[west]

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2) woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west

[Patrick Turner]

west wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2) woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west

I hope I don't get into too much trouble answering your post to the
group.

I have a client with infinities who has aked me to rebuild them
because he has noise problems, response problems ( I did a test ),
and amp problems. The speakers have been tweaked by someone years ago
and monster thick cabling is absurdly thick, but has oxidised badly
under the clear
insulation plastic. ( why oh lord do they do this crap? )
Many inductors have iron lamination cores.

But I have no time to attend the project now, but I sympathise with your
concerns.

If you have two 8 ohm speakers in series, you get 16 ohms.
If in parallel, you get 4 ohms.

But in parallel, the change in Z will mean the existing crossover coils
and caps
will be all but useless, and cut off the bass at an F of 1/4 of where it
is now.

dc winding R of coils will be high compared to 4 ohms, so losses will be
higher, however the
sensitivity of bass for a given input voltage will be much greater.
8Vrms into 16 ohms = 4 watts, and maybe you get 93dB SPL,
and 8V into 4 ohms = 16 watts, maybe you get 99dB SPL, less say 1 db for
power lost in dcr.
So unless you re-design the whole speaker circuitry
just paralleling the two bass will put a big trough in the response, and
a peak in low bass,
and I doubt that will improve the sound.

These speakers are impressive speakers when set up well.

Let us suppose you do have the bass speakers in parallel and you place
series-paralel
resistance networks before the midrange and treble units to get a nice
flat response with everything
done just right and with all air core inductors and quality caps.

The box size will still be valid for the bass response.

But the speaker Z will fall to low levels and probably down to maybe 2
ohms
at the bass to midrange Xover pole, so that may strain ****e out of your
amp whatever it is.

The only solution then is to use a toroidal impedance matching tranny
to match 2 ohms to maybe 4, 8 or 16 etc, or as it is possible with
whatever taps exist on the matching tranny.
Another customer of mine invested in a pair of impedance matching
speaker trannies and
said they improved the outcome significantly. He found he didn't need
all the watts ppl say he should have,
and matched the speakers so his amps see a higher number of ohms than
the speaker Z.
This is a "lighter" load which allows a PP amp to remain in class A for
all of its working
but with less maximum power which is limited to about 40% of the total
idle Pda of the output tubes,
so if this = say 100 watts for 4 x 6550, then 40 watts of PURE class A1
is possible in UL, beam tetrode,
or about 30watts in triode.
SE amps will also have less power with a higher loading ohms but there
should be less distortion
and more effectively low Rout.
And just connecting 4 ohms to an amp straight where its Rout = 1 ohm
gives a DF = 4 only,
and less at the Xover pole, so having the amp think its connected to 16
ohms
will improve the DF to 16, or much more because it will work in class A
rather than class AB where
the Rout is always lower than class A.
( The change in Rout between A and AB is responsible for the increase in
3H
witnessed where loads are too low for class A to be maintained to high
levels ).

I tested the trannies which he bought from the US
but which had no box enclosures which were very expensive although nice
looking.
I made cheaper timber boxes, potted the trannies in compacted sand,
and added binding posts able to take 4mm banana plugs for rapid on-site
Z changes.
He can match anything to anything and is happy.
I couldn't seetoo much wrong with the trannies which had a full power
bandwidth
from less than 10Hz to 1MHz.

Methinks such matching trannies would be a godsend for those struggling
to stop
tubes burning out in their OTL amplifiers. a 4 ohm load is usually
hellishly difficult for
any OTL amp, but where the amp can "see" 24 ohms, output power can be
greater than
with 4 ohms, and there will be much more class A, less thd/imd, less
severe pda,
much better DF, and no wonder ppl say a matching tranny with OTL is a
great idea.
( So why omit the OPT in the darn first instance then??...its another
story. )

The transformers I boxed for my customer came from Paul Speltz at
http://www.zeroimpedance.com/

These were not particularly neatly wound as i would have liked,
indicating
they were wound by cheap labour goodness knows where, ( China? Mexico? )
but they worked, and insulation isn't critical
because speaker voltages are low, and heat is minimal.

