Impedance matching anyone

[corlioni1976]

Hello All,

I'm trying to impedance match two systems together. I have an audio source
(600 ohms) and a speaker (15 ohms) and I need to find a way of getting them
to work.

Can anyone point me in the direction of a good book or website that will
explain how this might be done. I have built an op amp amplifier and this
works. But if a passive approach is possible I would like to take that as
my interface will be battery powered.

Many thanks for any help/advise.

Naveed

 

[Don Pearce]

On Tue, 17 May 2005 12:08:53 +0100, "corlioni1976"
<corlioni1976REMOVE@yahoo.co.uk> wrote:

Hello All,

I'm trying to impedance match two systems together. I have an audio source
(600 ohms) and a speaker (15 ohms) and I need to find a way of getting them
to work.

Can anyone point me in the direction of a good book or website that will
explain how this might be done. I have built an op amp amplifier and this
works. But if a passive approach is possible I would like to take that as
my interface will be battery powered.

Many thanks for any help/advise.

Naveed

You have already done the right thing. Passive impedance matching is
certainly possible with a transformer, but there are two problems. I,
there will not be enough power to driver the speaker, and 2, speakers
are not designed to be matched - they need a zero impedance voltage
source.

d

Pearce Consulting
http://www.pearce.uk.com

 

[James Meyer]

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.

Speakers will work just fine when connected to a wide range of source
impedances.

Speakers (drivers) in a two-way system, for instance, seldom "see" a
zero impedance voltage source.

Jim

 

[Don Pearce]

On Tue, 17 May 2005 11:55:52 GMT, James Meyer <jmeyer@nowhere.net>
wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.

No they won't. One look at the impedance curve of a speaker should
tell you that they need a low - if not zero - source impedance if they
are not going to honk like geese.

Speakers (drivers) in a two-way system, for instance, seldom "see" a
zero impedance voltage source.

Jim

Such speakers generally have no pretensions to, nor need for,

fidelity.

d

Pearce Consulting
http://www.pearce.uk.com

 

[John Fields]

On Tue, 17 May 2005 12:14:01 GMT, donald@pearce.uk.com (Don Pearce)
wrote:

On Tue, 17 May 2005 11:55:52 GMT, James Meyer <jmeyer@nowhere.net
wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.


No they won't. One look at the impedance curve of a speaker should
tell you that they need a low - if not zero - source impedance if they
are not going to honk like geese.

---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...

--
John Fields
Professional Circuit Designer

 

[Don Pearce]

On Tue, 17 May 2005 08:02:13 -0500, John Fields
<jfields@austininstruments.com> wrote:

On Tue, 17 May 2005 12:14:01 GMT, donald@pearce.uk.com (Don Pearce)
wrote:

On Tue, 17 May 2005 11:55:52 GMT, James Meyer <jmeyer@nowhere.net
wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.


No they won't. One look at the impedance curve of a speaker should
tell you that they need a low - if not zero - source impedance if they
are not going to honk like geese.

---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...

I'm talking about the considerable mismatch of an electrically matched
system - where the source impedance would be 8 ohms instead of zero.
That makes for considerable unflatness in response.

I've never come across a speaker designed to be powered from anything
other than as low an impedance as possible - can you point me at one?
In fact I didn't even know it was an electromechanical possibility.

d

Pearce Consulting
http://www.pearce.uk.com

 

[corlioni1976]

"Don Pearce" <donald@pearce.uk.com> wrote in message
news:428fd1a9.185844171@news.plus.net...

On Tue, 17 May 2005 12:08:53 +0100, "corlioni1976"
corlioni1976REMOVE@yahoo.co.uk> wrote:

Hello All,

I'm trying to impedance match two systems together. I have an audio
source
(600 ohms) and a speaker (15 ohms) and I need to find a way of getting
them
to work.

Can anyone point me in the direction of a good book or website that will
explain how this might be done. I have built an op amp amplifier and
this
works. But if a passive approach is possible I would like to take that
as
my interface will be battery powered.

Many thanks for any help/advise.

