YEA CopperTop
Hooray for CopperTop! Time, frequency, and energy are the three domains for amplifiers and transducers. Be warned if anyone post outside the frequency domain on DIYaudio expect everything but letter bombs to show up railing against anything but frequency measurements. Never mind things like current sources being orders of magnitude slower in time than a good resistor or the propagation time through an amplifier to close the global feedback loop or the fact output load is very well reflected back through output transistors used as emitter followers to the last voltage gain stage. These things are not to be discussed and preferably not even mentioned falling into categories of blasphemy.
The lone thread where this seems to not be true is the http://www.diyaudio.com/forums/solid-state/154070-watt-sucking-fireball-series.html
Hooray for CopperTop! Time, frequency, and energy are the three domains for amplifiers and transducers. Be warned if anyone post outside the frequency domain on DIYaudio expect everything but letter bombs to show up railing against anything but frequency measurements. Never mind things like current sources being orders of magnitude slower in time than a good resistor or the propagation time through an amplifier to close the global feedback loop or the fact output load is very well reflected back through output transistors used as emitter followers to the last voltage gain stage. These things are not to be discussed and preferably not even mentioned falling into categories of blasphemy.
The lone thread where this seems to not be true is the http://www.diyaudio.com/forums/solid-state/154070-watt-sucking-fireball-series.html
thanks for the link!! i had meant to revisit that thread when i had enough learning under the belt to start to make sense of it, but forgot all about it. will the WSF pic directly linked become a museum piece of what could have been, but never was?
very interesting discourse and i'm very intrigued by that differential input. also the description of why N and P are so different.
not taboo for me, but rather just what the doctor ordered
thanks again
very interesting discourse and i'm very intrigued by that differential input. also the description of why N and P are so different.
not taboo for me, but rather just what the doctor ordered
thanks again
@ sreten: thanks for taking the time to reply, but i dont think you understood my question and BTW the 'standard circuit' for attenuating a speaker output to drive headphones is not an issue
for starters, i dont own the headphones in question, nor do i really wish to experiment with raising the output impedance of my dedicated headamps, which are all solid state and the best ones to my ears are pretty much zero. my headphones are 300ohms HD600 and 28ohms JH-13 custom fit in-ear monitors (but at low frequencies these present much less than 10ohms as there is 6 drivers per ear and a 3 way XO in a W||W-M||M-T||T setup).
ive got both of these situations fairly well under control and do not really want for better, just playing with new. I wouldnt touch a circuit like you describe as it seems just something to write into the specs for a reciever rather than representing any real effort to provide a headphone output., i'm not sure ive ever actually used one, being that i'm 37 and my entry into higher quality headphones came about the same time as getting into headphone amps
my question was more a wish for understanding, why someone would want to purposely design an amp with poor damping factor to drive 28ohms single driver headphones with a pretty much open back wooden cup and thus little by way of mechanical or atmospheric impedance. speakers i can understand how there might be more to it as far as interactions in this regard, but headphones and in particular these ones, the only factor i can think of would be that the drivers and frame are not so well physically damped and thus set up a resonance that way.
so i'm a bit puzzled as to how they could be driven well this way, let alone better, but i'm open to learning something new if there is something i havent taken into account, so i asked
for starters, i dont own the headphones in question, nor do i really wish to experiment with raising the output impedance of my dedicated headamps, which are all solid state and the best ones to my ears are pretty much zero. my headphones are 300ohms HD600 and 28ohms JH-13 custom fit in-ear monitors (but at low frequencies these present much less than 10ohms as there is 6 drivers per ear and a 3 way XO in a W||W-M||M-T||T setup).
ive got both of these situations fairly well under control and do not really want for better, just playing with new. I wouldnt touch a circuit like you describe as it seems just something to write into the specs for a reciever rather than representing any real effort to provide a headphone output., i'm not sure ive ever actually used one, being that i'm 37 and my entry into higher quality headphones came about the same time as getting into headphone amps
my question was more a wish for understanding, why someone would want to purposely design an amp with poor damping factor to drive 28ohms single driver headphones with a pretty much open back wooden cup and thus little by way of mechanical or atmospheric impedance. speakers i can understand how there might be more to it as far as interactions in this regard, but headphones and in particular these ones, the only factor i can think of would be that the drivers and frame are not so well physically damped and thus set up a resonance that way.