If you have a GOSS spiral core rated for 500VA mains use,
and about 300turns of 1.5mm dia for the primary and secondary
and with appropriate taps, its very easy to get wide bw and low losses
and THD/IMD.

Ordinary E&I lam trannies also are possible with 2" stack 1.5" tongue
GOSS fully meshed core,
just fill up the bobbin with layers PSPSPSPSP interleaved layers of say
1.6mm dia wire
and vary wind arrangements to suit whatever may be required.

Patrick Turner.

[[email protected]]

On Tue, 13 Feb 2007 12:22:34 GMT, "west" wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2) woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west


First of all, if they are 16 ohms, why do you think they need lots of current?

If you want to drive them with an SS amp, wiring them in parallel is a good
idea, don't forget to remove the crossover circuitry and get an electronic
crossover for the amp. Direct drive is the BEST way to go!

Any decent amp will drive them, make sure you have the required power.

[Patrick Turner]

wrote:

On Tue, 13 Feb 2007 12:22:34 GMT, "west" wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2) woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west


First of all, if they are 16 ohms, why do you think they need lots of current?

If you want to drive them with an SS amp, wiring them in parallel is a good
idea, don't forget to remove the crossover circuitry and get an electronic
crossover for the amp. Direct drive is the BEST way to go!

Any decent amp will drive them, make sure you have the required power.

Hmm, you left a lot out of what else he would have to do eh.

Direct drive of the bass units in parallel is doable, but it won't give
better bass than if the speakers are in series; it may give more bass
only, ie, play louder.

The series connection is reliant on each driver being exactly the same
in its Z characteristic.
This means ech driver is +/- 5% within tolerance of what the average
chaacteristic is, and unless a speaker has a serious defect,
indeed the drivers will be matched well enough. One cannot use different
drivers in series, ( eg, different brands of 12" bass )
or have series speakers where the reflex enclosures behind them are
different volumes for overlapping
peaks in Z for each driver in each volume. ( a nice trick to even out
the peaks in bass speaker Z and obtain impressively
flat Z character for a given bass speaker system. Don't ask me if this
sounds better ).

But if paralleled, and direct driven, then you must have bi-amping or
tri-amping,
and aditional amps are needed for midrange and treble, and active or
passive crossover filters must be designed and constructed to
control the input signal voltages to the amps which are direct connected
to the speakers without
their original passive LCR crossover networks.
let me tell ya, its only possible to achieve good results with a heck of
a lot of work
for which you must be a trained expert to do properly, and if you look
at the very complex Infinity passive Xovers
then any expert would tell ya that reproducing the response with direct
connected amps and
preceeding Xover filters is a real challenge.

But of course such expert re-doing of filters to get a response is how I
earn part of my living,
and usually its not all that hard to improve on the work done by
designers of
commercially mass produced speakers, very few of which survive the mass
production process
and still have the response the designer intended, because capacitor
values may be +/- 100% away from
where they should be, L values are out, and maybe the 2nd and 3rd
attempt to tailor the response to be flat
for the final production spec has never even been attempted.

Nearly all speakers act like having a perfect flat speaker powered with
an amp with de-facto graphic equalizer,
and the many different brands are like having wildly different slide
settings on the equalizer.
This is the reason why different brands sound different; they have
different responses.
And this aspect doesn't include box resonances and box effects....

When the box effects are very much more suppressed below what is normal
for so many boxed speakers, ie, bloody horrible,
and they are made to be +/- 2dB right across the band, its remarkable
how the sound becomes very similar,
providing the power handling is adequate, and filtering is also adequate
to prevent cone break up,
or bass affecting tweeters etc.

Patrick Turner.

[west]

"Patrick Turner" wrote in message
...


west wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you
know,
it demands lots of current. All my tube amps sound great with the
Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2)
woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a
Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them
in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west

I hope I don't get into too much trouble answering your post to the
group.