Naveed


You have already done the right thing. Passive impedance matching is
certainly possible with a transformer, but there are two problems. I,
there will not be enough power to driver the speaker, and 2, speakers
are not designed to be matched - they need a zero impedance voltage
source.

d

Pearce Consulting
http://www.pearce.uk.com

Hi Don

Many thanks for your reply.

Naveed

 

[Tam/WB2TT]

"John Fields" <jfields@austininstruments.com> wrote in message
news:lkqj81dsbn84eolqm7a6pt49k73n3j0t54@4ax.com...

On Tue, 17 May 2005 12:14:01 GMT, donald@pearce.uk.com (Don Pearce)
wrote:

On Tue, 17 May 2005 11:55:52 GMT, James Meyer <jmeyer@nowhere.net
wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce)
wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.


No they won't. One look at the impedance curve of a speaker should
tell you that they need a low - if not zero - source impedance if they
are not going to honk like geese.

---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...

--
John Fields
Professional Circuit Designer

John,

Any tube amplifier that claimed to be high fidelity had enough negative
feedback to bring the output impedance below 1 Ohm. The ancient Radiotron
Designers Handbook talks of damping factors of around 20.

Tam

 

[Tim Wescott]

John Fields wrote:

On Tue, 17 May 2005 12:14:01 GMT, donald@pearce.uk.com (Don Pearce)
wrote:


On Tue, 17 May 2005 11:55:52 GMT, James Meyer <jmeyer@nowhere.net
wrote:


On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:


........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.


No they won't. One look at the impedance curve of a speaker should
tell you that they need a low - if not zero - source impedance if they
are not going to honk like geese.


---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...

More like designed to match the ratio of tube supply voltage to

available plate current to the loudspeaker's nominal impedance. Then,
as mentioned in other posts, negative feedback (or triode finals) were
used to bring the impedance of the speaker terminals down.

--
-------------------------------------------
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Don Pearce]

On 17 May 2005 07:49:36 -0700, "Mark" <makolber@yahoo.com> wrote:

Yes,

some Fender guitar amplifiers are designed with current feedback that
actually raises their output impedance. They drive the speakers fine.

I'm not talking about amplifier design, but speaker design. And of
course guitar amplifier designers strive for the sound they want, not
fidelity.

I modified one to change the feedback to voltage feedback to lower the
output impedance. There was a subtle change to the bass response.

A speaker driven with a hi Z will have a more pronounced peak at the
bass resonant frequency and will have a bit more treble.

As I said.

The difference is a bit subtle and depending on the OPs application may
not be a big issue.

I believe his application is to monitor what is happening on a line

out, therefore we must, as best guess, consider that fidelity is a
prerequisite.

d

Pearce Consulting
http://www.pearce.uk.com

 

[John Fields]

On Tue, 17 May 2005 07:34:01 -0700, Tim Wescott <tim@seemywebsite.com>
wrote:

John Fields wrote:

---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...

More like designed to match the ratio of tube supply voltage to
available plate current to the loudspeaker's nominal impedance.

---
Since the ratio of the supply voltage, V, to the plate current, I, is


V
---
I

and since that comes out to


V
--- = R
I

and R is the plate resistance, I think that's what I said.

That is, if we were running a couple of 6L6's in push-pull with a 400V
supply and about 100mA of plate current we'd be looking at
transforming about 4000 ohms down to whatever the speaker was, say 8
ohms.
---

Then, as mentioned in other posts, negative feedback (or triode finals) were
used to bring the impedance of the speaker terminals down.

---
I believe that that feedback was used to _linearize_ the amp, not to
bring its impedance down, per se. Think about it like this: Say
we've got a tube amp with an 8 ohm output and we have an 8 ohm and a 4
ohm loudspeaker. First, with the volume turned down, we take the 8
ohm speaker, connect it to the 8 ohm output, and connect an audio
oscillator to the amp's aux input. Next, we adjust the input to the
amp to 1VRMS and 1000Hz. Then we crank up the volume until we get
1VRMS across the loudspeaker. Once we've done that, we turn the audio
source down to zero and substitute the 4 ohm speaker for the 8 ohm
one. Now, we adjust the output of the audio source back up to 1VRMS
and measure the voltage across the loudspeaker. Do you think it's
going to be 1VRMS? I don't, and if it isn't, the amplifier won't be a
voltage source and its output impedance won't be zero ohms.

