Bailey TL speakers from WW 1972
malcolm
I am just about to make a pair of these Loud Speakers,
Bailey TL speakers from WW 1972
anybody else out there done the same and how do they sound.
regards malcolm
--
¸.·´¯`·.¸><((((º>.·´¯`·.¸¸.·´¯`·.¸><((((º>.·´¯`·.¸
LED Headlamps and Sea Fishing UK
http://www.geocities.com/malc_hurn
http://groups.yahoo.com/group/seafishinguk
http://www.brockstock.org.uk
tam...@hickorytech.com
malc...@yahoo.com
¸.·´¯`·.¸><((((º>.·´¯`·.¸¸.·´¯`·.¸><((((º>.·´¯`·.¸
Trevor Wilson
"malcolm" <tam...@hickorytech.net> wrote in message
news:ym9%a.8876$Ly2.1...@cletus.bright.net...
> hello folk
> I am just about to make a pair of these Loud Speakers,
> Bailey TL speakers from WW 1972
> anybody else out there done the same and how do they sound.
> regards malcolm
**Excellent! A few points to note, however:
* The original KEF crossovers are appalling. I used the very excellent
Radford crossovers, in my design.
* You should transmission line load the midrange as well (as suggested by
Bailey), by using a tube all the way through the back of the enclosure.
* The KEF T27 tweeter is well past its use by date. I am certain better
devices can now be sourced.
* You MUST use long haired sheep's wool and be prepared to adjust the
filling for the appropriate tuning.
* Use very solid construction. I found big differences after I built my set
with 19mm chipboard and then covered it with benchtop laminate. The laminate
cut resonances dramatically. One of my mates built his from 19mm chipboard,
then covered the whole lot in another layer of chipboard. The difference was
not subtle.
When you do, finally build them and I assume you'll carefully tune them,
you'll have one of the few, genuinely 20Hz capable loudspeakers. They are
very good indeed. Pay careful attention to construction details, stuffing,
etc and you will be rewarded handsomely.
BTW: I used to drive mine with a 300 Watt/channel Marantz Model 500 power
amp. They coped quite nicely. The poor old Radford crossovers did not do so
well. Replacing the horrible bi-polar electros soon sorted that out. If I
had been using KEF crossovers, I would have melted the inductors as well.
--
Trevor Wilson
www.rageaudio.com.au
Frank Johnson
> hello folk
> I am just about to make a pair of these Loud Speakers,
> Bailey TL speakers from WW 1972
> anybody else out there done the same and how do they sound.
> regards malcolm
>
>
Can't help with the TL design but before you start....
You obviously want bass response down to 20Hz but if this is for
domestic (ie home lounge and not stage/auditorium) then why not consider
a sealed box design with "Linkwitz transform" circuit to equalise the
bass response ? http://www.linkwitzlab.com
Using 3 way active crossover, my design using the Peerless 831709 bass
unit in a 58 litre sealed box, acheives a response -3dB at 20Hz. The
box volume is uncritical and there is no tuning, specialised woodwork or
materials.
Electronic crossover PCB's are available as is all the design information.
The only drawback is that you need 6 channel amplification but the
results are well worth the effort.
Just a thought anyway and good luck with your project.
My efforts can be seen here: http://www.fdj.org.uk/loudspeakers
Frank
harrogate
Irrespective of what anyone else has been said here, this is one decision
you will never regret - provided of course you have a room big enough to let
them breathe. Don't expect them to be big, brash, and lively - if you like
rock or its like then these are not the speakers for you. But on the other
hand if you like classical, MoR, organ, or jazz, you will never regret the
effort in building them.
Again irrespective of other comments you do not need a lot to drive them as
they are remarkably efficient - and they produce such a realistic and easy
sound you will find that you tend not to turn them up so much (you'll end up
with holes where your windows were if you do!) When I first had mine (a
colleague built them but then got married and bought a small terraced house
so they had to go) my 'hi-fi' was an Ecko ZU4G music centre (remember them?)
It was a 25W amp but of very good design with a superbly flat response -
using a Shure M75EDII cartridge you just bathed in the music. Over the years
I used a Leak Stereo 70 (both Mk1 and MkII versions,) a home-built MOSFET
power amp driven by the front end of the Stereo 70 but later a Toshiba
pre-amp and by a Quad 33 (both original in form and with my own amps in it.)
They all produced MUSIC, not hi-fi!
The thing that was most noticable was the smooth, easy, natural but
frighteningly deep bass. I used to have a Denon sampler/demonstration CD
which had a track of the organ of Limburg Cathedral on it. I never heard
before and have never heard since ANY speaker that could reproduce the pedal
reeds as they actually sound and not just a boomy grunt.
I did two things over the years to them. First I fitted Coles 4001G super
tweeters with the additional components bought as a kit from (the then)
Wilmslow Audio. This did much to improve and brighten the top end that the
T27's had difficulty handling. I also bi-wired them and even though others
will dispute it (go away Pinky - comments not required here) even my wife
commented on the improvement!
The nearest commercial equivalent for size and constriction was the IMF
TLS80 - the nearest electrically and sound-wise was the Cambridge R50 but
she wouldn't let me keep them as they were even bigger!
I still regret the day they had to go but management gave me the ultimatum.
I replaced them with Spendor BC1's which, despite being BBC monitors, were
much 'warmer' and nothing like so easy on the ear. I still have those but
now use KEF Q55's which are much brighter and so better suited to our room.
The cabinets went to the tip, but I still have the drive units just in case
one day.................
--
Woody
malcolm
"Trevor Wilson" <tre...@rageaudio.com.au> wrote in message
news:3f3d...@news.comindico.com.au...
hello Trevor
thanks for replying and the info.
I most likely will be using 3/4" chipboard, but at our nearest DIY depot
they have MDF and
other interesting wood sheets, Oak and Lauran faced MDFs and similar.
is MDF better or worse than chipboard for LS enclosures?
also I have read on a few websites about people using Miraflex fibreglass
insulation,
this would be easier and cheaper to find here in the USA than long fibre
wool.
I intended to use full range Loudspeakers for the moment.
and maybe scale up the enclosures by 33% all sides, depends how my Wife
reacts ;)
malcolm
"harrogate" <harr...@ntlworld.com> wrote in message
news:O%o%a.12886$Kx1.1...@newsfep4-glfd.server.ntli.net...
hello Woody
Classical and folk and vocal music mostly.
at the last Company I worked at, a photocopy of Baileys 1972 Wireless World
article was floating around.
the Technical Director had built a pair and swore they were magic.
and I like the simple elegant way they are built.
and I am sure they can be tweaked to sound ok no matter how they turn out.
regards malcolm
malcolm
news:bhkti6$nkj$1$8302...@news.demon.co.uk...
thanks Frank
I will put pictures on my website when I start.
Trevor Wilson
"harrogate" <harr...@ntlworld.com> wrote in message
news:O%o%a.12886$Kx1.1...@newsfep4-glfd.server.ntli.net...