I have a client with infinities who has aked me to rebuild them
because he has noise problems, response problems ( I did a test ),
and amp problems. The speakers have been tweaked by someone years ago
and monster thick cabling is absurdly thick, but has oxidised badly
under the clear
insulation plastic. ( why oh lord do they do this crap? )
Many inductors have iron lamination cores.

But I have no time to attend the project now, but I sympathise with your
concerns.

If you have two 8 ohm speakers in series, you get 16 ohms.
If in parallel, you get 4 ohms.

But in parallel, the change in Z will mean the existing crossover coils
and caps
will be all but useless, and cut off the bass at an F of 1/4 of where it
is now.

dc winding R of coils will be high compared to 4 ohms, so losses will be
higher, however the
sensitivity of bass for a given input voltage will be much greater.
8Vrms into 16 ohms = 4 watts, and maybe you get 93dB SPL,
and 8V into 4 ohms = 16 watts, maybe you get 99dB SPL, less say 1 db for
power lost in dcr.
So unless you re-design the whole speaker circuitry
just paralleling the two bass will put a big trough in the response, and
a peak in low bass,
and I doubt that will improve the sound.

These speakers are impressive speakers when set up well.

Let us suppose you do have the bass speakers in parallel and you place
series-paralel
resistance networks before the midrange and treble units to get a nice
flat response with everything
done just right and with all air core inductors and quality caps.

The box size will still be valid for the bass response.

But the speaker Z will fall to low levels and probably down to maybe 2
ohms
at the bass to midrange Xover pole, so that may strain ****e out of your
amp whatever it is.

The only solution then is to use a toroidal impedance matching tranny
to match 2 ohms to maybe 4, 8 or 16 etc, or as it is possible with
whatever taps exist on the matching tranny.
Another customer of mine invested in a pair of impedance matching
speaker trannies and
said they improved the outcome significantly. He found he didn't need
all the watts ppl say he should have,
and matched the speakers so his amps see a higher number of ohms than
the speaker Z.
This is a "lighter" load which allows a PP amp to remain in class A for
all of its working
but with less maximum power which is limited to about 40% of the total
idle Pda of the output tubes,
so if this = say 100 watts for 4 x 6550, then 40 watts of PURE class A1
is possible in UL, beam tetrode,
or about 30watts in triode.
SE amps will also have less power with a higher loading ohms but there
should be less distortion
and more effectively low Rout.
And just connecting 4 ohms to an amp straight where its Rout = 1 ohm
gives a DF = 4 only,
and less at the Xover pole, so having the amp think its connected to 16
ohms
will improve the DF to 16, or much more because it will work in class A
rather than class AB where
the Rout is always lower than class A.
( The change in Rout between A and AB is responsible for the increase in
3H
witnessed where loads are too low for class A to be maintained to high
levels ).

I tested the trannies which he bought from the US
but which had no box enclosures which were very expensive although nice
looking.
I made cheaper timber boxes, potted the trannies in compacted sand,
and added binding posts able to take 4mm banana plugs for rapid on-site
Z changes.
He can match anything to anything and is happy.
I couldn't seetoo much wrong with the trannies which had a full power
bandwidth
from less than 10Hz to 1MHz.

Methinks such matching trannies would be a godsend for those struggling
to stop
tubes burning out in their OTL amplifiers. a 4 ohm load is usually
hellishly difficult for
any OTL amp, but where the amp can "see" 24 ohms, output power can be
greater than
with 4 ohms, and there will be much more class A, less thd/imd, less
severe pda,
much better DF, and no wonder ppl say a matching tranny with OTL is a
great idea.
( So why omit the OPT in the darn first instance then??...its another
story. )

The transformers I boxed for my customer came from Paul Speltz at
http://www.zeroimpedance.com/

These were not particularly neatly wound as i would have liked,
indicating
they were wound by cheap labour goodness knows where, ( China? Mexico? )
but they worked, and insulation isn't critical
because speaker voltages are low, and heat is minimal.