so i'm a bit puzzled as to how they could be driven well this way, let alone better, but i'm open to learning something new if there is something i havent taken into account, so i asked
Last edited:
Hi,
My point was what driving impedance are the headphones designed for ?
according to here : Between the Ears: the art and science of measuring headphones Page 3 | Stereophile.com
If they are designed to work well with the poor typical amplifier headphone
output, then it stands to reason they might work well with a high output
impedance headphone amplifier, or at least be a different flavour.
It is conjecture, but FWIW if a driver is amenable to "current drive",
some of the distortion mechanisms of moving coil drivers are bypassed.
Anyway some info here : Variable Amplifier Impedance
if someone wants to build current drive type amplifiers properly.
rgds, sreten.
My point was what driving impedance are the headphones designed for ?
according to here : Between the Ears: the art and science of measuring headphones Page 3 | Stereophile.com
If they are designed to work well with the poor typical amplifier headphone
output, then it stands to reason they might work well with a high output
impedance headphone amplifier, or at least be a different flavour.
It is conjecture, but FWIW if a driver is amenable to "current drive",
some of the distortion mechanisms of moving coil drivers are bypassed.
Anyway some info here : Variable Amplifier Impedance
if someone wants to build current drive type amplifiers properly.
rgds, sreten.
A guess: most headphones don't need to be efficient so plenty of mechanical damping can be included, so little or no external electrical damping is needed.
Hi,
Agreed. They certainly don't do bass by having low Fs's, so damping must be high.
Between the Ears: the art and science of measuring headphones Page 4 | Stereophile.com
Implies generally 120 ohm will give warmer bass around 100Hz, for 3 of 4
headphones tested, the problem with the 4th is which response is right ?
rgds, sreten.
Grado themselves have a headphone amp called the RA-1, which is basically a glorified CMOY in a wooden box, so one would assume they think near zero is the best, so very much a voltage source vs the current source some Grado fans often seem to recommend. now most appear to be recommending a tube or hybrid, so perhaps they just prefer this subjectively and are making up psuedo science to support their choice =)
the 120r standard should be a thing of the past, headphones, including studio headphones are very rarely above 250ohms and most less these days; it was only ever a stopgap solution at best
thats what i would have thought too. Grados are handmade wooden cup headphones that have changed little since the company started, all models are 28ohms and fairly open by design, perhaps they really arent damped very well physically so need this electrical damping to prevent resonance caused by this fault. unsurprisingly when i posed the question this way i didnt receive a response =)
sorry for the OT, but i thought it an interesting adjunct to the topic to discuss the effects of electrical damping on a system that is devoid of many of the physical problems associated with loudspeakers
the 120r standard should be a thing of the past, headphones, including studio headphones are very rarely above 250ohms and most less these days; it was only ever a stopgap solution at best
thats what i would have thought too. Grados are handmade wooden cup headphones that have changed little since the company started, all models are 28ohms and fairly open by design, perhaps they really arent damped very well physically so need this electrical damping to prevent resonance caused by this fault. unsurprisingly when i posed the question this way i didnt receive a response =)
sorry for the OT, but i thought it an interesting adjunct to the topic to discuss the effects of electrical damping on a system that is devoid of many of the physical problems associated with loudspeakers
Last edited:
Off the subject but about headphone amps... number one problem seen here with many many designs is very excessive DC offset. Can recall a very expensive tube unit >$700 MSRP with 65mV DC offset with the headphones plugged in!!! Talk about forcing the diaphragm far from center creating lots of even ordered distortion, large (>20mV) of DC offset is the norm in most of these headphone amp designs with reviewers lauding the wonderful music coming from these badly distorting due to DC offset designs. A very few get it right though with less than 1mv DC offset. 20mV is almost to high for a regular power amp of quality.