--
John Fields
Professional Circuit Designer

 

[Joerg]

Hello Tim,

More like designed to match the ratio of tube supply voltage to
available plate current to the loudspeaker's nominal impedance. Then,
as mentioned in other posts, negative feedback (or triode finals) were
used to bring the impedance of the speaker terminals down.

Only one out of several tube amps I repaired many moons ago had
feedback. The biggest ones didn't, yet they sounded really cool.

Same with our Hammond organ which came to us as a basket case. So I had
to get schematics and dive in. Same thing, no feedback, yet fidelity is
really crucial on those instruments.

Regards, Joerg

http://www.analogconsultants.com

 

[John Popelish]

corlioni1976 wrote:

Hello All,

I'm trying to impedance match two systems together. I have an audio source
(600 ohms) and a speaker (15 ohms) and I need to find a way of getting them
to work.

Can anyone point me in the direction of a good book or website that will
explain how this might be done. I have built an op amp amplifier and this
works. But if a passive approach is possible I would like to take that as
my interface will be battery powered.

Many thanks for any help/advise.

Naveed


You don't need a perfect impedance match to transfer most of the

available power. An 800 ohm to 8 ohm transformer rated for 400mW
(42TU008) is probably close enough and costs about $2 from Mouser

 

[Tim Wescott]

John Fields wrote:

On Tue, 17 May 2005 07:34:01 -0700, Tim Wescott <tim@seemywebsite.com
wrote:


John Fields wrote:


---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...


More like designed to match the ratio of tube supply voltage to
available plate current to the loudspeaker's nominal impedance.


---
Since the ratio of the supply voltage, V, to the plate current, I, is


V
---
I

and since that comes out to


V
--- = R
I

and R is the plate resistance, I think that's what I said.

Yes, but you were wrong. The plate resistance of a tube is defined as
the slope of the V/I curve at the operating point; for a 6L6 class A
amplifier operating at 350V plate voltage and 60mA plate current the
plate resistance is 33k ohm. The specified _load_ resistance in this
case is 4.2k ohm.

That is, if we were running a couple of 6L6's in push-pull with a 400V
supply and about 100mA of plate current we'd be looking at
transforming about 4000 ohms down to whatever the speaker was, say 8
ohms.

Correct, and that'll make sure that you run out of current at the same
time you run out of voltage. It _won't_ make the amplifier impedance
equal to 8 ohms however -- with a 33k plate impedance and a 500:1
impedance ratio the impedance at the amplifer terminals will be around
66 ohms.

---


Then, as mentioned in other posts, negative feedback (or triode finals) were
used to bring the impedance of the speaker terminals down.


---
I believe that that feedback was used to _linearize_ the amp, not to
bring its impedance down, per se. Think about it like this: Say
we've got a tube amp with an 8 ohm output and we have an 8 ohm and a 4
ohm loudspeaker. First, with the volume turned down, we take the 8
ohm speaker, connect it to the 8 ohm output, and connect an audio
oscillator to the amp's aux input. Next, we adjust the input to the
amp to 1VRMS and 1000Hz. Then we crank up the volume until we get
1VRMS across the loudspeaker. Once we've done that, we turn the audio
source down to zero and substitute the 4 ohm speaker for the 8 ohm
one. Now, we adjust the output of the audio source back up to 1VRMS
and measure the voltage across the loudspeaker. Do you think it's
going to be 1VRMS? I don't, and if it isn't, the amplifier won't be a
voltage source and its output impedance won't be zero ohms.

That all depends on how the amp is designed, doesn't it? A 2A3 final

stage would have a load resistance of 2500k at the plates for about a
300:1 impedance transformation ratio. A 2A3 in a class-A amp has an 800
ohm plate resistance, so the amplifier as a whole would have a terminal
resistance of about 2.7 ohms. Not zero, but much less than 8 if my math
skills haven't abandoned me. Similar results could be had with a
pin-toad amplifier using negative feedback.

-------------------------------------------
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Tim Wescott]

Joerg wrote:

Hello Tim,

More like designed to match the ratio of tube supply voltage to
available plate current to the loudspeaker's nominal impedance. Then,
as mentioned in other posts, negative feedback (or triode finals) were
used to bring the impedance of the speaker terminals down.