If you have a GOSS spiral core rated for 500VA mains use,
and about 300turns of 1.5mm dia for the primary and secondary
and with appropriate taps, its very easy to get wide bw and low losses
and THD/IMD.

Ordinary E&I lam trannies also are possible with 2" stack 1.5" tongue
GOSS fully meshed core,
just fill up the bobbin with layers PSPSPSPSP interleaved layers of say
1.6mm dia wire
and vary wind arrangements to suit whatever may be required.

Patrick Turner.

I appreciate your comprehensive reply, Patrick It's another couple of pages
for my theory folder. With all the posts you made like this, perhaps you
should just collate them and write a practical audio book, seriously.
However, I was thinking more on the lines of keeping it simple, at least for
now by using a 2-way crossover such as the tube Marchland. This way I'll use
the passive part of the Infinity's xrv.. that pertains to the EMIT & Mid.
The active xrv. would feed that (tube amp) and also feed the woofers amp. I
still end up with the woofers' amp looking at 16 ohms. I think you seen the
schematic of the Infinity's xrv a while back. Sorry I didn't make my self
clearer in the OP (is that what you meant by getting in trouble?).
Finally again, what do you think about a professional line of amps, like
Pyramid, driving the woofers only. I think I'm missing a point somewhere
because this approach seems too simple. No ridiculing please. Only Allison
is allowed.

Cordially,
west

[[email protected]]

On Wed, 14 Feb 2007 02:05:52 GMT, Patrick Turner
wrote:



wrote:

On Tue, 13 Feb 2007 12:22:34 GMT, "west" wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2) woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west


First of all, if they are 16 ohms, why do you think they need lots of current?

If you want to drive them with an SS amp, wiring them in parallel is a good
idea, don't forget to remove the crossover circuitry and get an electronic
crossover for the amp. Direct drive is the BEST way to go!

Any decent amp will drive them, make sure you have the required power.

Hmm, you left a lot out of what else he would have to do eh.

Not really... he can use his tube amp to drive the other speakers WITH their
existing crossovers, no reason to change anything there. He can adj. volume,
phase, and bandpass in the SS amp/woofer setup to match. He just needs a decent
electronic crossover. If he can't proceed from here, he shouldn't be doing the
work...

Direct drive of the bass units in parallel is doable, but it won't give
better bass than if the speakers are in series; it may give more bass
only, ie, play louder.

It will be a better match for his SS amp - very few SS amps want to put 100
watts into 16 ohms!

The series connection is reliant on each driver being exactly the same
in its Z characteristic.

Well no.... but I don't recommend it.

This means ech driver is +/- 5% within tolerance of what the average
chaacteristic is, and unless a speaker has a serious defect,
indeed the drivers will be matched well enough. One cannot use different
drivers in series, ( eg, different brands of 12" bass )

sure you can... ask Bose!

or have series speakers where the reflex enclosures behind them are
different volumes for overlapping
peaks in Z for each driver in each volume. ( a nice trick to even out
the peaks in bass speaker Z and obtain impressively
flat Z character for a given bass speaker system. Don't ask me if this
sounds better ).

As you say - different speakers will even themselves out. You are actually
contradicting yourself...

But if paralleled, and direct driven, then you must have bi-amping or
tri-amping,

That was his plan, Stan.

and aditional amps are needed for midrange and treble, and active or
passive crossover filters must be designed and constructed to
control the input signal voltages to the amps which are direct connected
to the speakers without
their original passive LCR crossover networks.

No and no and no... no reason you HAVE to use electronic crossovers on the mids
and tweets. In fact, an amp and electronic filter for a tweeter is a huge waste
of time and money if you aren't playing the Forum...

let me tell ya, its only possible to achieve good results with a heck of
a lot of work
for which you must be a trained expert to do properly, and if you look
at the very complex Infinity passive Xovers
then any expert would tell ya that reproducing the response with direct
connected amps and
preceeding Xover filters is a real challenge.

That's why I didn't tell him to do that...