>
> news:ym9%a.8876$Ly2.1...@cletus.bright.net...
> > hello folk
> > I am just about to make a pair of these Loud Speakers,
> > Bailey TL speakers from WW 1972
> > anybody else out there done the same and how do they sound.
> > regards malcolm
> >
> >
> > --
> > ¸.·´¯`·.¸><((((º>.·´¯`·.¸¸.·´¯`·.¸><((((º>.·´¯`·.¸
> > LED Headlamps and Sea Fishing UK
> > http://www.geocities.com/malc_hurn
> > http://groups.yahoo.com/group/seafishinguk
> > http://www.brockstock.org.uk
> > tam...@hickorytech.com
> > malc...@yahoo.com
> > ¸.·´¯`·.¸><((((º>.·´¯`·.¸¸.·´¯`·.¸><((((º>.·´¯`·.¸
> >
> >
>
> Irrespective of what anyone else has been said here, this is one decision
> you will never regret - provided of course you have a room big enough to
let
> them breathe. Don't expect them to be big, brash, and lively - if you like
> rock or its like then these are not the speakers for you. But on the other
> hand if you like classical, MoR, organ, or jazz, you will never regret the
> effort in building them.
**Utter nonsense. Like any quality speaker, the KEF T-lines are equally at
home with rock, chamber music, or jazz. Back in the 1970s, I pounded mine
with rock every day. When I wanted to impress my more conservative friends,
I played Also Srach Zarathustra (Zubin Mehta, Decca version) and blew them
away with the organ stuff.
>
> Again irrespective of other comments you do not need a lot to drive them
as
> they are remarkably efficient - and they produce such a realistic and easy
> sound you will find that you tend not to turn them up so much (you'll end
up
> with holes where your windows were if you do!)
**More nonsense. They are (slightly) below average efficiency. Mine measured
87dB/Watt. Adequate, when operated with a (say) 100 Watt/ch amp. I used to
rattle the windows with my 300 watt/ch Marantz, but that is another story.
When I first had mine (a
> colleague built them but then got married and bought a small terraced
house
> so they had to go) my 'hi-fi' was an Ecko ZU4G music centre (remember
them?)
> It was a 25W amp but of very good design with a superbly flat response -
> using a Shure M75EDII cartridge you just bathed in the music. Over the
years
> I used a Leak Stereo 70 (both Mk1 and MkII versions,) a home-built MOSFET
> power amp driven by the front end of the Stereo 70 but later a Toshiba
> pre-amp and by a Quad 33 (both original in form and with my own amps in
it.)
> They all produced MUSIC, not hi-fi!
**On that, we agree. Back when I had mine, a mate popped in and had a
listen. After a few minutes, he proclaimed that they lacked bass (??!!),
we're as clear and clean in the mids and highs, etc, etc. After two hours of
listening, he announced: "Geez, you can really live with these speakers,
can't you?" They did not sound as big and splashy as his Infinity 2.5s, but
they provided good, clean sound quality, with a genuine low bass capability.
And, naturally, MOSFET amplifiers are the worst possible choice for T-lines
(and any other speaker).
--
Trevor Wilson
www.rageaudio.com.au
Trevor Wilson
>
>
> hello Trevor
> thanks for replying and the info.
> I most likely will be using 3/4" chipboard, but at our nearest DIY depot
> they have MDF and
> other interesting wood sheets, Oak and Lauran faced MDFs and similar.
> is MDF better or worse than chipboard for LS enclosures?
**MDF should be better.
> also I have read on a few websites about people using Miraflex fibreglass
> insulation,
> this would be easier and cheaper to find here in the USA than long fibre
> wool.
**Bailey was quite specific on the wool thing. I have never heard a
successful pair of Bailey T-lines using anything but wool.
> I intended to use full range Loudspeakers for the moment.
> and maybe scale up the enclosures by 33% all sides, depends how my Wife
> reacts ;)
> regards malcolm
**Don't. Stick to Bailey's design. It does not need to be any bigger. IMO,
you need more solid enclosure walls, with the dimensions provided by Bailey.
A larger enclosure would require much more substantial enclosure walls than
19mm MDF.
--
Trevor Wilson
www.rageaudio.com.au
harrogate
Each to their own opinion - you have yours and I have mine, so is there any
need to be so bombastic?
The only comment I would make is that if you were driving they as described
I am surprised that you did not do damage - with the KEF crossover they were
system rated at 50W, although I seem to recall it was higher (100W?) with
the Radford crossover.
We will have to agree to differ over MOSFET designs. If they were good
enough for John Lindsey-Hood and Doug Self they were and are good enough for
me.
--
Woody
Chris Morriss
<tre...@rageaudio.com.au> writes
>**On that, we agree. Back when I had mine, a mate popped in and had a
>listen. After a few minutes, he proclaimed that they lacked bass (??!!),
>we're as clear and clean in the mids and highs, etc, etc. After two hours of
>listening, he announced: "Geez, you can really live with these speakers,
>can't you?" They did not sound as big and splashy as his Infinity 2.5s, but
>they provided good, clean sound quality, with a genuine low bass capability.
>And, naturally, MOSFET amplifiers are the worst possible choice for T-lines
>(and any other speaker).
>
>
Aah, you follow Doug Self's view on MOSFETs do you? I can't agree.
Douglas Self for all his virtues seems to have a bee in his bonnet about
MOSFETs, mainly as a result of some questionable SPICE simulations.
Mosfets don't make good source followers, (although the modern
'low-threshold' ones work fairly well).
They do make superb CFP output stages though, both in the real world and
in SPICE simulations.
--
Chris Morriss
Stewart Pinkerton
<tre...@rageaudio.com.au> wrote:
>They did not sound as big and splashy as his Infinity 2.5s, but
>they provided good, clean sound quality, with a genuine low bass capability.
>And, naturally, MOSFET amplifiers are the worst possible choice for T-lines
>(and any other speaker).
Bollocks, some of the finest amps ever made have been MOSFET designs
(the big Adcoms, for one example).
--
Stewart Pinkerton | Music is Art - Audio is Engineering
Stewart Pinkerton
<tam...@hickorytech.net> wrote:
>I most likely will be using 3/4" chipboard, but at our nearest DIY depot
>they have MDF and
>other interesting wood sheets, Oak and Lauran faced MDFs and similar.
>is MDF better or worse than chipboard for LS enclosures?
MDF is significantly superior to chipboard, and laminated MDF is even
better due to the added 'constrained layer' damping. OTOH, there's a
thread running on rec.audio.high-end which pretty universally agrees
that homebuilt speakers just don't cut it on sound quality against
commercial designs.
>also I have read on a few websites about people using Miraflex fibreglass
>insulation,
>this would be easier and cheaper to find here in the USA than long fibre
>wool.
