Understanding impedence matching

I am a psychologist and my eyes start crossing when the conversation gets too technical, but that being said, I am beginning to wonder what impedence matching means and what I should pay attention to as I upgrade my setup.

The two pieces in my system that are likely here to stay for a while are my SF Electra Amator II, that evidently don't drop below 6 ohmns, making them receptive to an OTL amp, and my Cary pre, that has phono impedence @ 47K (evidently this can be adjusted) and output at 800 ohms. I am in the process of upgrading my amp and TT.

What does this mean in terms of how I think about these two acquisitions?? Are there specific numbers I need to pay attention to, and/or formulas for these, or is it all in the listening?

As always, thanks for your help. This community is amazing.
Think of impedence as language. The closer they are the same, the greater the amount of information can pass.
Sometimes the mismatch is on purpose, and in those situations, the language analogy does not work. Mostly those deliberate mismatched situations are to increase stability of the components.
(electrical engineers are free to flog me...)

First - when you are matching speakers and amps - you are not
doing "impedance matching". Impedance matching is when you
are dealing with transmission lines - like the coax or
twin-lead antenna cable going to your TV. You want to match
impedances so you don't get internal reflections. Imagine
light going through air, then hitting a glass window or
aquarium. There is a mismatch in impedance and you will get
partial reflection at the surface. That's why you can see
yourself in a window - like a partially silvered mirror.
If you match impedances - you don't get the reflection.

What you are doing in matching speakers and amp is not
impedance matching. The lower the impedance of the speakers
the more current has to flow through them for a given
voltage. For example, a 4 ohm speaker [ on average ] needs
twice the current as an 8 ohm speaker given the same drive
voltage. The amp may not have the current reserves to
supply the current required. Thats why an 8 ohm speaker
is less taxing on the amp than a 4 ohm speaker. You follow
the amp manufacturer's recommendation for the load impedance.

If you have too low an impedance, the current reserves of
the amp will be taxed. At low volumes, it'll work - but as
you crank up the volume, and hence voltage, the amp has to
deliver more and more current to the load. If it runs out
of current, the amp will "clip" - it will chop the top off
the music waveform. That type of distortion is death to
speakers - it will fry the tweeters.

Dr. Gregory Greenman
A couple things....

Elizibeth..how goes it? Your analogy to language is fine except it covers only part of the issue. If impedances are matched you get more power. (Power = current x voltage.)But a mismatch in impedances is not just for stability.

Power is not always the goal and sometimes you want to avoid it because you do not want to load down the prior circuit. For example, if you are in the front-end (pre) of a system with a small voltage that you want to pass to the amp you want a small to larger impedance in order to do it. One of the reasons a tube's grid is used to develope the signal is that it has an almost* infinite resistance. Perfect for developing a voltage signal. Anyway, what I wanted to emphasize is that in order to think about "matching impedance" you have to figure out what the goal is. transferring voltage or current.

Esteemed Dr. Greenman:
I am not a physicist but I know how amps work. You certainly are matching impedances from amp to speaker. This is at the heart of the OTL issue in fact. The primary purpose of a transfermer in a tube amp is to match the output tube plate circuit, which requires a large (some thousands of ohms) impedance as a load, to a speaker that is small impedance (2-8typically). The impedance is changed by the square of the turns ratio of the output transformer. This is the fist thing you must spec in a transformer. This is the primary reason conventional tube amps have transformers and why OTLs are, well, different. Now most folks don't need to know this except it makes it easier to understand that when they look at the 4 and 8 ohm taps on their amp they are looking at different taps to the transformer's secondary and using one tap or the other
changes the turns ration, which in turn, changes the nominal impedance match from amp to speaker.

A further note, most amps would do a lot better if the folks who built knew what speakers were going to be attached to them because they would know the load. Unfortunately we are all in love with separates.

Matching Z takes place through the entire system, is not limited to transmission lines as you seem to suggest, and IMHO, if folks understood it (and a little RLC) about 3/4 of the magic and synergy would disappear from the wires used in audio. It is fatal if overlooked with passive pres and such. There are many places you can go to get a little more on this simple point and, if anyone is interested, you might try online. Here is one short piece.


I remain
not a physicist, and
This is so cool. I'm all over google trying to make sense out of what you good people are saying. It's slow, it's slow. Clueless, you say the point about impedence is that I need to figure out what the goal is, transferring voltage or current. Can you elaborate on this? How do you tell what the goal should be?

I am not even going to touch all the other questions I have - I'll keep reading around and try to learn a little more about the basics.