But of course such expert re-doing of filters to get a response is how I
earn part of my living,
and usually its not all that hard to improve on the work done by
designers of
commercially mass produced speakers, very few of which survive the mass
production process
and still have the response the designer intended, because capacitor
values may be +/- 100% away from
where they should be, L values are out, and maybe the 2nd and 3rd
attempt to tailor the response to be flat
for the final production spec has never even been attempted.

All this great work for people to listen to mp3's of crap bands!


Nearly all speakers act like having a perfect flat speaker powered with
an amp with de-facto graphic equalizer,
and the many different brands are like having wildly different slide
settings on the equalizer.
This is the reason why different brands sound different; they have
different responses.
And this aspect doesn't include box resonances and box effects....

When the box effects are very much more suppressed below what is normal
for so many boxed speakers, ie, bloody horrible,
and they are made to be +/- 2dB right across the band, its remarkable
how the sound becomes very similar,
providing the power handling is adequate, and filtering is also adequate
to prevent cone break up,
or bass affecting tweeters etc.

Patrick Turner.

I've noticed that when reviewing different QUALITY speakers in stores, it's very
very hard to pick which is best. It comes down to picking the best QUIRKS in the
speaker that you can live with! But can one really say one is better? Better
for you, really.

BTW I just finished constructing a sub and satellite speaker system for my own
listening pleasure. I used a 150 watt SS amp for the woofer, a LAB12 in a 2.75'
box. The amp is in the box as well, as it's design. The sats each use an
Eminence 8" and bullet tweeter. The tweets are crossed over passively, and I'm
going to build a high pass filter at between 80 and 120hz for the amp, since I
don't like the overlap with the sub. I haven't decided on the amp yet, I'm using
an old Technics at the moment till I can build a nice class AB 18 watt/ch tube
amp, probably PP EL84. Or I may go to a 60 watt design, since I have some bigger
tubes...

One thing I can say about the bass is it's awesome! That LAB12 is smooth! It's
basically flat down to 25hz.

As for the sats, the passive crossover needs work... I more or less guessed the
design, but I'm going to have to bring the wires out so I can tweak the values
while I listen.

It is a near field sys by the way, for my DAW, but will also be used for the
elec. drums, hence the high power capacity mids! (225 watts RMS)

I've listened with both a 30 watt and a 400 watt amp, and either will do for
most cases. I don't want to go deaf! For non critical listening IE background
music while I work I turn the sub off. The sats have a response similar to other
nearfields like the Tannoys. (60 - 20k)

[Patrick Turner]

west wrote:

"Patrick Turner" wrote in message
...


west wrote:

I like the sound of my Infinity RS-IIIB speakers, but as most of you
know,
it demands lots of current. All my tube amps sound great with the
Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2)
woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a
Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them
in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

Cordially,
west

I hope I don't get into too much trouble answering your post to the
group.

I have a client with infinities who has aked me to rebuild them
because he has noise problems, response problems ( I did a test ),
and amp problems. The speakers have been tweaked by someone years ago
and monster thick cabling is absurdly thick, but has oxidised badly
under the clear
insulation plastic. ( why oh lord do they do this crap? )
Many inductors have iron lamination cores.

But I have no time to attend the project now, but I sympathise with your
concerns.

If you have two 8 ohm speakers in series, you get 16 ohms.
If in parallel, you get 4 ohms.

But in parallel, the change in Z will mean the existing crossover coils
and caps
will be all but useless, and cut off the bass at an F of 1/4 of where it
is now.

dc winding R of coils will be high compared to 4 ohms, so losses will be
higher, however the
sensitivity of bass for a given input voltage will be much greater.
8Vrms into 16 ohms = 4 watts, and maybe you get 93dB SPL,
and 8V into 4 ohms = 16 watts, maybe you get 99dB SPL, less say 1 db for
power lost in dcr.
So unless you re-design the whole speaker circuitry
just paralleling the two bass will put a big trough in the response, and
a peak in low bass,
and I doubt that will improve the sound.

These speakers are impressive speakers when set up well.

