[obobskivich]

I was looking at the 70V speaker systems as while as ESL speakers, I apologise for any mistakes I made. I always thought that DF was the one that had to do with FR changes more so than just low Zout(since line outs ain't that low impedance either) but its just about impedance swings relative to the Zout is that right?

DF is a marketing creation, just like PMPO (okay, not *that* bad, but you get the idea).

FR changes are the result of the relationship between Zsource and Zload - line outs are higher impedance but consider that line inputs are also higher impedance. There's still a potential for mismatch depending on what you hook up. DF is some formula to make Zsource look pretty, iirc it's Zsource/Zload (or Zload/Zsource, whatever makes the bigger number basically) - it's completely arbitrary. It's easier to market things with a "single linear variable" than complex specs; which looks better:

Amp XYZ with output impedance 1 ohm.
Amp ABC with a DF of 5000.

XYZ can actually express a "DF of 5000" into a high impedance load, and ABC might have an output impedance even higher than 1 ohm because that's a unitless number. Make sense?

Actually Impedance is the algebraic sum of Resistance and Inductive & Capacitive Reactance,  so they are two inter-related things.

I meant "R" and "Z" are different.

 

[Chris J]

DF is a marketing creation, just like PMPO (okay, not *that* bad, but you get the idea).
FR changes are the result of the relationship between Zsource and Zload - line outs are higher impedance but consider that line inputs are also higher impedance. There's still a potential for mismatch depending on what you hook up. DF is some formula to make Zsource look pretty, iirc it's Zsource/Zload (or Zload/Zsource, whatever makes the bigger number basically) - it's completely arbitrary. It's easier to market things with a "single linear variable" than complex specs; which looks better:
Amp XYZ with output impedance 1 ohm.
Amp ABC with a DF of 5000.
XYZ can actually express a "DF of 5000" into a high impedance load, and ABC might have an output impedance even higher than 1 ohm because that's a unitless number. Make sense?
I meant "R" and "Z" are different.

Just because damping factor is a unitless number does not mean it is meaningless, damping factor MUST always be referenced to a load Z, if it isn't, it truly is meaningless.
But I do agree, damping factors greater than 50 or 100 are useless as they will be swamped by the cable impedance and contact resistance of the various connectors, plus whatever determental impedance may be contained inside a loudspeaker.

YOU need a Scotch, laddie!

 

[Chris J]

Quote:

high damping factor is from low output impedance and vice versa. usually anything above 70@8ohms is cause of too much/abuse of negative feedback. also you forget not all amps mention where that damping factor was measured. most amps only measure it at 1khz. some amps do however list damping factor of multiple frequencies like 20/50hz,1khz, and 10khz or don't even list the frequency measured at all. i don't even think headamps measure like speaker amps cause they never list the damping factor at any particular load at all and just say it's at that specific output impedance and that's it.

Rex,
Ouch!
That hurts!
Too much negative feedback?
You didn't do your homework, did you?  LOL
Keep in mind that headphone amps like the Grado headphone amp or the O2 or the CMOY which only use Op Amps will have approx. 100 dB of feedback at low frequencies.
 
Other than that you raise some excellent points:  whether we are talking about a headphone amp or a "speaker amp", to some degree, damping factor always varies with frequency.

 

[RexAeterna]

Rex,
Ouch!
That hurts!
Too much negative feedback?
You didn't do your homework, did you?  LOL
Keep in mind that headphone amps like the Grado headphone amp or the O2 or the CMOY which only use Op Amps will have approx. 100 dB of feedback at low frequencies.

Other than that you raise some excellent points:  whether we are talking about a headphone amp or a "speaker amp", to some degree, damping factor always varies with frequency.

well, that's why i have you here to correct me you silly goose! to learn and all that fun stuff. i'm still trying to wrap my mind around the last papers you shared with me

 

[obobskivich]

Just because damping factor is a unitless number does not mean it is meaningless, damping factor MUST always be referenced to a load Z, if it isn't, it truly is meaningless.
But I do agree, damping factors greater than 50 or 100 are useless as they will be swamped by the cable impedance and contact resistance of the various connectors, plus whatever determental impedance may be contained inside a loudspeaker.
YOU need a Scotch, laddie!

Yeah. If manufacturers stated real values, I'd be happy. There's an entire Rane Note devoted to this rant that I'll link instead of ranting:
http://www.rane.com/note145.html

Rex,
amps like the Grado headphone amp or the O2 or the CMOY which only use Op Amps will have approx. 100 dB of feedback at low frequencies.

So basically mean you the CMOY and the O2, right?

I would agree with your point here too - but I like Rex's thinking: just give me Zout and be done with it, I can do the math if I'm really curious. Again, mfgrs need to see the Rane Note. :angry_face:

 

[proton007]

Haha I guess the manufacturers don't really want customers to go sneaking around, do they?
Nor do they expect the intellect of the customer to be higher than that of a trained monkey. Everything is getting dumbed down. Technology is dictating human behavior, when 20 years ago it was the user who was responsible for getting the technology to work, program your own code, games etc.
They only care to mention the key specs most buyers know from hearsay. Stuff like PMPO, screen resolution, GHz, dual/quad etc.

 

[obobskivich]

Haha I guess the manufacturers don't really want customers to go sneaking around, do they?
Nor do they expect the intellect of the customer to be higher than that of a trained monkey. Everything is getting dumbed down. Technology is dictating human behavior, when 20 years ago it was the user who was responsible for getting the technology to work, program your own code, games etc.
They only care to mention the key specs most buyers know from hearsay. Stuff like PMPO, screen resolution, GHz, dual/quad etc.

