossover to split the bass from
the rest of the system is just the first stage. I fully intend to get rid of
the passive crossovers, put the bass drivers into better boxes in an
isobaric configuration, equalise the limitations of the bass system and room
out of the equation as much as I can, and pile on the watts to be sure that
straining amplifiers aren't a factor.

Once it's got to that stage, *then* I ask myself if it's an improvement over
what I started with, establish what limitations there may be, and look into
ways of improving it further if I think that's warranted. I know I'll be
relying on memory of the previous configuration, but I have had these
speakers most of the last 25 years - they're the one thing in my system that
has never changed, so I reckon I'm pretty familiar with their sound. In
other words, aside from the memory aspect, which I think is mitigated, the
thing that I'm seeking to test is the difference between the old set up and
the tri-amped one. It'll be interesting to note the effects of the changes
that are made along the way to get to that point, but it's really the end
result that I want to check out.


Afraid I don't know any figures for the valve amp you mentioned.
However if you look in reviews, values when quoted are often of the
order of 0.5 Ohms or more. i.e. somewhat larger than the value for
your Cyrus,

I never did do an A-B comparison when I got the Cyrus - I just changed out
the valve amp and got on with it. My impression at the time was that the
fidelity of each amp was very similar - very difficult to separate them,
other than the Cyrus being capable of producing more clean volume. I don't
know what effect a higher o/p impedance would likely have.


Maybe that depends on what is causing the change in frequency
response.

Yes. Indeed, there are some good reasons for feeling that an active
crossover and 'direct drive' of the speaker units can be optimum. But
this then may well require the crossover to also do some other
alterations to cater for any inherent variations in sensitivity with
frequency of the speaker units. However it is difficult to say
without more specific measured data.

It strikes me that the basic principle is sound. I can't claim that the
drivers have a flat response in their working bandwidths, but I'm not aware
of the Kef passive crossovers having stuff in them to smooth out any
howlers. I stand to be corrected, but I gather they're basically dividing
networks.


My concern here is that people may buy relatively expensive active
crossovers and bi- or tru-amp, but then get result which could have
been obtained more easily and cheaply with a slightly better amp and
a tweaked speaker or room acoustic. Hence I am inclined to be wary of
this without suitable measurements, etc.

Cost is a factor, but, while I'm not willing to pay audiophile prices, I'm
willing to spend a bit to get a marked improvement over what I see as the
current system's shortcomings, which are primarily messy, uneven bass with
too much roll-off, and not enough clarity when played up loud. There's also
a curiosity factor, in that I want to see what difference is to be had by
eliminating the crossovers and driving the speakers directly.


--
Wally
www.artbywally.com/FiatPandaRally/index.htm
www.wally.myby.co.uk