Let us suppose you do have the bass speakers in parallel and you place
series-paralel
resistance networks before the midrange and treble units to get a nice
flat response with everything
done just right and with all air core inductors and quality caps.

The box size will still be valid for the bass response.

But the speaker Z will fall to low levels and probably down to maybe 2
ohms
at the bass to midrange Xover pole, so that may strain ****e out of your
amp whatever it is.

The only solution then is to use a toroidal impedance matching tranny
to match 2 ohms to maybe 4, 8 or 16 etc, or as it is possible with
whatever taps exist on the matching tranny.
Another customer of mine invested in a pair of impedance matching
speaker trannies and
said they improved the outcome significantly. He found he didn't need
all the watts ppl say he should have,
and matched the speakers so his amps see a higher number of ohms than
the speaker Z.
This is a "lighter" load which allows a PP amp to remain in class A for
all of its working
but with less maximum power which is limited to about 40% of the total
idle Pda of the output tubes,
so if this = say 100 watts for 4 x 6550, then 40 watts of PURE class A1
is possible in UL, beam tetrode,
or about 30watts in triode.
SE amps will also have less power with a higher loading ohms but there
should be less distortion
and more effectively low Rout.
And just connecting 4 ohms to an amp straight where its Rout = 1 ohm
gives a DF = 4 only,
and less at the Xover pole, so having the amp think its connected to 16
ohms
will improve the DF to 16, or much more because it will work in class A
rather than class AB where
the Rout is always lower than class A.
( The change in Rout between A and AB is responsible for the increase in
3H
witnessed where loads are too low for class A to be maintained to high
levels ).

I tested the trannies which he bought from the US
but which had no box enclosures which were very expensive although nice
looking.
I made cheaper timber boxes, potted the trannies in compacted sand,
and added binding posts able to take 4mm banana plugs for rapid on-site
Z changes.
He can match anything to anything and is happy.
I couldn't seetoo much wrong with the trannies which had a full power
bandwidth
from less than 10Hz to 1MHz.

Methinks such matching trannies would be a godsend for those struggling
to stop
tubes burning out in their OTL amplifiers. a 4 ohm load is usually
hellishly difficult for
any OTL amp, but where the amp can "see" 24 ohms, output power can be
greater than
with 4 ohms, and there will be much more class A, less thd/imd, less
severe pda,
much better DF, and no wonder ppl say a matching tranny with OTL is a
great idea.
( So why omit the OPT in the darn first instance then??...its another
story. )

The transformers I boxed for my customer came from Paul Speltz at
http://www.zeroimpedance.com/

These were not particularly neatly wound as i would have liked,
indicating
they were wound by cheap labour goodness knows where, ( China? Mexico? )
but they worked, and insulation isn't critical
because speaker voltages are low, and heat is minimal.

If you have a GOSS spiral core rated for 500VA mains use,
and about 300turns of 1.5mm dia for the primary and secondary
and with appropriate taps, its very easy to get wide bw and low losses
and THD/IMD.

Ordinary E&I lam trannies also are possible with 2" stack 1.5" tongue
GOSS fully meshed core,
just fill up the bobbin with layers PSPSPSPSP interleaved layers of say
1.6mm dia wire
and vary wind arrangements to suit whatever may be required.

Patrick Turner.

I appreciate your comprehensive reply, Patrick It's another couple of pages
for my theory folder. With all the posts you made like this, perhaps you
should just collate them and write a practical audio book, seriously.

I don't have the time to collect all my posts together as well as the
many wonderful
posts from everyone else, them collate them to produce a book along the
line of
Tremain's Audio Encyclopedia, which is done with question and answer
format.

I have to make a living from audio, and writing books is something one
does when one is on a pension
and doesn't have to wrestle with customers and money.


However, I was thinking more on the lines of keeping it simple, at least for
now by using a 2-way crossover such as the tube Marchland. This way I'll use
the passive part of the Infinity's xrv.. that pertains to the EMIT & Mid.
The active xrv. would feed that (tube amp) and also feed the woofers amp. I
still end up with the woofers' amp looking at 16 ohms.