Perhaps so, but Bailey used long fibre wool for a good acoustic reason
- it has excellent self-damping and produces an adiabatic
compression/rarefaction state in the line. Also, stick to the design
dimensions, if you make the panels bigger then they will be much more
resonant. Oh, by the way - hi there Harrogate - forget biwiring!
Chris Morriss
<pat...@dircon.co.uk> writes
'Holofill' acrylic fibre is a good damping medium for transmission lines
as well, and certainly better than the plastic soft foam that my old
TLS50s had in. Admittedly not as good as wool, but easier to get hold
of. (Take a few old Holofill pillows apart!)
Bailey's TL suffered from having a right-angle bend close to the rear of
the bass unit. I wonder if a narrower, but deeper design with a 45
degree reflection fillet might stop the higher bass frequencies being
reflected back through the cone?
Compared with an IB design, TL speakers have much lower internal
pressures, so I wouldn't have thought that panel resonance was that much
of an issue. (I've only ever had the TLS50s so I can't compare with
other TL models)
--
Chris Morriss
malcolm
> And, naturally, MOSFET amplifiers are the worst possible choice for
T-lines
> (and any other speaker).
what about class 'D' audio amps?
they work very well in that application
harrogate
"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
news:3f3f88b1...@news.fsnet.co.uk...
Thanks Pinky.
--
Woody
Trevor Wilson
"Chris Morriss" <cr...@oroboros.demon.co.uk> wrote in message
news:9h+XFzIvB5P$Ew...@oroboros.demon.co.uk...
**Wool is worth the effort to track down.
>
> Bailey's TL suffered from having a right-angle bend close to the rear of
> the bass unit. I wonder if a narrower, but deeper design with a 45
> degree reflection fillet might stop the higher bass frequencies being
> reflected back through the cone?
**Bailey suggested this in his design notes.
>
> Compared with an IB design, TL speakers have much lower internal
> pressures, so I wouldn't have thought that panel resonance was that much
> of an issue. (I've only ever had the TLS50s so I can't compare with
> other TL models)
**That was what I thought too. I was wrong. Cabinet strength and rigidity is
very important.
--
Trevor Wilson
www.rageaudio.com.au
Trevor Wilson
>
>
> Each to their own opinion - you have yours and I have mine, so is there
any
> need to be so bombastic?
**When I see myth perpetrated as truth: Yes. The efficiency of the KEF
T-lines is simply not as bad as what has been reported. The T-lines do
deliver very CLEAN, low bass. Perhaps this is why people IMAGINE them to be
low efficiency.
>
> The only comment I would make is that if you were driving they as
described
> I am surprised that you did not do damage - with the KEF crossover they
were
> system rated at 50W, although I seem to recall it was higher (100W?) with
> the Radford crossover.
**I DID damage my Radford crossovers. I replaced the bi-polar electros with
film caps. With the air core inductors of the Radfords and suitably rated
caps, the power handling is enormous. Just like any decent speaker.
>
> We will have to agree to differ over MOSFET designs. If they were good
> enough for John Lindsey-Hood and Doug Self they were and are good enough
for
> me.
**Good for you. Think on this:
* MOSFETs require ten times the bias current, to deliver approximately
equivalent levels of distortion as BJTs.
* MOSFETs are around double the cost, for the same peak current ability of
BJTs (Complementary types, of course).
* Modern BJTs demonstrate better linearity than MOSFETs.
On the plus side of the balance sheet, MOSFETs require about 5% less cost in
protection systems, as is required by BJTs. MOSFETs can also operate at far
higher speeds. An advantage which is wasted in audio applications.
--
Trevor Wilson
www.rageaudio.com.au
Trevor Wilson
> <tre...@rageaudio.com.au> wrote:
>
> >They did not sound as big and splashy as his Infinity 2.5s, but
> >they provided good, clean sound quality, with a genuine low bass
capability.
> >And, naturally, MOSFET amplifiers are the worst possible choice for
T-lines
> >(and any other speaker).
>
> Bollocks, some of the finest amps ever made have been MOSFET designs
> (the big Adcoms, for one example).
**The Adcom I have on the bench uses BJTs. Having said that, I'm sure you
are aware that:
* MOSFETs require more bias current to remain as linear as BJTs.
* Modern BJTs are more linear than MOSFETs.
* MOSFETs are around double the cost of BJTs, for the same peak current
ability (complementary types, of course).
MOSFETs do require approximately 5% fewer Dollars spent on protection
systems, however.
--
Trevor Wilson
www.rageaudio.com.au
Trevor Wilson
**For frequencies up to around 1kHz, they're great. When Class D amps reach
5MHz switching frequencies, I'll get excited. Until then, they'll remain a
curiosity to me.
--
Trevor Wilson
www.rageaudio.com.au
Stewart Pinkerton
<tre...@rageaudio.com.au> wrote:
>
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:3f3f88b1...@news.fsnet.co.uk...
>> On Sat, 16 Aug 2003 19:52:28 GMT, "Trevor Wilson"
>> <tre...@rageaudio.com.au> wrote:
>>
>> >They did not sound as big and splashy as his Infinity 2.5s, but
>> >they provided good, clean sound quality, with a genuine low bass
>capability.
>> >And, naturally, MOSFET amplifiers are the worst possible choice for
>T-lines
>> >(and any other speaker).
>>
>> Bollocks, some of the finest amps ever made have been MOSFET designs
>> (the big Adcoms, for one example).
>
>**The Adcom I have on the bench uses BJTs.
OK, I was referring to the Nelson Pass designs such as the 4802.
>Having said that, I'm sure you
>are aware that:
>
>* MOSFETs require more bias current to remain as linear as BJTs.
Yes - so what? Many folks *like* the idea of a degree of Class A
operation.
>* Modern BJTs are more linear than MOSFETs.
Yes - so what? Ancient power triodes are more linear still.........
>* MOSFETs are around double the cost of BJTs, for the same peak current
>ability (complementary types, of course).
Yes - so what? It's the power supply that costs the big bucks in any
decent power amp.
>MOSFETs do require approximately 5% fewer Dollars spent on protection
>systems, however.
Yes - so what?
Trevor Wilson
> On Mon, 18 Aug 2003 03:13:51 GMT, "Trevor Wilson"
> <tre...@rageaudio.com.au> wrote:
>
> >
> >"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
> >news:3f3f88b1...@news.fsnet.co.uk...
> >> On Sat, 16 Aug 2003 19:52:28 GMT, "Trevor Wilson"
> >> <tre...@rageaudio.com.au> wrote:
> >>
> >> >They did not sound as big and splashy as his Infinity 2.5s, but
> >> >they provided good, clean sound quality, with a genuine low bass
> >capability.
> >> >And, naturally, MOSFET amplifiers are the worst possible choice for
> >T-lines
> >> >(and any other speaker).
> >>
> >> Bollocks, some of the finest amps ever made have been MOSFET designs
> >> (the big Adcoms, for one example).
