Re: Questions on impedance matching


Ted Rook
 

My experience in the lab tends to roll-off above 20kHz so there isn't much I can usually
contribute to threads here however I wish to offer some remarks in response to soome fuzzy
information below.

Ted

On 21 May 2018 at 7:56, KeepIt SimpleStupid via Groups.Io wrote:

This

http://ward-beck.com/news/2014/03/07/600-ohms-is-it-neccessary/
might help you understand where the 600 ohms came from.

It's unlikely that a consumer grade piece of audio equipment would be able to drive a 600
ohm input.

In the commercial sense, connections are usually balanced XLR's using 600 ohm
transmission line.  You do have long cable to deal with.

In audio transmission lines are not used. Rather the system operates with low source
impedances and high load impedances, ratios of ten to one are common. This is true
for balanced and unbalanced operations.

The only real object is to have x volts p-p at the end of the cable, so you drive with the lowest
impeadance into something reasonable like 10K.  Tubes were an entirely different animal.

With a power amp, you have say an 8 ohm speaker and 100 W/ch, both channels driven.  If I
had an 8 ohm output Z at my amplifier that would not be very good.  So, I really need the
lowest output Z.  This is actually expressed as Damping factor.  A damping factor of 100,
means the output Z of the amp is 8 ohms/100.

Speaker impeadance is actually measured at 1 kHz and is usually about 6 ohms for an "8
ohm speaker".

It is the minimum value of the impedance that matters from the point of view of power
amplifier loading. Speaker impedances vary substantially over the audio frequency
range usually being a minimum at the frequency at which the driver low frequency
resonance occurs, often in the range 20-100Hz. A value of 6 ohms means the minimum
impedance is 6 ohms, this is not the impedance at 1kHz, which is probably much
higher, and irrelevant.

We also have losses in the wires going to the speaker.  We don't have sense leads in our
audio amplifiers.

Professional audio is usually more precise.  So, the output is assumed to be driving a 600
ohm balanced load.  That's not the case in consumer audio.

Balanced operation is common in professional audio but loads are 10k or higher. 600
ohm loads were encountered mostly in telephony in the days when it was analogue,
but not in audio where they were rare, and are now non existent.

==

A comment about scope impeadance.  Two standards exist of 1 M and 50 ohms.  They may
have different parallel capacitances.  1M || 22 pF is common.
Why the capacitance?  We try to create the perfect voltage divider.  So of the scope has 1 M
|| 22 pF and the probe has 1 M || with x-22 pF, then it's possible to create a near perfect
divider at "all" frequencies.
  By doing probe compensation (vary the capacitor on the probe with a square wave input
such that the waveform is square) does just that.  In return, we get a probe that sees 10 meg
ohms resistive and attenuates the signal by 10x usually.    The scope probe cable has
capacitance too.

==

You also need to remember a couple of things:
Twisting helps eliminate EMI because the EMI radiates into both conductors "equally" and the
differential receiver removes the common mode signal as well as the act of twisting.

Shielding reduces EMI.  Shields should be connected at one end only and this is typically the
signal source.

Thus, twisted-pair shielded cable is very common for process control and professional audio
too.

Ground loops are really our enemy.

A square wave by Fourier analysis requires infinate bandwidth to re-produce.
It's made up of the sum of the odd harmonics of sine waves.

Bandwith is usually measured at the -3db frequency or when the signal is down by ~0.707. 
there is a rule of thumb that relates rise time and bandwidth.  See:
https://www.edn.com/electronics-blogs/bogatin-s-rules-of-thumb/4424573/Rule-of-Thumb--1-
-The-bandwidth-of-a-signal-from-its-rise-time
Worry about stuff when it's important.

Is three resistances in series e.g. 3K equivalent to three 1K resistors in series?   I can argue
no and be right.

Does a moving wire in free space on earth develop a current?    The answer is yes.  The
practical answer is no.  It is a wire that's rotating in the Earth's magnetic field, so it generates
a current based on physics.  Have I measured such a current - Yes.

In a college glass we had a test and I successfully argued that my answer was correct.   His
answer was, "Your not supposed to know that yet."


On Friday, February 16, 2018, 6:00:25 PM EST, @Nielsentelecom
<@Nielsentelecom> wrote:

In telephony, bridging is a common concept. The impedance could be
anything. on repeater equipment, there is a switch setting if the unit
was at the end of the trunk, or in the middle somewhere. The settings
are; 150 ohm, 600 ohm, or 1200 ohm. The 600 is the usual setting. The
1200 ohm is considered a compromise setting. You have to be careful with
the settings. Setting 1200 ohm can cause overdriving an amplifier or
repeater with the usual accompanied distortion or 'ringing'. When
sending test tones to check level, or realigning a circuit, the
transmission test set had to be set properly to either 'terminated' or
'bridged' to measure the tone correctly. Otherwise a bad reading will
result.

I've always thought the term came from the schematic drawing of a
bridged circuit, 2 parallel lines and 2 perpendicular lines crossing the
parallel lines. It looks like a bridge at the crossing points.

I use the term 'bridged' to describe any paralleled devices. Such as
duplex wall outlets.


NielsenTelecom

Join TekScopes@groups.io to automatically receive all group messages.