If you just have bi-amping, its relatively simple.
And at least you can adjust levels of bass to mid-treble, and possibly
the Xover poles.
This removes the need for padding the mid/treb to balance with the bass
levels.
personally, I prefer the more flexible approach of having all passive
Xovers and one good amp.

I think you seen the
schematic of the Infinity's xrv a while back. Sorry I didn't make my self
clearer in the OP (is that what you meant by getting in trouble?).
Finally again, what do you think about a professional line of amps, like
Pyramid, driving the woofers only. I think I'm missing a point somewhere
because this approach seems too simple. No ridiculing please. Only Allison
is allowed.

Cordially,
west

The sound you experience from a system is always the engineered result
of its components
along with room and furnishing effects.

There is more than one way to achieve good audio, and whether it is
better than
the next guy's system is always a moot point.

So if you take ENOUGH TROUBLE TO GET IT RIGHT then probably your efforts
will reward you.
Nearly all SS amps are good enough for bass at least and many good ones
compare favourably with
the good tube amps for full range audio. I prefer having all tubes but
if i must use SS I would say OK,
use it for the bass, and leave my midrange and treble for the tubes.

So whatever you choose, if the assembly is optimised in a wholistic (
synergistic ?? ) way,
then you get good sound.
I like simple signal paths myself, and see nothing wrong with CD player
- preamp - power amp - 3 way xovers - speakers.
Some would say that having 2 power amps, and removing the passive
Xovers, or part of them, and adding active Xovers
ahead of the power amps is the same thing but to me the active Xovers
challenge the sound more than passive Xovers.
One could spend a month of sundays building passive crossovers to suit
say 5kohm source resistance and termination resistances
with L&C elements placed before the power amps which are passive, and
which don't load down the output of a preamp.
Potentially, this preserves the "purity of design" theory behind the
system, so nothing is lost due to adding yet more electronics
in the signal path. Some would say No, use opamps with NFB to make the
filters, some may say No,
use discrete j-fets or bjts or tubes for the active Xovers with NFB for
the filters.
Who is right? Well, that depends on the sound you get, and that depends
on your choice
of what you use and how you use it.

Say you have 20 opamps in series, each making 0.005% thd, and each
giving gain of 1.41
How much thd is produced at the end of the run?. Gain will be 1.41
raised to the power of 20, or "lots" of gain.

But will you hear anything different to where you have 2 stages of gain
giving the same gain,
with the same thd?

Who knows the answer until you do the tests to find out?

But if it is doubtful that 20 seriesed opamps would cause any
perceptable sound change, then adding ONE
more active lot of electronics should indeed be difficult to detect, and
perhaps no worse than
the passive LCR circuits being replaced.
May even be an improvement. Depends on who does it, and how well.

There is a point where any advice given could be either right or wrong,
and/or not worth
being given; the outcome depends on you and nobody else, or at least on
the total expertise
used to construct the system.

The use of active Xover and direct bass speaker drive should make the
bass
heard "better", ie, more accurate, fast, real, etc, because the speaker
cones are
connected to a source impedance that is just the LOW Rout of the amp
without
all the L and C crap attached.
Trouble is that even though the speaker is controlled better the sound
released
from it is not within any NFB loop, so harmonics are produced by box
resonances,
room effects, whatever.
Some folks believe the best bass is made with drivers fitted with motion
sensors
so that cone movement is converted to a feedback signal and compared
with an input signal
and this is the only way to properly control a bass driver.

I just don't have the experience to say yes or no.

Patrick Turner.

[Ian Iveson]

west wrote

I like the sound of my Infinity RS-IIIB speakers, but as most of you
know,
it demands lots of current. All my tube amps sound great with the
Infinitys
at low levels, but peter out with demanding music. The Infinitys
have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2)
woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a
Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire
them in
parallel (4 ohms)? Is this sound practice? Thanks in advance.


Fred's advice seems practical and reasonable, West.