Many thanks to you all. Happy listening. Chuck
Chuck --
investing a little time in understanding Ohm's law will take you a long way towards understanding these issues. It's a bit like learning a language: in the beginning it feels daunting and impossible and then, suddenly, you loosen up & speak the language... you've acquired the "knowledge" as it were.

OK, you may not be ready to design your own amps, etc -- but, as Clueless (and not a physicist:)) notes,
if folks understood ...(that)... (and a little RLC) about 3/4 of the magic and synergy would disappear from the wires used in audio
However, the magic of matching devices into a system would appear.

Try this link, too. It's a short introduction to electricity & components.
It can be fascinating!
A couple of practical points specific to your components: since your pre has an output impedence of 800 ohms, using the 1:10 rule your power amp should have an input impedence of at least 8k, which is easy to find. (My Alephs, for example, have an unusually very low input impedence of 10k ohms.) Secondly, and perhaps more importantly, use of a tube amp with a high output impedence (let's say above 1 ohm) WILL couple with your speakers' variable impedence differently as a function of frequency, resulting in the somewhat unpredictable trampolining of its freq response that can make tube amps so much fun (!?). However, a 6ohm speaker load min is pretty high, so MOST tube amps will show minimal timbral distortion. Again, think of the varying load response the amp sees as it's 1 ohm impedence meets a 6-20 ohm load across 30-20kHz window. (Yes, study ohm's law.) Then imagine what happens with a more common lower impedence load, often varying down to impedence minima of 2-3 ohms!
There can then easily be a 10x change in "load" across the frequency band, severely modulating the response. Usually the upper bass and mids get plumped up, the highs rolled off. Only works well a small percentage of the time since it's such a wild crapshoot. So even though your speaker is relatively amp friendly I'd chase an amp with a low output impedence to maximize your odds of a reasonably flat frequency response without having to tweak the hell out of your room, or other consequent bandaids. Good luck.
If the circuitry is stable and well designed, the source output impedance should match that of the load input impedance. If the load contains a high level of reactance, especially at sharp phase angles, the source output impedance should be measurably lower than that of the load input impedance. If our audio gear was designed properly, we wouldn't be using a 1 / 10 ratio or anything close to that.

The only company that i've ever seen that offers a product line that takes advantage of such design features would be Sierra Audio ( i "think" it was them ). Jud Barber of Joule Electra has also made comments similar to this, but not to the point of promoting matching input / output impedances. Sean
Thanks everyone. I have begun my inquiry into Ohm's law and its implications. I've gotten the E=IR concept down, but as to the implications, well I gotta keep studying. But since there is some conviction from you that understanding this does, in fact, help demystify the process, I am eager to learn.

Again, my thanks. The more I explore the web, the more amazed I am by its power to connect strangers.

I'm sorry - you are misusing the term "impedance matching".

When you have tranmission lines - and you do "impedance
matching" - you make the impedance of one element exactly
the SAME as the upstream components.

For example, if you have a transmission line with twin-lead -
the type that's used for antenna connections to TVs - that's
a 75 ohm transmission line. When you want to terminate the
transmission line - you use a 75 ohm resistor. That's
because a 75 ohm resistor looks to a twin-lead transmission
line like an infinite length of twin-lead transmission line.

THAT is impedance matching - making the impedances the SAME
hence the term "matching".

I would use the term "coupling" not "matching" in the example
you give with the tube amp. Even with the transformer -
the output impedance of the tube amp is not 2-8 ohms.

In fact, the ratio of the output impedance of the amp to
the load impedance of the speaker is called the "damping
factor" and it is not unity [ 1.0 ] in a good stereo setup.

For the amp to exert good control on the speaker - the
damping factor is usually a few hundred. That is the
output impedance of the amp is around a few thousand ohms.

In an audio system - the interconnects ARE transmission
lines - typically 47k ohm transmission lines.

I believe we have a semantics problem here. You evidently
well understand how amps and speakers work. However, you're
evidently unfamiliar with the terms "transmission lines"
and "impedance matching". These are well defined terms in
science and electrical engineering.

Courtesy of the Institute for Telecommunication Sciences:

Transmission line definition:


Impedance matching definition:


Dr.Gregory Greenman
"For the amp to exert good control on the speaker - the
damping factor is usually a few hundred. That is the
output impedance of the amp is around a few thousand ohms."

This one is wrong on both counts.

"For the amp to exert good control on the speaker - the
damping factor is usually a few hundred."

Damping factor doesn't determine the amount of "control" that the amp has over the driver. This is one of the most widely misunderstood and ill-repeated concepts in the industry.