Like I've said before: a nice number to slap on the box.

They're great at inventing new ones after they turn a given buzzword into absolutely useless dribble - e.g. "High Definition" and "HD" (there's even "HD" cars now!).

 

[Chris J]

Quote:

well, that's why i have you here to correct me you silly goose! to learn and all that fun stuff. i'm still trying to wrap my mind around the last papers you shared with me

 
Hey!
I'm not too comfortable with you calling me a silly goose!
LOL!
I think someone needs to write "Otala and Feedback for Dummies"!
Quote:

Yeah. If manufacturers stated real values, I'd be happy. There's an entire Rane Note devoted to this rant that I'll link instead of ranting:
http://www.rane.com/note145.html
So basically mean you the CMOY and the O2, right?

I would agree with your point here too - but I like Rex's thinking: just give me Zout and be done with it, I can do the math if I'm really curious. Again, mfgrs need to see the Rane Note.

 
NO!
You competely misunderstand me!
The Grado Headphone amp is completely different than the CMOY!
The Grado sounds wwarm and organic and luxurious because it comes in a warm, organic wooden case and costs a lot.
The CMOY sounds cold and bright because it comes in a cold metal case!
Other than that, they are almost identical............

 
I gotta agree with your "just gimme an Zout figure!" comment.
You can crank the damping factor number skyhigh by calculating it by using a 2,00 ohm 'phone.
You would need a big chart to keep the manufacturers honest:
damping fact WRT  headphone impedance WRT frequency, wouldn't you?
 
Is this what you hate about damping factor? 
 
BTW, go ahead and rant, you not, it's what Head Fi is for!

 

[Chris J]

Quote:

Haha I guess the manufacturers don't really want customers to go sneaking around, do they?
Nor do they expect the intellect of the customer to be higher than that of a trained monkey. Everything is getting dumbed down. Technology is dictating human behavior, when 20 years ago it was the user who was responsible for getting the technology to work, program your own code, games etc.
They only care to mention the key specs most buyers know from hearsay. Stuff like PMPO, screen resolution, GHz, dual/quad etc.

You forgot one of my favourite ones:
cheap cameras with crappy lenses but really high megapixel counts.

I need a beer!

 

[obobskivich]

Is this what you hate about damping factor? 

Yup. I don't have an issue with what it tries to relate, I have an issue with how it's been bastardized. Just like those super high-end cameras with their nifty plastic lenses and 400MP stamped on the box!

 

[Chris J]

Quote:

Yup. I don't have an issue with what it tries to relate, I have an issue with how it's been bastardized. Just like those super high-end cameras with their nifty plastic lenses and 400MP stamped on the box!

 
OK, now I can relate!
Thanks for the clarification.

 

[firev1]

Quote:

DF is a marketing creation, just like PMPO (okay, not *that* bad, but you get the idea).
FR changes are the result of the relationship between Zsource and Zload - line outs are higher impedance but consider that line inputs are also higher impedance. There's still a potential for mismatch depending on what you hook up. DF is some formula to make Zsource look pretty, iirc it's Zsource/Zload (or Zload/Zsource, whatever makes the bigger number basically) - it's completely arbitrary. It's easier to market things with a "single linear variable" than complex specs; which looks better:
Amp XYZ with output impedance 1 ohm.
Amp ABC with a DF of 5000.
XYZ can actually express a "DF of 5000" into a high impedance load, and ABC might have an output impedance even higher than 1 ohm because that's a unitless number. Make sense?
I meant "R" and "Z" are different.

Do manufacturers really do that? I hadn't see any in particular. The only DF measurements I saw are something like DF of XXXX@ 50 ohm, which is not useless at all.  But I agree that I would rather have the output impedance straight.

 

[Chris J]

Quote:

Yeah. If manufacturers stated real values, I'd be happy. There's an entire Rane Note devoted to this rant that I'll link instead of ranting:
http://www.rane.com/note145.html
So basically mean you the CMOY and the O2, right?

I would agree with your point here too - but I like Rex's thinking: just give me Zout and be done with it, I can do the math if I'm really curious. Again, mfgrs need to see the Rane Note.

 
Hey!
I finally got around to reading the "Rane Rant".
They raise more than a few very good points:
all these specs are meaningless unless they are fully referenced.
 
Another irritating spec:
Output impedance:  16-600 ohms.    What is that supposed to tell us?  Another meaningless spec! 

 

[Chris J]

Quote:

Great guide, a few notes though:
- Not all amplifiers decrease their output as Z increases; some are invariant (or nearly invariant) - the WA6 is a good example (it uses a transformer to accomplish this feat).
 

 
Just noticed this, but better late than never:
 
An amp with an output transformer is only load invariant if it is equipped with a series of taps to match the various load impedances.
 
For example, you can see this on a lot of Vacuum Tube power amps used to drive loudspeakers: there is typically a transformer tap for 4, 8 and 16 ohm loudspeakers.
Even this type of amp will deliver half power into a 32 ohm load as it doesn't have a 32 ohm output transformer tap.
Of course, this assumes that the load impedcance does not vary with frequency! hich it almost always does, especially in a loudspeaker system.

 

[obobskivich]

An amp with an output transformer is only load invariant if it is equipped with a series of taps to match the various load impedances.

Oh yeah, but every OT hp amp I've seen has such taps routed through a rotary switch (Cayin, Woo, etc) so I figured it was safe to assume as global.