Although this seems a bit snooty: "He just needs a decent
electronic crossover. If he can't proceed from here, he shouldn't be
doing the
work..."

Why not? If you get stuck on setting up the line-level xover, you can
always ask again...perhaps on rec.audio.tech.

Not knowing much about speakers and xovers, I wonder if simply
disconnecting the bass units from the existing speaker-level xovers
will change the way they deal with the remaining speakers?

It seems to me that the question of whether the bass units are best in
series or parallel depends on how well matched they are. If perfectly
matched, it shouldn't make any difference, but they won't be. I don't
know whether the consequent interactions between them will be
significant. Assuming their driving amplifier performs just as well
into 4ohms as into 16ohms, I suppose the best way of deciding is to
listen, considering it will be easy to switch between the two options.

However, even if the amp driving the bass units is happy with either
load, it is likely that its distortion characteristics will be
different. I guess you need a decent amp if you want to make a fair
comparison. If you're not interested in fair comparisons, just listen.

cheers, Ian

[Arny Krueger]

"west" wrote in message
news:e4iAh.948$hp4.706@trnddc02...

I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the
Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.

My thoughts are to drive the EMIT & mid with my tube amp and the (2)
woofers
with some SS amp. Two immediate questions come to mind ...

1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?

There's a well-known joke in the industry, two words: "Pyramid watts". They
are a fraction of real watts.

If you want a good cheap amp, try something like a Behringer A500 - a pretty
genuine 130 wpc.

2. Since the woofers are wired in series (16 ohms), should I rewire them
in
parallel (4 ohms)? Is this sound practice? Thanks in advance.

If there's no internal crossover parts, or if you bypass them, yes. When you
use two amplifiers like this, usual practice is to use an electronic
crossover to split the signal between the bass speakers and the rest. If you
don't do this, you're spinnning your wheels because you'll still be passing
the same amounts of bass to your amps for the upper-range drivers.

Usual practice would be to use an electronic crossover to split the signal
between the LF amp and the HF amp, and drive the woofer voice coils directly
from the LF amplifier.

[Trevor Wilson]

"west" wrote in message
news:e4iAh.948$hp4.706@trnddc02...
I like the sound of my Infinity RS-IIIB speakers, but as most of you know,
it demands lots of current. All my tube amps sound great with the
Infinitys
at low levels, but peter out with demanding music. The Infinitys have an
EMIT tweeter, Dome mid, & (2) 12 inch Woofers wired in series.
My thoughts are to drive the EMIT & mid with my tube amp and the (2)
woofers
with some SS amp. Two immediate questions come to mind ...
1. Can I use for the woofers an inexpensive commercial amp like a Pyramid
without compromising the sound?
2. Since the woofers are wired in series (16 ohms), should I rewire them
in
parallel (4 ohms)? Is this sound practice? Thanks in advance.


**You're on a path to a beating here. First off, connecting the drivers in
parallel will make the system MUCH more difficult to drive. Then, of course,
the crossover will be useless and you will need to re-engineer it to suit
the new bass impedance. Lastly, the RSIIIb uses 250mm (10 inch) bass
drivers, not 300mm drivers.

But wait: There's more. Substituting drivers in the RSIIIb is fraught with
problems. The 250mm drivers employ a 'curvilinear' cone, which has very
specific advantages not available with regular cones. If you replace the
drivers with regular cone drivers, you can expect serous problems. Then, of
course, is the fact that the drivers used are designed for use in a sealed
enclosure. You would need to find suitable drivers for such an enclosure AND
ones which worked with the existing crossovers.

Having said all that, you will find that the impedance of the RSIIIb varies
from around 4 Ohms down to around 2.5 Ohms. The bass impedance curve is
quite difficult and this explains why you have problems. Your BEST option is
to separate the bass from the mids and HF and operate the bass section from
it's own solid state amp and use the tube amp for mids and HF.


--
Trevor Wilson
www.rageaudio.com.au



--
Posted via a free Usenet account from http://www.teranews.com