> >
> >**The Adcom I have on the bench uses BJTs.
>
> OK, I was referring to the Nelson Pass designs such as the 4802.
**Name dropper. Pass was using BJTs back in the days when MOSFETs provided
superior linearity to BJTs. Now, BJTs are more linear than MOSFETs, he uses
MOSFETs. Curious. IMagine how much better his amps would be, if he used
lower distortion output devices....
>
> >Having said that, I'm sure you
> >are aware that:
> >
> >* MOSFETs require more bias current to remain as linear as BJTs.
>
> Yes - so what? Many folks *like* the idea of a degree of Class A
> operation.
**It's all about Dollars. BJTs do it cheaper and better. For a given amount
of Class A power, BJTs offer lower distortion levels. MOSFETs start behind
BJTs and they finish behind them. Higher levels of Class A bias, lower the
distortion levels of BJTs still further.
>
> >* Modern BJTs are more linear than MOSFETs.
>
> Yes - so what? Ancient power triodes are more linear still.........
**For a given amount of NFB, BJTs offer lower distortion levels. In fact,
their linearity is arguably better than most Triodes.
>
> >* MOSFETs are around double the cost of BJTs, for the same peak current
> >ability (complementary types, of course).
>
> Yes - so what? It's the power supply that costs the big bucks in any
> decent power amp.
**Dollars are Dollars. If a manufacturer can shave 20 Squid off the price of
an amp, at the manufacturing point, then around 100 Squid saving can be
passed on to the consumer. And further, MOSFETs require higher Voltages,
then BJTs, when driving low impedance loads. Thus, the power supply costs
more for the MOSFET amp.
>
> >MOSFETs do require approximately 5% fewer Dollars spent on protection
> >systems, however.
>
> Yes - so what?
**I was throwing you a bone. BJTs ain't perfect. They're just a whole lot
better than MOSFETs.
Oh yeah, one more thing: MOSFETs are not available in perfectly mirror
imaged devices. BJTs are. N and P channel MOSFETs are fundamentally
different devices. NPN and PNP BJTs are remarkably similar. Not a problem
for amplifiers which use large amounts of bias current and huge lumps of
NFB. However, for those amps which use low levels of bias current (99.99% of
them) and those amps which don't use much Global NFB (about 0.1% of them),
BJTs may make a much better choice.
--
Trevor Wilson
www.rageaudio.com.au
Old Fart at Play
Will you two stop arguing and get back to the subject?
Any opinions about the relative benefits of a TL and a bass reflex box
for the B139? In theory a bass reflex should give a flat response down
to ~30Hz whereas a TL is way down at that frequency. In practice most
bass reflex speakers sound like a mad drummer is playing on an oil
drum along with the music, almost in time. :-)
Roger.
Trevor Wilson
"Old Fart at Play" <roger....@talk21.com> wrote in message
news:3F40A06A...@talk21.com...
**I've never heard a B139 in a CORRECTLY aligned reflex enclosure. All the
reflex enclosures using B139s sounded dreadful. Particularly the Concertos.
In a T-line, they DO go down to 20Hz, as long as the line is suitably long.
I measured mine, in room, with a 1/3rd octave analyser and they were
satisfyingly flat to 20Hz. Just as predicted by Bailey. Bailey designed his
T-line to use the room gain at LF to provide a flat response. If I get some
time, I'll do a simulation in LEAP for the B139, with various reflex boxes.
LEAP is at least as accurate as required. LEAP's accuracy exceeds the unit
to unit variations with B139 drivers.
--
Trevor Wilson
www.rageaudio.com.au
Jim Lesurf
<tre...@rageaudio.com.au> wrote:
> * MOSFETs require more bias current to remain as linear as BJTs. *
> Modern BJTs are more linear than MOSFETs. * MOSFETs are around double
> the cost of BJTs, for the same peak current ability (complementary
> types, of course).
> MOSFETs do require approximately 5% fewer Dollars spent on protection
> systems, however.
My well-established tendency to being 'old fashioned' comes out here. :-)
I just use fuses in the d.c. powerlines as 'protection' and then ensure the
amp can survive mistreatments like being driven into a short with no damage
beyond blown fuses. If nothing else, having to replace fuses is a good
tutorial excercise for the user in knowing the limits of their equipment,
and the need to treat it properly. ;->
In my case, in the cobwebbed past when I designed poweramps, FETs were not
really up to the required currents and operating areas, so I used bipolars.
My tendency, though, is to leave the choice of device type to the
individual designer, and just judge the result as a system. I'd tend to use
bipolars as I'm familiar with them, but if someone else uses FETs I'd
assume they can get good results if they know what they are doing.
Slainte,
Jim
--
Electronics http://www.st-and.ac.uk/~www_pa/Scots_Guide/intro/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
Stewart Pinkerton
<tre...@rageaudio.com.au> wrote:
>
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:3f407a8a...@news.fsnet.co.uk...
>> On Mon, 18 Aug 2003 03:13:51 GMT, "Trevor Wilson"
>> <tre...@rageaudio.com.au> wrote:
>>
>> >
>> >"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>> >news:3f3f88b1...@news.fsnet.co.uk...
>> >> On Sat, 16 Aug 2003 19:52:28 GMT, "Trevor Wilson"
>> >> <tre...@rageaudio.com.au> wrote:
>> >>
>> >> >They did not sound as big and splashy as his Infinity 2.5s, but
>> >> >they provided good, clean sound quality, with a genuine low bass
>> >capability.
>> >> >And, naturally, MOSFET amplifiers are the worst possible choice for
>> >T-lines
>> >> >(and any other speaker).
>> >>
>> >> Bollocks, some of the finest amps ever made have been MOSFET designs
>> >> (the big Adcoms, for one example).
>> >
>> >**The Adcom I have on the bench uses BJTs.
>>
>> OK, I was referring to the Nelson Pass designs such as the 4802.
>
>**Name dropper. Pass was using BJTs back in the days when MOSFETs provided
>superior linearity to BJTs. Now, BJTs are more linear than MOSFETs, he uses
>MOSFETs. Curious. IMagine how much better his amps would be, if he used
>lower distortion output devices....
Well, all his own 'Pass Labs' amps use MOSFETs, and rumour has it that
a complementary pair of MOSFETs operating in class A has the lowest
distortion of *any* output topology.
>> >Having said that, I'm sure you
>> >are aware that:
>> >
>> >* MOSFETs require more bias current to remain as linear as BJTs.
>>
>> Yes - so what? Many folks *like* the idea of a degree of Class A
>> operation.
>
>**It's all about Dollars. BJTs do it cheaper and better. For a given amount
>of Class A power, BJTs offer lower distortion levels. MOSFETs start behind
>BJTs and they finish behind them. Higher levels of Class A bias, lower the
>distortion levels of BJTs still further.
See above................
Not that there's any real-world *sonic* advantage over a decent Class
AB BJT design, of course.................