Only one out of several tube amps I repaired many moons ago had
feedback. The biggest ones didn't, yet they sounded really cool.

Same with our Hammond organ which came to us as a basket case. So I had
to get schematics and dive in. Same thing, no feedback, yet fidelity is
really crucial on those instruments.

Regards, Joerg

http://www.analogconsultants.com

Did any of them use the ultra-linear connection in the final? That's a
non-obvious form of feedback.

I would class guitar and your Hammond completly differently than a hi-fi
amp: in the case of of the amp being part of the instrument you can
tailor the whole system to include the speaker response; in the case of
the hi-fi amp you would have to be more careful about how the sound is
reproduced.

"Roger" mentioned on an adjoining thread that with a high-impedance
amplifier output you need to match the speaker to the amplifier. IIRC
the cool early hi-fi speakers (and possibly current, I just listen to
boom boxes) were "low efficiency", presumably with sacrifices made in
the speaker to get better sound. If you're dinking around with the
speakers at the same time you're dinking around with the amplifier it's
hard to tell what changes are affecting what...

-------------------------------------------
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Mac]

On Tue, 17 May 2005 11:55:52 +0000, James Meyer wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.

Speakers (drivers) in a two-way system, for instance, seldom "see" a
zero impedance voltage source.

Jim

Well, a 16 Ohm speaker is not going to work very well when driven by a 600
Ohm source. ;-(

--Mac

 

[James Meyer]

On Wed, 18 May 2005 05:06:09 GMT, Mac <foo@bar.net> wroth:

On Tue, 17 May 2005 11:55:52 +0000, James Meyer wrote:

On Tue, 17 May 2005 11:14:13 GMT, donald@pearce.uk.com (Don Pearce) wroth:

........... speakers
are not designed to be matched - they need a zero impedance voltage
source.


Speakers will work just fine when connected to a wide range of source
impedances.

Speakers (drivers) in a two-way system, for instance, seldom "see" a
zero impedance voltage source.

Jim

Well, a 16 Ohm speaker is not going to work very well when driven by a 600
Ohm source. ;-(

--Mac

That's why the subject line of this thread includes the words "impedance
matching".

A 16 ohm speaker driven from a 600 ohm source *will* work fine. It will
simply have less output energy at a fixed voltage from the 600 ohm source than
when a 600 to 16 ohm matching transformer is used between the source and
speaker.

A zero ohm source impedance such as Don recommends is *very* difficult
to achieve with passive components. The original question asked about
transformers. A zero ohm impedance winding on a transformer will result in zero
volts of output.

Jim

 

[Joerg]

Hello Tim,

Did any of them use the ultra-linear connection in the final? That's a
non-obvious form of feedback.

I have not seen a screen grid connection to a tap of the output
transformer in any of them. Probably I didn't get to see the Rolls Royce
class of tube amps in those days.

I would class guitar and your Hammond completly differently than a hi-fi
amp: in the case of of the amp being part of the instrument you can
tailor the whole system to include the speaker response; ...

It is likely that they did that. Although I believe Hammond had the
electronics module designed by an outside contract firm.

"Roger" mentioned on an adjoining thread that with a high-impedance
amplifier output you need to match the speaker to the amplifier. IIRC
the cool early hi-fi speakers (and possibly current, I just listen to
boom boxes) were "low efficiency", presumably with sacrifices made in
the speaker to get better sound. If you're dinking around with the
speakers at the same time you're dinking around with the amplifier it's
hard to tell what changes are affecting what...

I remember my teenage days when it was customary to pack speakers with
bales of damping materials. Then you needed hundreds of watts to drive
them but luckily there was that old cheap work horse, the 2N3055. I did
prefer tubes in those days. More oompf. The meanest tube amp delivered
around 700-800 watts, mostly depending of how much the mains circuit
breaker could stomach. One strong twang on the electric guitar and the
lights went out.

The real freaks built Klipsch horns. But that required access to
precision wood working tools which none of us had.

