[Chris J]

Quote:

 

You can currently buy 16 ohm speakers off the shelf from a variety of sources (speakers for guitar amps, hi-fi, and automotive) and historically speakers were made in a VERY VERY wide variety of impedances, with some going as high as several hundred ohms. 
 
Regarding SS amps: none of that mattered when everything was transformer coupled. You just got the transformer wound for the speaker you had. When speakers were almost only sold in big integrated boxes (you know the old style tube radios) this was not difficult or even prohibitive. 
 
Regarding why we dont have other impedances TODAY I'l ask you a question, and tell you the answer. 
Are 4 and 8ohms better, or is that just what non-technical users are familiar and comfortable with? Its the later. The best item for the job depends on the job. Period. Anyone who gives you a default answer on BEST is a fool. People default to 4&8ohms because that is what they are familiar with. 

I would argue that 8 ohm speaker are standard because most audio power amplifiers are optimized to work best into 8 ohm speakers.
 
BTW, this has nothing to do with impedance matching or Maximum Power Transfer.
 

 

[nikongod]

 
Quote:

I would argue that 8 ohm speaker are standard because most audio power amplifiers are optimized to work best into 8 ohm speakers.
 
BTW, this has nothing to do with impedance matching or Maximum Power Transfer.
 

 
Based on what if not maximum power transfer? 

 

[Chris J]

Quote:

 
 
Based on what if not maximum power transfer? 

OK
Deep breath.
Not too sure what you know and what you don't know...........so forgive me if this sounds like a lecture

 or if I sound like Mr. Spock.
Maximum Power Transfer means (as you probably know) the Source Impedance = the Load Impedance.
In our case this would mean the headphone impedance would be equal to the headphone amplifier output impedance.
The voltage divider principle would apply here.
As you know no doubt know:
If headphone impedance = Amp output impedance then you would have poor damping factor AND poor efficiency in the headphone/amp interface.
In a loudspeaker system most loudspeakers have an average impedance of approx. 8 ohms (just stating the obvious) and the power amplifier has an output impedance in the range of 0.1 ohm or less. 
When the audio power amplifier manufacturers design a 100 Watt amp to work with an 8 ohm speaker they use a power supply with enough voltage and current to provide 100 watts (for example) and select output transistors to deliver enough current.
 In addition, Audiophile amplifier manufacturers usually design in enough capacity to also support a 4 ohm load or even lower.
How do I know this? I'm an EE.
Regards, C
    
.......any thoughts on the Nikon D90? I'm thinking about getting one.
 
 

 

[nikongod]

Quote:

OK
Deep breath.
Not too sure what you know and what you don't know...........so forgive me if this sounds like a lecture

 or if I sound like Mr. Spock.
Maximum Power Transfer means (as you probably know) the Source Impedance = the Load Impedance.
In our case this would mean the headphone impedance would be equal to the headphone amplifier output impedance.
The voltage divider principle would apply here.
As you know no doubt know:
If headphone impedance = Amp output impedance then you would have poor damping factor AND poor efficiency in the headphone/amp interface.
In a loudspeaker system most loudspeakers have an average impedance of approx. 8 ohms (just stating the obvious) and the power amplifier has an output impedance in the range of 0.1 ohm or less. 
When the audio power amplifier manufacturers design a 100 Watt amp to work with an 8 ohm speaker they use a power supply with enough voltage and current to provide 100 watts (for example) and select output transistors to deliver enough current.
 In addition, Audiophile amplifier manufacturers usually design in enough capacity to also support a 4 ohm load or even lower.
How do I know this? I'm an EE.
Regards, C
    
.......any thoughts on the Nikon D90? I'm thinking about getting one.
 
 

In a crop-sensor Id lean towards the D7000 or maybe 5100 unless you need a feature that you can only get on the D90. The high ISO performance and dynamic range of the newer bodies totally dominates the older digital cams. 
 
I thought we weren't talking about maximum power transfer, specifically ignoring it per your post at the top of the page. 
 
Skipping that, and the next bit about the same thing with headphones we get to the part about a hypothetical SS amplifier with a 100watt power rating, and a 0.1ohm output impedance. 
What about this amplifier would not work better with a 16ohm speaker than an 8ohm one? 
The electrical damping factor would be better with the 16ohm speaker than the 8ohm speaker. 
You stated that the transistors are selected to supply enough current into 8ohms, so why wouldn't they supply adequate current for 16ohms? 
The absolute max power would probably (although there are conditions where this does not apply) be lower with the 16ohm speaker than the 8ohm one, but if the amp had decent overhead/safety factor in its power rating this should not be a problem. 
So whats left? Various distortions? Crosstalk? You can find pretty solid numbers for both in chip-amp data sheets, and from a few manufacturers who specify the performance of their amplifiers into 16ohms. 
 
Why 8ohms:
 
8ohm speakers became the default choice for a few reasons, none of which have really anything to do with being better, and most have to do with being cheaper. 
 