>> >* Modern BJTs are more linear than MOSFETs.
>>
>> Yes - so what? Ancient power triodes are more linear still.........
>
>**For a given amount of NFB, BJTs offer lower distortion levels. In fact,
>their linearity is arguably better than most Triodes.
No argument at all - it isn't. OTOH, we all know how this trick is
done, so you certainly wouldn't go building a complete low-distortion
amp with triodes!
>> >* MOSFETs are around double the cost of BJTs, for the same peak current
>> >ability (complementary types, of course).
>>
>> Yes - so what? It's the power supply that costs the big bucks in any
>> decent power amp.
>
>**Dollars are Dollars. If a manufacturer can shave 20 Squid off the price of
>an amp, at the manufacturing point, then around 100 Squid saving can be
>passed on to the consumer. And further, MOSFETs require higher Voltages,
>then BJTs, when driving low impedance loads. Thus, the power supply costs
>more for the MOSFET amp.
Component cost of course has *nothing* to do with what the market will
bear for say a Mark Levinson 'Red Rose' amp - which just happens to be
a rebadged mainland Chinese jobby.....................
>> >MOSFETs do require approximately 5% fewer Dollars spent on protection
>> >systems, however.
>>
>> Yes - so what?
>
>**I was throwing you a bone. BJTs ain't perfect. They're just a whole lot
>better than MOSFETs.
Nah, they're much slower and they have a smaller SOA.
>Oh yeah, one more thing: MOSFETs are not available in perfectly mirror
>imaged devices. BJTs are. N and P channel MOSFETs are fundamentally
>different devices. NPN and PNP BJTs are remarkably similar. Not a problem
>for amplifiers which use large amounts of bias current and huge lumps of
>NFB. However, for those amps which use low levels of bias current (99.99% of
>them) and those amps which don't use much Global NFB (about 0.1% of them),
>BJTs may make a much better choice.
So use a little more bias and little more NFB. It ain't rocket science
with the parts we have in 2003!
Stewart Pinkerton
<roger....@talk21.com> wrote:
>Stewart Pinkerton wrote:
>
>> On Mon, 18 Aug 2003 03:13:51 GMT, "Trevor Wilson"
>> <tre...@rageaudio.com.au> wrote:
>
>>>MOSFETs do require approximately 5% fewer Dollars spent on protection
>>>systems, however.
>>>
>>
>> Yes - so what?
>>
>Will you two stop arguing and get back to the subject?
We're having fun, and Trev is easy to wind up on amp topologies, being
a devotee of weird stuff like 'zero NFB'.
>Any opinions about the relative benefits of a TL and a bass reflex box
>for the B139?
Already done that - forget it, go buy a good modern speaker.
> In theory a bass reflex should give a flat response down
>to ~30Hz whereas a TL is way down at that frequency.
In theory, it depends on the driver, and especially on the bass
alignment of the reflex cabinet. For the B139, remember that the
Bailey 'Transmission Line' is actually a well-damped reflex. If it
wasn't, the end of the line wouldn't have to be open. See the B&W
Nautilus for a *real* TL design. Bottom line is that the Bailey
cabinet gives excellent results with this venerable driver, but it
can't hold a candle to the NHT 1259 in a 4-6 cu.ft sealed box.
> In practice most
>bass reflex speakers sound like a mad drummer is playing on an oil
>drum along with the music, almost in time. :-)
In theory *and* in practice, it all depends on how you set the bass
alignment. A sealed box can be a lot boomier than a reflex if you do
it wrong. Anyone ever heard the big reflex ATCs accused of being
boomy? Or the B&W N800?
Chris Morriss
<roger....@talk21.com> writes
There is no such thing as a bass-reflex loudspeaker with a flat
response. It defies the laws of physics. A bass-reflex is a Helmholtz
resonator for God's sake.
(Yes I know that you can theoretically tune one to give a flat response,
but only on days of the week without a 'Y' in them.
A transmission line, if it is long enough and properly damped is a
speaker to give real bass. Not 'bass' as it is heard emanating from a
Vauxhall Nova.
--
Chris Morriss
Chris Morriss
<pat...@dircon.co.uk> writes
>In theory *and* in practice, it all depends on how you set the bass
>alignment. A sealed box can be a lot boomier than a reflex if you do
>it wrong. Anyone ever heard the big reflex ATCs accused of being
>boomy? Or the B&W N800?
I've never heard the N800, but the big ATCs are underdamped. Sure, a
sealed box is dreadful at a Q of above about 0.75, just listen to the
LS3/5s. These were never hi-fi speakers. Adequate monitors if you were
siting only a metre away from them perhaps!
(Mind you, my TLS50s were very underdamped until I'd put extra damping
in the line.)
Perhaps it's just me, I prefer speakers with a Qtc at around 0.5 which
means that they take up a lot of space.
--
Chris Morriss
Chris Morriss
<tre...@rageaudio.com.au> writes
>
The latest Class-D chips from TI are not far off that.
--
Chris Morriss
Trevor Wilson
**Rumours schmoomers. Let's deal with facts.
>
> >> >Having said that, I'm sure you
> >> >are aware that:
> >> >
> >> >* MOSFETs require more bias current to remain as linear as BJTs.
> >>
> >> Yes - so what? Many folks *like* the idea of a degree of Class A
> >> operation.
> >
> >**It's all about Dollars. BJTs do it cheaper and better. For a given
amount
> >of Class A power, BJTs offer lower distortion levels. MOSFETs start
behind
> >BJTs and they finish behind them. Higher levels of Class A bias, lower
the
> >distortion levels of BJTs still further.
>
> See above................
>
> Not that there's any real-world *sonic* advantage over a decent Class
> AB BJT design, of course.................
**Of course.
>
> >> >* Modern BJTs are more linear than MOSFETs.
> >>
> >> Yes - so what? Ancient power triodes are more linear still.........
> >
> >**For a given amount of NFB, BJTs offer lower distortion levels. In fact,
> >their linearity is arguably better than most Triodes.
>
> No argument at all - it isn't. OTOH, we all know how this trick is
> done, so you certainly wouldn't go building a complete low-distortion
> amp with triodes!
**It would seem to be an exercise in folly.
>
> >> >* MOSFETs are around double the cost of BJTs, for the same peak
current
> >> >ability (complementary types, of course).
> >>
> >> Yes - so what? It's the power supply that costs the big bucks in any
> >> decent power amp.
> >
> >**Dollars are Dollars. If a manufacturer can shave 20 Squid off the price
of
> >an amp, at the manufacturing point, then around 100 Squid saving can be
> >passed on to the consumer. And further, MOSFETs require higher Voltages,
> >then BJTs, when driving low impedance loads. Thus, the power supply costs
> >more for the MOSFET amp.
>
> Component cost of course has *nothing* to do with what the market will
> bear for say a Mark Levinson 'Red Rose' amp - which just happens to be
> a rebadged mainland Chinese jobby.....................