Power transfer and loading characteristics determine the amount of control the amplifier has over the driver. The more power that the amp can deliver into the driver with the least amount of reactance within the passband being used, the more "control" or "influence" the amp has over the driver.

Damping factor is simply a measure of how susceptable the amplifier is to having a specific impedance speaker modulate its' output. The bigger the impedance mismatch, the less power that the speaker can load back into the amp. This is a good thing since speakers are not purely resistive and will always demonstrate various levels of reactance. Impedance matching increases power transfer and communications in both directions. Since we don't want the speaker "talking back" to the amp, we purposely design in an impedance mismatch. We simply want the amp telling the speaker what it should do, not the other way around.

The closer the output impedance of the amp to the input impedance of the speaker, the easier it is for the speaker's reactance to modulate or interfere with the amplifier's normal operation. This is true of both SS and tube designs, but more-so with tubes due to their typically higher output impedance. Hence the wide frequency response deviations found in SET or OTL amps when connected to various speakers / levels of reactance & impedances. This demonstrates the previous rules i stated above i.e. when reactance is involved, the source output impedance should be much lower than the load input impedance. Using this approach, the higher levels of reactance are somewhat minimized due to the reduced ability to modulate the output of the source.

"That is the output impedance of the amp is around a few thousand ohms."

I'll give you the benefit of doubt here and assume this was a mistake. The output of an amp should be infinitesimally low, not around a few thousand ohms. If such were the case, the amplifier would be loaded down and current limited before you played the first note. You probably meant to say that "the output impedance of the amp is around a few thousandth's of an ohm. Sean

You are correct about the amp output impedance.

I know I have an amp with a damping factor of 400 into an
8 ohm load. I got my ratio upside down.

The amp's output impedance is 20 milliohms.

Let me correct some more of my previous post:

Twin lead antenna cable is a 300 ohm transmission line.

Coax cable is the 75 ohm transmission line.

The connection between an audio cartridge and the phono
preamp is the 47k ohm transmission line.

Dr. Gregory Greenman

Well said in your explanation.

I was being "loose" in my use of the word "control". In a
way, a speaker that can modulate the amp is not under "control".

You are quite correct that in the case of the interface
between the speaker and amp - "impedance matching" is the
last thing you want to do - since as you state - it would
allow the speaker to transfer what limited energy it has
back to the amp.

It is desirable for the interface NOT to be "impedance
matched". That way - only the amp - which has more power
than the speaker - can dictate what is going on.

As you state - there is a purposeful impedance mismatch
at the amp / speaker interface so the speaker can't talk
back to the amp - but of course - the powerful amp can
make the speaker do its bidding.

Dr. Gregory Greenman
Still with us, Cc? Just be careful to home-demo ANY amp you try with a relatively high output impedence, like many tube amps. Simple.
Thanks for asking, Subaruguru. I am tracking, except for when I'm not! The main upshot of all of this is that I am going to want to find some coherent, organized introduction to these issues, as I would like to pull back the veil a little. Any suggestions? The links people have provided seem to start in the middle, given my starting position.

And, I also understand that there is the same kind of impedence issues with the TT/cartridge and the pre-amp. I assume the same principles are involved. Is that right?

But, whether I understand it all or not, I am really enjoying reading and appreciate people's input.

The interface between the TT/cartridge and the preamp does
have to be impedance matched. However, cartridge makers and
preamp makes have standardized this interface to be a 47k ohm
transmission line - so you don't have to do anything special.

Likewise, all your interconnects are 75 ohm transmission lines
which is also standardized - and therefore impedance matched.

However, for the power amp to speaker interface there is an
intentional mismatch - so as Sean states - the speakers
can't feedback to the amp and modulate it.

For this interface, you check the specs of the power amp
or receiver - that will tell you the minimum load impedance
that the particular amp will handle properly.

Best wishes.

Dr. Gregory Greenman
...if it's a tube amp ask for the oftentimes dirty little secret of it's output impedence. This will give you a hint as to how much relative frequency response trampolining can occur compared to another amp. Again, you have to home-demo WITH your speakers. Don't make a career of reading about this...it'll come down to how the amp mates with your speakers. If you become obsessed with amplifier output impedence just get a nice ss amp....
Hi Subaruguru,

The only problem is that I am probably going to buy it over audiogon, so home audition may not be an option.... I had another thread earlier asking people who have the same speakers what amps they have found works well, and of course, got several differing opinions! I got a book by Robert Harley that explains some of the electronic basics, so I am learning, slowly. Thanks again for your help and input.