Regards, Joerg

http://www.analogconsultants.com

 

[Tam/WB2TT]

"Tim Wescott" <tim@seemywebsite.com> wrote in message
news:118li5329unjt20@corp.supernews.com...

John Fields wrote:

On Tue, 17 May 2005 07:34:01 -0700, Tim Wescott <tim@seemywebsite.com
wrote:


John Fields wrote:


---
Strangely, loudspeakers connected to toob amps with those big ol'
output tranformers seem to honk OK, and they're certainly nowhere near
a zero source impedance. Designed to match the plate resitance of the
toobs to the speaker impedance, I believe...


More like designed to match the ratio of tube supply voltage to available
plate current to the loudspeaker's nominal impedance.


---
Since the ratio of the supply voltage, V, to the plate current, I, is


V
---
I

and since that comes out to


V
--- = R
I and R is the plate resistance, I think that's what I said.

Yes, but you were wrong. The plate resistance of a tube is defined as the
slope of the V/I curve at the operating point; for a 6L6 class A amplifier
operating at 350V plate voltage and 60mA plate current the plate
resistance is 33k ohm. The specified _load_ resistance in this case is
4.2k ohm.

Tim,
Keep in mind, though, that later tube amplifiers used an "ultra linear"
configuration, where the screens were connected to taps on the output
transformer. This made a pentode look more like a triode.

That is, if we were running a couple of 6L6's in push-pull with a 400V
supply and about 100mA of plate current we'd be looking at
transforming about 4000 ohms down to whatever the speaker was, say 8
ohms.

Correct, and that'll make sure that you run out of current at the same
time you run out of voltage. It _won't_ make the amplifier impedance
equal to 8 ohms however -- with a 33k plate impedance and a 500:1
impedance ratio the impedance at the amplifer terminals will be around 66
ohms.

Now add in 20 DB of negative feedback to the input of the phase inverter,
plus the local negative feedback to the screens. Unless I am getting
confused, a damping factor of 20 with a 16 Ohm speaker implies an output
impedance of 0.8 Ohms.

Tam

---


Then, as mentioned in other posts, negative feedback (or triode finals)
were used to bring the impedance of the speaker terminals down.


---
I believe that that feedback was used to _linearize_ the amp, not to
bring its impedance down, per se. Think about it like this: Say
we've got a tube amp with an 8 ohm output and we have an 8 ohm and a 4
ohm loudspeaker. First, with the volume turned down, we take the 8
ohm speaker, connect it to the 8 ohm output, and connect an audio
oscillator to the amp's aux input. Next, we adjust the input to the
amp to 1VRMS and 1000Hz. Then we crank up the volume until we get
1VRMS across the loudspeaker. Once we've done that, we turn the audio
source down to zero and substitute the 4 ohm speaker for the 8 ohm
one. Now, we adjust the output of the audio source back up to 1VRMS
and measure the voltage across the loudspeaker. Do you think it's
going to be 1VRMS? I don't, and if it isn't, the amplifier won't be a
voltage source and its output impedance won't be zero ohms.
That all depends on how the amp is designed, doesn't it? A 2A3 final
stage would have a load resistance of 2500k at the plates for about a
300:1 impedance transformation ratio. A 2A3 in a class-A amp has an 800
ohm plate resistance, so the amplifier as a whole would have a terminal
resistance of about 2.7 ohms. Not zero, but much less than 8 if my math
skills haven't abandoned me. Similar results could be had with a pin-toad
amplifier using negative feedback.

-------------------------------------------
Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

 

[Paul Hovnanian P.E.]

corlioni1976 wrote:

Hello All,

I'm trying to impedance match two systems together. I have an audio source
(600 ohms) and a speaker (15 ohms) and I need to find a way of getting them
to work.

I think they refer to these devices as audio amplifiers.

Can anyone point me in the direction of a good book or website that will
explain how this might be done. I have built an op amp amplifier and this
works. But if a passive approach is possible I would like to take that as
my interface will be battery powered.

A passive approach will work only if you don't need any gain.

Many thanks for any help/advise.

Naveed

--
Paul Hovnanian mailtoaul@Hovnanian.com
------------------------------------------------------------------
I want to die quietly in my sleep, like my grandfather,
not screaming in terror, like his passengers.