**Strawman argument. The rest of the audio world IS concerned about
manufacturing costs. If a product can be built at lower cost, it is usually
sold at a lower price. A tiny minority of products are sold at a
'what-the-market-will-bear' model.
>
> >> >MOSFETs do require approximately 5% fewer Dollars spent on protection
> >> >systems, however.
> >>
> >> Yes - so what?
> >
> >**I was throwing you a bone. BJTs ain't perfect. They're just a whole lot
> >better than MOSFETs.
>
> Nah, they're much slower and they have a smaller SOA.
**Their speed is irrelevant in audio. Modern, high speed power BJTs possess
an fT in ecessof 20MHz. Acheiving a 3dB point of (say) 200kHz, is a trivial
excercise. Their SOA limitations are known and decent designers allow for
it.
>
> >Oh yeah, one more thing: MOSFETs are not available in perfectly mirror
> >imaged devices. BJTs are. N and P channel MOSFETs are fundamentally
> >different devices. NPN and PNP BJTs are remarkably similar. Not a problem
> >for amplifiers which use large amounts of bias current and huge lumps of
> >NFB. However, for those amps which use low levels of bias current (99.99%
of
> >them) and those amps which don't use much Global NFB (about 0.1% of
them),
> >BJTs may make a much better choice.
>
> So use a little more bias and little more NFB. It ain't rocket science
> with the parts we have in 2003!
**A LOT more bias (around 10 times the level of a BJT amp). In any case, by
doing so, you require more heat sinking. MOSFETs cost more to implement.
Products using MOSFETs cost more to buy, for a specific level of
performance. If you increase the bias and NFB of a BJT amp, it's performance
increases still further.
--
Trevor Wilson
www.rageaudio.com.au
Trevor Wilson
**Really? What frequency do they switch at?
--
Trevor Wilson
www.rageaudio.com.au
Chris Morriss
>
>"Chris Morriss" <cr...@oroboros.demon.co.uk> wrote in message
>news:LqYqnKEJ5RQ$Ew...@oroboros.demon.co.uk...
>> In message <3f4044bc$1...@news.comindico.com.au>, Trevor Wilson
>> <tre...@rageaudio.com.au> writes
>> >
>> >"malcolm" <tam...@hickorytech.net> wrote in message
>> >news:%oN%a.8993$Ly2.1...@cletus.bright.net...
>> >>
>> >> > And, naturally, MOSFET amplifiers are the worst possible choice for
>> >> T-lines
>> >> > (and any other speaker).
>> >>
>> >> what about class 'D' audio amps?
>> >> they work very well in that application
>> >
>> >**For frequencies up to around 1kHz, they're great. When Class D amps
>reach
>> >5MHz switching frequencies, I'll get excited. Until then, they'll remain
>a
>> >curiosity to me.
>> >
>> >
>>
>> The latest Class-D chips from TI are not far off that.
>
>**Really? What frequency do they switch at?
>
>
We use them in Class-D public address amps that we design. I'll check
up tomorrow and post the answer on here.
--
Chris Morriss
Chris Morriss
<m...@privacy.net> writes
>On Mon, 18 Aug 2003 19:42:01 +0100, Chris Morriss used
><+rE3nlD53RQ$Ew...@oroboros.demon.co.uk> to say...
>
>>Perhaps it's just me, I prefer speakers with a Qtc at around 0.5 which
>>means that they take up a lot of space.
>
>
>You'd have had a Hi-Q then :)
>
>Boom Boom.
>
>Sorry, couldn't resist :)
>
>
Oh heavens Kurt, that went right over my head the first time I read it!
Luckily my brain caught on the second time!
--
Chris Morriss
Chris Morriss
<tre...@rageaudio.com.au> writes
>
>**Their speed is irrelevant in audio. Modern, high speed power BJTs possess
>an fT in ecessof 20MHz. Acheiving a 3dB point of (say) 200kHz, is a trivial
>excercise. Their SOA limitations are known and decent designers allow for
>it.
An open loop 3dB before you close the NFB loop? The open loop -3dB
point of the output stage of an amp using overall NFB needs to be up at
a few MHz to ensure unconditional stability. Just model an amp as two
stages, with the first main gain stage as an integrator and you might be
surprised how wide band the output stage has to be to ensure that you
have plenty of phase margin when you close the loop. Very difficult to
achieve with transistors. It's eye-opening to SPICE model some
published circuits of BJT amps and see how marginally stable they are.
If it is a closed loop -3dB point, well that's neither here nor there.
If you think an amp needs a closed loop bandwidth of 200kHz that's your
business.
>**A LOT more bias (around 10 times the level of a BJT amp). In any case, by
>doing so, you require more heat sinking. MOSFETs cost more to implement.
>Products using MOSFETs cost more to buy, for a specific level of
>performance. If you increase the bias and NFB of a BJT amp, it's performance
>increases still further.
>
>
No! As Doug Self has pointed out many times. For a particular class-B
output stage there is one, and only one optimal bias setting.
--
Chris Morriss
Trevor Wilson
> On Mon, 18 Aug 2003 10:46:18 +0100, Old Fart at Play
> <roger....@talk21.com> wrote:
>
> >Stewart Pinkerton wrote:
> >
> >> On Mon, 18 Aug 2003 03:13:51 GMT, "Trevor Wilson"
> >> <tre...@rageaudio.com.au> wrote:
> >
> >>>MOSFETs do require approximately 5% fewer Dollars spent on protection
> >>>systems, however.
> >>>
> >>
> >> Yes - so what?
> >>
> >Will you two stop arguing and get back to the subject?
>
> We're having fun, and Trev is easy to wind up on amp topologies, being
> a devotee of weird stuff like 'zero NFB'.
**Yeah, like Stew said: "We're having fun." For the record: I am not a
devotee of "zero NFB" stuff. All amplifiers use NFB. Every single one.I AM,
however, a devotee of PROPERLY designed and implemented Zero GLOBAL NFB
amplifiers.
>
> >Any opinions about the relative benefits of a TL and a bass reflex box
> >for the B139?
>
> Already done that - forget it, go buy a good modern speaker.
**I am inclined to agree. Those Peerless XLS drivers are pretty damned nice.
>
> > In theory a bass reflex should give a flat response down
> >to ~30Hz whereas a TL is way down at that frequency.
>
> In theory, it depends on the driver, and especially on the bass
> alignment of the reflex cabinet. For the B139, remember that the
> Bailey 'Transmission Line' is actually a well-damped reflex. If it
> wasn't, the end of the line wouldn't have to be open.
**Read the original notes from Bailey. He states, quite clearly, that the
line can be left closed.
See the B&W
> Nautilus for a *real* TL design. Bottom line is that the Bailey
> cabinet gives excellent results with this venerable driver, but it
> can't hold a candle to the NHT 1259 in a 4-6 cu.ft sealed box.
