Questionable Article on Impedance

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Don't believe all you read.

What most punters need is 50 laps with Ohm's Law, DC networks.

That article wanders off to frequency-varying impedance (which we usually avoid) and some lab-only 600 Ohm Standard (real longlines were 100-900 Ohms).
 
Gosh...now what I understood is all in question...sigh.

Sorry for posting this. Mods, you can take this down if all it’s doing is spreading misinformation. Cause not being an EE, I really thought this was a good article.

Mike

Ps maybe it would be instructive for the geniuses here to explain what’s wrong with this for clarity...
 
Phrazemaster said:
Gosh...now what I understood is all in question...sigh.

Sorry for posting this. Mods, you can take this down if all it’s doing is spreading misinformation. Cause not being an EE, I really thought this was a good article.

Mike

Ps maybe it would be instructive for the geniuses here to explain what’s wrong with this for clarity...
I have ignored this so far...

I used to write a column about audio myths some 40 years ago, and not surprisingly all those myths I debunked (most of them) are still around.  So misinformation can have a life of its own. Being repeated in print or on forums like this helps keep them alive.

If you really want this deleted (you could remove the link) or we could delete the entire thread (seriously ask us).

If you want it picked apart, that could be done too, by any capable (multiple people here) willing to invest the time and effort (less people here). 

Right now, I can't even get myself to read it. Even after debunking it a fraction will still believe so any extra attention could perpetuate the misinformation (note: I have not read it myself but trust the judgement of Abbey, Paul, and others. )

JR

 
moamps said:
Excellent?
I think it would be worthwhile explaining the faults. Maybe they'll read our comments and fix it (or just take it down I suppose). But we need to provide concrete examples. So ...

Regarding moamps Figure 2 reference, there are multiple problems:

1) 600 ohms out Z into 600 ohms in Z yields a 6dB drop but for some reason the author uses a weird unrelated example of loading 600 ohms with two 600 ohms in parallel. That would NOT yield a 6dB drop. 600 ohms into 300 ohms is a 9.6dB drop.

2) I don't know of any gear other than some super old lab filter sets or tube outs that actually have 600 ohm outputs which means the whole example would probably cause more confusion that it resolves. Output Z specs sometimes refer to what load it can handle but 99% of the time outputs are zero ohms plus a build-out resistor which is usually 50-100 ohms.

3) The whole example is useless because most inputs are more like 10K (exceptions being stuff like old 1176 and Pultec gear which is 600 ohms) but with a 100 ohm source the line is effectively bridged not matched.

I have not read the article all the way through but here's another mistake:

Impedance & Frequency Response
The output impedance of a device and the capacitance of its connecting cable form a simple first-order low-pass filter, producing a 6dB/octave attenuation above a certain frequency. However, you need either quite a low output impedance or quite a long high-capacitance cable to bring the turnover of this filter into the audio band.

Clearly the author meant high not low. Low out Z would push the corner frequency up not down.
 
Ok I’m asking - mods please just delete this whole thread.

Sorry for confusion I spread. I thought I was learning something. All I learned is I don’t know anything.

Thx

Mike
 
Phrazemaster said:
Ok I’m asking - mods please just delete this whole thread.

Sorry for confusion I spread. I thought I was learning something. All I learned is I don’t know anything.

Thx

Mike

Nah, you changed your title to reflect there are errors and you didn't spread any confusion Mike. 
 
offering up alternate articles would turn around the situation and help others to further their knowledge along the lines of AC signal theory at this juncture of universal understanding of the universe,  can i say that legally?  :D
 
CJ said:
offering up alternate articles would turn around the situation and help others to further their knowledge along the lines of AC signal theory at this juncture of universal understanding of the universe,  can i say that legally?  :D
It’s a great idea.

Obviously needs to come from someone who understands this!
 
Phrazemaster said:
Sorry for confusion I spread. I thought I was learning something. All I learned is I don’t know anything.

That's not your fault. This article is there more than 15 years. That's the problem. 
 
squarewave said:
1) 600 ohms out Z into 600 ohms in Z yields a 6dB drop but for some reason the author uses a weird unrelated example of loading 600 ohms with two 600 ohms in parallel. That would NOT yield a 6dB drop. 600 ohms into 300 ohms is a 9.6dB drop.

And that illustration is confusing as it discusses 'signal power' that isn't really a 'thing' in line level interconnection where we are looking at the voltage and need only enough current to produce that at the load.
To be fair it does go on to discuss more standard Lo-Z outputs / Higher Z loads but I do see that the article could be misleading and the other technical flaws that have been highlighted.
 
Unless you are a circuit designer there is not a lot you need to know about impedance.

Resistors, as their name implies, resist the flow of electricity. The higher the value of the resistor the more it resists. A higher resistance needs a higher voltage across it to get the same current flow. This is ohms law:

I = V/R

A pure resistor does not care if the voltage across it is ac or dc. It works the same. Inductors and capacitors behave differently. Their apparent resistance varies with frequency. All real components have stray resistance, capacitance an inductance. Even cables have resistance inductance and capacitance.

When you mathematically combine the contributions from resistance, capacitance and inductance in a component you get its frequency dependent value called impedance. In more complex circuits you need combine all the contributions from all the  relevant components.

If you 'look into' the output of a line amp, mixer, microphone or any audio source you will see an impedance. This is called the source impedance or output impedance of the source. If you look into a mixer mic or line input, or and input transformer or any kind of input you will see another impedance. This is called the input impedance or load impedance.

The fun begins when you connect an output to an input. The audio is fed from a source impedance into a load impedance, both of which vary with frequency. Now you can see why designers strive to make input and output impedances constant over the audio band and 99% of the time you can assume they do so.

When you connect a source to a load (output to input) you want as much of the output signal to appear across the input. No sense throwing signal away. The simplest way to do this is to make the source impedance as small as possible ans the load impedance as large as possible. That way you create a pot divider with only a small fraction of a dB loss.

As a rule of thumb, if you make the load impedance 10 times the source impedance, the loss will be negligible. This is why mic  pre input impedances are often in the 1K5 region for 150 ohm (source impedance) microphones. Most modern op am based outputs have an output (source) impedance of 100 ohms or so. Most line inputs have an input impedance of about 10K which is 100 times the the source impedance of the signals fed to them. This is called a bridging load. You can connect several bridging loads to a typical line output without violating the 1 to 10 rule.

Things get more complicated if long runs of cable are involved but most of the tme that does not apply.

Hope this helps.

Cheers

ian
 
Well said and concise Ian. The exception would be old passive filter sets with 600 ohm output characteristic impedance, guitars which have ~15K output impedance and some guitar pedals like Fuzz which can have pretty high output Z. And of course impedance must be considered for speaker loads.
 
Phrazemaster said:
Ok I’m asking - mods please just delete this whole thread.

Sorry for confusion I spread. I thought I was learning something. All I learned is I don’t know anything.

Thx

Mike
I think it's too late for that as  forum members have started adding useful content.

No regrets. If you didn't share it, you wouldn't get this feedback. Keep learning...

JR
 
Back
Top