One of the reasons was that on less expensive consumer grade transformer coupled tube amps a popular place to save a few bucks was in the transformer. Aside from poor quality in general, the various output taps seen on better amps were dropped. The designers of this cheap gear were faced with a decision of what 1 tap to use. 8 ohms was the most common choice between the also common 4 and 16 ohms, so that's what they picked. A few dollars less sells amps all day long... people quickly realized that their new amps only had taps for 8ohm speakers so that's what people bought. 
 
Tube amps with multi tapped output transformers dont have these problems, and work GREAT into any load they have a tap for, but the extra taps add cost to the amp and most people just dont care. The few people who did care enough to spring for extra taps wanted all sorts of cool in their amps (McIntosh, Harmon-Cardon, Leak, and others) , which drove the cost even higher so aside from a fairly small niche the 4 and 16ohm speakers had a foot in the grave before SS even became the defacto-standard. 
 
When SS amps hit the scene in force they really put the nail in the coffin for both the 4 and 16ohm speaker. 8ohm speakers generally play a bit louder on a given SS amp than a 16ohm one. It is a well known fact that a few db louder is perceived as better by the vast majority of people so that was the end of the 16ohm speaker. 4 ohm speakers generally go a little louder still on the same SS amp, but there weren't many 4-ohm stable SS amps until well after 8ohms became basically *the* standard. The lack of 4ohm stable amps killed the 4ohm speaker - once a speaker gets a rap for blowing up amps (even though its because the amp is crap) that reputation gets around, and sticks. After this point speaker designers had basically no choice but to build "8ohm" speakers. 
 
What sucks about the whole thing is that speaker designers are forced to make "8" ohm speakers with extremely low reactance so that they dont anger the amplifier designers & builders. Building a speaker is a BUNCH harder than building an amp, the amp guys should bow humbly to the needs of the speaker builders, but it never happens. 

 

[Chris J]

Quote:

In a crop-sensor Id lean towards the D7000 or maybe 5100 unless you need a feature that you can only get on the D90. The high ISO performance and dynamic range of the newer bodies totally dominates the older digital cams. 
 
I thought we weren't talking about maximum power transfer, specifically ignoring it per your post at the top of the page. 
 
Skipping that, and the next bit about the same thing with headphones we get to the part about a hypothetical SS amplifier with a 100watt power rating, and a 0.1ohm output impedance. 
What about this amplifier would not work better with a 16ohm speaker than an 8ohm one? 
The electrical damping factor would be better with the 16ohm speaker than the 8ohm speaker. 
You stated that the transistors are selected to supply enough current into 8ohms, so why wouldn't they supply adequate current for 16ohms? 
The absolute max power would probably (although there are conditions where this does not apply) be lower with the 16ohm speaker than the 8ohm one, but if the amp had decent overhead/safety factor in its power rating this should not be a problem. 
So whats left? Various distortions? Crosstalk? You can find pretty solid numbers for both in chip-amp data sheets, and from a few manufacturers who specify the performance of their amplifiers into 16ohms. 
 
Why 8ohms:
 
8ohm speakers became the default choice for a few reasons, none of which have really anything to do with being better, and most have to do with being cheaper. 
 
One of the reasons was that on less expensive consumer grade transformer coupled tube amps a popular place to save a few bucks was in the transformer. Aside from poor quality in general, the various output taps seen on better amps were dropped. The designers of this cheap gear were faced with a decision of what 1 tap to use. 8 ohms was the most common choice between the also common 4 and 16 ohms, so that's what they picked. A few dollars less sells amps all day long... people quickly realized that their new amps only had taps for 8ohm speakers so that's what people bought. 
 
Tube amps with multi tapped output transformers dont have these problems, and work GREAT into any load they have a tap for, but the extra taps add cost to the amp and most people just dont care. The few people who did care enough to spring for extra taps wanted all sorts of cool in their amps (McIntosh, Harmon-Cardon, Leak, and others) , which drove the cost even higher so aside from a fairly small niche the 4 and 16ohm speakers had a foot in the grave before SS even became the defacto-standard. 
 
When SS amps hit the scene in force they really put the nail in the coffin for both the 4 and 16ohm speaker. 8ohm speakers generally play a bit louder on a given SS amp than a 16ohm one. It is a well known fact that a few db louder is perceived as better by the vast majority of people so that was the end of the 16ohm speaker. 4 ohm speakers generally go a little louder still on the same SS amp, but there weren't many 4-ohm stable SS amps until well after 8ohms became basically *the* standard. The lack of 4ohm stable amps killed the 4ohm speaker - once a speaker gets a rap for blowing up amps (even though its because the amp is crap) that reputation gets around, and sticks. After this point speaker designers had basically no choice but to build "8ohm" speakers. 
 
What sucks about the whole thing is that speaker designers are forced to make "8" ohm speakers with extremely low reactance so that they dont anger the amplifier designers & builders. Building a speaker is a BUNCH harder than building an amp, the amp guys should bow humbly to the needs of the speaker builders, but it never happens. 

I'm looking for a Nikon which can use my old AF lenses. I'm talking about lenses which are approx. 7 years old. 24 mm, 50 mm and 105 mm. I understand that the the 5100 does not support old AF lenses?
 