**Agreed. Or the Peerless XLS in a 30 Litre box. Are you guys still using
those clunky Imperial measurements? Must be a nightmare trying to deal with
Europe. I thought that only a piddling 5% of the planet's population still
used Imperial measurements.
--
Trevor Wilson
www.rageaudio.com.au
Old Fart at Play
> There is no such thing as a bass-reflex loudspeaker with a flat
> response. It defies the laws of physics. A bass-reflex is a Helmholtz
> resonator for God's sake.
What horrible stinky bait.
> (Yes I know that you can theoretically tune one to give a flat response,
> but only on days of the week without a 'Y' in them.
Indeed. In fact a long wool sock shoved in the port is often enough
to tame a Hi-Q speaker.
> A transmission line, if it is long enough and properly damped is a
> speaker to give real bass. Not 'bass' as it is heard emanating from a
> Vauxhall Nova.
A Nova would be about the right size for a decent reflex enclosure.
Roger.
Trevor Wilson
> In message <3f41...@news.comindico.com.au>, Trevor Wilson
> <tre...@rageaudio.com.au> writes
> >
> >**Their speed is irrelevant in audio. Modern, high speed power BJTs
possess
> >an fT in ecessof 20MHz. Acheiving a 3dB point of (say) 200kHz, is a
trivial
> >excercise. Their SOA limitations are known and decent designers allow for
> >it.
>
> An open loop 3dB before you close the NFB loop?
**Points:
* You're assuming a feedback loop. Such things are not necessary for high
fidelity performance. Local feedback, correclty applied can allow a BJT to
amp to deliver inaudible levels of distortion, with very high levels of
stability.
* A closed loop response of 200kHz is simple enough to acheive with modern
BJTs.
The open loop -3dB
> point of the output stage of an amp using overall NFB needs to be up at
> a few MHz to ensure unconditional stability. Just model an amp as two
> stages, with the first main gain stage as an integrator and you might be
> surprised how wide band the output stage has to be to ensure that you
> have plenty of phase margin when you close the loop. Very difficult to
> achieve with transistors. It's eye-opening to SPICE model some
> published circuits of BJT amps and see how marginally stable they are.
**For stuff designed back in the 1970s, maybe.
>
> If it is a closed loop -3dB point, well that's neither here nor there.
> If you think an amp needs a closed loop bandwidth of 200kHz that's your
> business.
**Sure. As long as the phase response in the audible band is flat, then the
BW is unimportant. 50kHz is probably fine.
>
> >**A LOT more bias (around 10 times the level of a BJT amp). In any case,
by
> >doing so, you require more heat sinking. MOSFETs cost more to implement.
> >Products using MOSFETs cost more to buy, for a specific level of
> >performance. If you increase the bias and NFB of a BJT amp, it's
performance
> >increases still further.
> >
> >
> No! As Doug Self has pointed out many times. For a particular class-B
> output stage there is one, and only one optimal bias setting.
**Does he say (precisely) that? Or does he suggest that Class A designs have
more distortion than a properly designed Class A/B design? In any case, I
am inlcined to agree. Once the bias current is set at an appropriate level,
more bias current does very little, short of adding more heat into the
troposphere.
--
Trevor Wilson
www.rageaudio.com.au
Dave Plowman
Chris Morriss <cr...@oroboros.demon.co.uk> wrote:
> Sure, a
> sealed box is dreadful at a Q of above about 0.75, just listen to the
> LS3/5s. These were never hi-fi speakers. Adequate monitors if you were
> siting only a metre away from them perhaps!
Err, that's what they are designed specifically for - sitting on top of in
a very small room where you won't get any real bass extension anyway.
Doesn't stop them sounding rather better than many in 'normal' use,
though.
--
*It doesn't take a genius to spot a goat in a flock of sheep *
Dave Plowman dave....@argonet.co.uk London SW 12
RIP Acorn
Stewart Pinkerton
<tre...@rageaudio.com.au> wrote:
>
>"Stewart Pinkerton" <pat...@dircon.co.uk> wrote in message
>news:3f40fec6...@news.fsnet.co.uk...
>> On Mon, 18 Aug 2003 10:46:18 +0100, Old Fart at Play
>> <roger....@talk21.com> wrote:
>> >Will you two stop arguing and get back to the subject?
>>
>> We're having fun, and Trev is easy to wind up on amp topologies, being
>> a devotee of weird stuff like 'zero NFB'.
>
>**Yeah, like Stew said: "We're having fun." For the record: I am not a
>devotee of "zero NFB" stuff. All amplifiers use NFB. Every single one.I AM,
>however, a devotee of PROPERLY designed and implemented Zero GLOBAL NFB
>amplifiers.
See what I mean? :-)
Stewart Pinkerton
<cr...@oroboros.demon.co.uk> wrote:
>There is no such thing as a bass-reflex loudspeaker with a flat
>response. It defies the laws of physics. A bass-reflex is a Helmholtz
>resonator for God's sake.
Clearly, you don't understand how to design speakers.........
Get a copy of Vance Dickasons Cookbook, and find out how it's done.
Jim Lesurf
<cr...@oroboros.demon.co.uk> wrote:
> We use them in Class-D public address amps that we design. I'll check
> up tomorrow and post the answer on here.
Do you have any spectra of their output extending from LF up to and beyond
their nominal switching frequency? I have seen some spectra on items like
the 'Tact' power amps, but never with a properly defined set of scales that
can allow the power spectral density to be determined. The plots published
by Paul Miller in HFN are essentially without meaning in terms of power
spectral density. I asked him about this a while ago, but got no meaninful
reply. :-/
Jim Lesurf
<cr...@oroboros.demon.co.uk> wrote:
> >
> No! As Doug Self has pointed out many times. For a particular class-B
> output stage there is one, and only one optimal bias setting.
In that case we may all be doomed. :-) My own impression is that the
actual bias wanders about in practice, and that the 'optimal' value may
depend upon what the amp is doing at any moment. :-)
Trevor Wilson
"Jim Lesurf" <jc...@st-and.demon.co.uk> wrote in message
news:4c2453a...@st-and.demon.co.uk...
> In article <UYtN77Be6SQ$Ew...@oroboros.demon.co.uk>, Chris Morriss
> <cr...@oroboros.demon.co.uk> wrote:
>
>
> > We use them in Class-D public address amps that we design. I'll check
> > up tomorrow and post the answer on here.
>
> Do you have any spectra of their output extending from LF up to and beyond
> their nominal switching frequency? I have seen some spectra on items like
> the 'Tact' power amps, but never with a properly defined set of scales
that
> can allow the power spectral density to be determined. The plots published
> by Paul Miller in HFN are essentially without meaning in terms of power
> spectral density. I asked him about this a while ago, but got no meaninful
> reply. :-/
**The early model Tact was one of those chronically flawed (but very
expensive) designs, which the reviewers fell about the place raving how good
the thing was. A great pity nobody thought to measure what the thing was
doing into a real loudspeaker.
--
Trevor Wilson
www.rageaudio.com.au
Chris Morriss
>
>"Chris Morriss" <cr...@oroboros.demon.co.uk> wrote in message
>news:LqYqnKEJ5RQ$Ew...@oroboros.demon.co.uk...
>> In message <3f4044bc$1...@news.comindico.com.au>, Trevor Wilson
>> <tre...@rageaudio.com.au> writes
>> >
>> >"malcolm" <tam...@hickorytech.net> wrote in message
>> >news:%oN%a.8993$Ly2.1...@cletus.bright.net...
>> >>
>> >> > And, naturally, MOSFET amplifiers are the worst possible choice for
>> >> T-lines
>> >> > (and any other speaker).
>> >>
>> >> what about class 'D' audio amps?
>> >> they work very well in that application
>> >
>> >**For frequencies up to around 1kHz, they're great. When Class D amps
>reach
>> >5MHz switching frequencies, I'll get excited. Until then, they'll remain
>a
>> >curiosity to me.
>> >
>> >
>>
>> The latest Class-D chips from TI are not far off that.
>
>**Really? What frequency do they switch at?
>
>
I was wrong about the TI chips we use, they still have a maximum clock
of 300kHz. It was another company that are looking at much higher
effective clock rates with a direct Delta-Sigma feed.
See www.imec.be/esscirc/esscirc2001/proceedings/data/62.pdf for an
interesting paper on the technique.
There is also www.tripath.com as well.
--
Chris Morriss
Chris Morriss
<pat...@dircon.co.uk> writes
>On Mon, 18 Aug 2003 19:36:34 +0100, Chris Morriss
><cr...@oroboros.demon.co.uk> wrote:
>
>>There is no such thing as a bass-reflex loudspeaker with a flat
>>response. It defies the laws of physics. A bass-reflex is a Helmholtz
>>resonator for God's sake.
>
>Clearly, you don't understand how to design speakers.........
>
>Get a copy of Vance Dickasons Cookbook, and find out how it's done.
I've got it, and a copy of WinSpeakerz for my PC and I still hold to my
point. In the fourth edition of Dickason's book there isn't a single
graph of any so-called reflex 'alignment' that shows an acceptable
response. They are all underdamped. (They have to be really to make a
reflex speaker have any bass extension over that of an IB of the same
volume.)
Can you really live with the phase problems that a 24dB per octave bass
roll-off gives?
I'm not one to talk though, as my current Magneplanar 2.5R speakers
(dipoles) can be very 'bloomy' if not well away from any walls.
Consider a simple (but accurate) model of a bass-reflex loudspeaker:
Hang a weight on a spring. This will have a natural resonant frequency
and is analogous to the compliance of the air in a reflex l/s,
resonating with the mass of air in the port. Now take a second weigh
and attach it to the top of the spring, and hang the whole lot on
another spring.
Now move the new hanging point up and down at various frequencies and
see how many frequencies there are at which the two weights are moving
in phase.
The two weights are the cone of the l/s and the mass of air in the port.
It's an acoustic nightmare!
We're just about to buy a 4133 mic and the MLSSA system at work so I
might once again be able to get my hands on some decent measuring
equipment.
--
Chris Morriss
Chris Morriss
<cr...@oroboros.demon.co.uk> writes
>I was wrong about the TI chips we use, they still have a maximum clock
>of 300kHz. It was another company that are looking at much higher
>effective clock rates with a direct Delta-Sigma feed.
>See www.imec.be/esscirc/esscirc2001/proceedings/data/62.pdf for an
>interesting paper on the technique.
>
>There is also www.tripath.com as well.
Rats! The www.imec.be isn't working here! It was OK from work today.
(it was a very interesting paper from ST-Micro)
--
Chris Morriss
Jim Lesurf
<tre...@rageaudio.com.au> wrote:
Having read the review of one in a recent HFN I would not rush to disagree
with you! :-)
The problem, though, is that although I have some worries about the amp,
the reviews I have seen give no really useful data that could be used to
resolve the doubts I have.
Stewart Pinkerton
<cr...@oroboros.demon.co.uk> wrote:
>In message <3f41caf4....@news.fsnet.co.uk>, Stewart Pinkerton
><pat...@dircon.co.uk> writes
>>On Mon, 18 Aug 2003 19:36:34 +0100, Chris Morriss
>><cr...@oroboros.demon.co.uk> wrote:
>>
>>>There is no such thing as a bass-reflex loudspeaker with a flat
>>>response. It defies the laws of physics. A bass-reflex is a Helmholtz
>>>resonator for God's sake.
>>
>>Clearly, you don't understand how to design speakers.........
>>
>>Get a copy of Vance Dickasons Cookbook, and find out how it's done.
>
>I've got it, and a copy of WinSpeakerz for my PC and I still hold to my
>point. In the fourth edition of Dickason's book there isn't a single
>graph of any so-called reflex 'alignment' that shows an acceptable
>response. They are all underdamped. (They have to be really to make a
>reflex speaker have any bass extension over that of an IB of the same
>volume.)
Ah well, that depends on why you're using a reflex alignment, now
doesn't it? A reflex design does *not* have to be underdamped, it just
turns out that people *like* that slightly boomy sound. I happen to
know that ATC revoiced the SCM100 from the dead flat of the intial
production run, because the customers thought it 'lacked bass'.
>Can you really live with the phase problems that a 24dB per octave bass
>roll-off gives?
Yes. Can you live with system Q above 0.6?
>I'm not one to talk though, as my current Magneplanar 2.5R speakers
>(dipoles) can be very 'bloomy' if not well away from any walls.
Try mounting them at 45 degrees to the wall - see my page on
http://www.apah69.dsl.pipex.com/
>Consider a simple (but accurate) model of a bass-reflex loudspeaker:
>
>Hang a weight on a spring. This will have a natural resonant frequency
>and is analogous to the compliance of the air in a reflex l/s,
>resonating with the mass of air in the port. Now take a second weigh
>and attach it to the top of the spring, and hang the whole lot on
>another spring.
>
>Now move the new hanging point up and down at various frequencies and
>see how many frequencies there are at which the two weights are moving
>in phase.
>
>The two weights are the cone of the l/s and the mass of air in the port.
>It's an acoustic nightmare!
Now add resistive damping to the springs, to bring the system Q to
less than 0.7, and it's an acoustic dream..............
Chris Morriss
<pat...@dircon.co.uk> writes
>
>Now add resistive damping to the springs, to bring the system Q to
>less than 0.7, and it's an acoustic dream..............
Quite possibly. The old 'drinking straws' in the port did a good job of
adding resistive damping. Haven't seen resistive loading in the port
for many years though.
--
Chris Morriss
Stewart Pinkerton
Ah yes, Questor................ :-)
Mind you, nowadays many manufacturers supply foam bungs which do much
the same job, but are removable if you prefer an open port.