If an amplifier was designed to max out at 100 Watts into an 8 ohm load, it would max out at approx. 50 Watts into a 16 ohm load. In actual practice you would get slightly more than 50 Watts because you would get less voltage drop across the output transistors and the power supply would sag less.   So it would work fine into a 16 ohm load but draw less power because the 16 ohm speakers would draw less current.
I agree, damping factor would be higher too.
Poorly designed amplifiers can become unstable when driving a speaker system with too much reactance, for example QUAds.

 

[Tribbs]

Quote:

Hi,
Please refer to this:
Go to "Head Fi" Home page
scroll to bottom of page
you will see a list of articles
click on "all articles"
then go to page 3 of "all articles"
on page three you will find a link to an article called "Headphone Impedance"
this may go some way to answer your question
<snip>

 
Ah ha!  This article > http://www.head-fi.org/a/headphone-impedance
 
Chris and others, thanks for your posts.  Not at all taken as lecturing.  It is discussions such as these that I enjoy the most.
 
If I may, I would be most curious of your (and others) opinions about Dr. Gilmore's DynaLo design with Grados.

 

[Chris J]

Quote:

 
Ah ha!  This article > http://www.head-fi.org/a/headphone-impedance
 
Chris and others, thanks for your posts.  Not at all taken as lecturing.  It is discussions such as these that I enjoy the most.
 
If I may, I would be most curious of your (and others) opinions about Dr. Gilmore's DynaLo design with Grados.

Thanks for the link to the headphone amp.
A very interesting design, he obviously knows what he is doing.
The real short answer is:
In some ways it is a fairly simple circuit, but very, very well thought out.
I like his thinking: first create a good design, then add a little feedback.
OTOH, it is a tremendous amount of voltage and current available to drive a headphone!
I suspect 0.5 Watts would destory a pair of Grados, I guarantee it would deafen you!
His suggestions for running a pair of SR-80s sounds a bit like overkill for a pair of $100 'phones!
 
More later.........Chris
 

 

[Tribbs]

Speaking of overkill, here is a link to interesting measurements of a Kevin Gilmore designed "DynaHi".
It is a follow-on design based on Gilmore's "Dynalo".
http://www.ibiblio.org/tkan/audio/dynahi.html
 
FYI - another DynaLo thread:
http://www.head-fi.org/t/461997/lost-link-to-a-dynalo-amp

 

[Chris J]

Quote:

Speaking of overkill, here is a link to interesting measurements of a Kevin Gilmore designed "DynaHi".
It is a follow-on design based on Gilmore's "Dynalo".
http://www.ibiblio.org/tkan/audio/dynahi.html
 
FYI - another DynaLo thread:
http://www.head-fi.org/t/461997/lost-link-to-a-dynalo-amp

Good lord!
It "only" dissipates 70 Watts?
Mindboggling...................it puts out a crazy amount of voltage and power!
If this is a desk top amp, you better have a BIG desk.
 
Thanks Tribbs, that is insane!
C
 

 

[nikongod]

if it were a real amp it would dissipate 70W...... (wait for it)..... in the tube heaters. 

 

[Chris J]

 
Quote:

if it were a real amp it would dissipate 70W...... (wait for it)..... in the tube heaters. 

 
Tube amps get a special efficiency exemption!

 

[rohan575]

even class A amps waste more than 70% of the power just to keep them properly biased and in class A opertation. What matters is the actual power delivered to the headphone which is given by 
   E X I cos(x) where cos(x) is the power factor which makes life bad for EE guys.
 
Long live tubes (with better,faster MOSFET's coming up i dont think they will)
Rohan

 

[bcschmerker4]

This was a lively debate sometime before I joined Head-Fi™, and I consider many of its points still valid.

I actually think of headphones in the 600Ω range as medium-impedance.  For me, truly high-impedance headphones have voice-coil impedances of 2 - 10 kΩ inclusive, center-tapped, for direct feed from paired 6K6 or similar tubes in push-pull, and use 6- or 7-pin Cannon® or functionally-similar connectors to allow application of +Vpp to the taps, plus a dedicated protective ground.

Modern electrostatic headsets are a special case of the high-impedance headphone, as, being capacitive rather than inductive in nature, they are of infinite DC resistance and require one conductor for each diaphragm and two for the corresponding field grids, plus protective ground.

 

[Redkitties]

So Low impedance is good?

 

[bcschmerker4]

  So Low impedance is good?

Depends on your hardware.  Most portable devices are voltage-limited by their power supply and therefore require headphones of under 64Ω.  I'm after a medium-impedance set myself, for an Asus® CM1630-06 that packs a XONAR® Essence™ STX™, as its Texas Instruments® TPA6120A2 headphone amplifier (Zs = 10.7Ω) is spot-on for a 120 Ω headphone consistent with the design specification proposed in IEC 61938-1996.  And old tube hardware wants greater than 500Ω due to the inherent very-low-current characteristics of 6V6GT, EL84, and similar beam power tubes, even with output transformers.