Hey, no problem, I think it's quite a good adjunct.
I've long been frustrated that headphone outputs on receivers are (were?) typically power amp outputs padded by a big resistor, meaning every different headphone's response would vary because of the amp. For portables, the output impedance is high to protect inexpensive output chips, and the same problem occurs. Buying headphones then becomes less about the intrinsic sound quality of the 'phones and instead becomes a crapshoot of synergy-an expensive crapshoot because the only way to do it is keep buying different headphones.
I'd be more interested in getting a very portable headphone amp if I thought the output impedance would be low, but I'm not recalling that is ever tested.
We should bug the various magazines and websites to start testing output impedance of headphone jacks, and ideally publish frequency response curves with several sets of simulated headphone impedances. THAT would raise consciousness about this issue!
Kinda like what Car Audio and Electronics did here:
Alpine PDX-F4 Review - Amplifier Reviews - Car Audio and Electronics
Look at the "Typical Class D amplifier" curves about 2/3 of the way down the page (search "a big peak occurs"). You can really see the high frequency ugliness of the usual passive output filter Class D design.
As a speaker engineer, I'll say DC at the amp is NOT a good thing.
Ugh, now I'm wondering how much DC is coming out of these portables...
Headphone jacks typically don't have much power available, so headphones DO need to be sensitive.
I haven't done transducer work with headphones. My loudspeaker experience makes me feel headphones have similar problems and so lots of damping would mean less bass. But we need a headphone engineer to comment.
Oh, voice coil transducers don't really work with current drive. The resonance makes a big bass peak. You would need to integrate an amp with special feedback into the speaker/headphone.
Last edited:
Mmmm. Maybe some at boutique firms, but none of the many big-company folks I know in the business. They all have big, typical receivers or amps they listen with. Source impedance is assumed to be low and typical, and they don't worry about it at all. Rather as Bob Cordell said:
So for the vast majority of speakers to work as designed, your amp's output impedance should be tiny. I don't know what headphone engineers do...tear out their hair? The transducers themselves still want low output impedance.
http://www.amazon.com/Designing-Audio-Power-Amplifiers-Cordell/dp/007164024X
Last edited:
Well, there's another "can of worms"!
How about when the amplifier starts to clip? The damping just dies? Or does it depend a lot on the amplifier design?
By the way, your book has good reviews, congrats!
http://www.amazon.com/Designing-Audio-Power-Amplifiers-Cordell/dp/007164024X
For most of the drek out there yes. But i'm diying because i want something better than i can buy. And i like to take advantage of the advanrages of high output impedance.
dave
Might the idea of higher damping factor be applied to in other areas of technology? Maybe our garage doors and car starter motors have been designed without the correct 'voicing', and a little series resistance would do wonders for their performance? After all, electrical damping, and therefore low output impedance, is only needed if there is insufficient mechanical damping, right..?
10's of mW is a lot of power when the transducer is almost in your ear. With really sensitive headphones, such as used with crystal sets, we can hear uW. Hi-fi units can afford to be inefficient by comparison.head_unit said:
In most cases amplifiers clip when they hit a voltage limit, so the impedance will be low but perhaps without the assistance of feedback.
Advantages? Do you mean that literally, or do you mean you prefer the sound of amps whose output impedance happens to be high? (like tubes)
Literally. Much less sensitivity to power compression, back-EMF for starters... voice coil driven devices are driven bt current, there are decided advantages to directly current driving them as opposed to using the driver impedance as a voltage to current converter.
http://www.diyaudio.com/forums/vend...ook-current-drive-loudspeakers-published.html
Tubes are often easier to get high impedance out of, but we have a couple SS variable transimpedance amplifiers here that are real eye openers.
dave
Hi head_unit,
All bets are off for DF when an amplifier clips, and it will depend on the design. Keep in mind that negative feedback has a great influence on DF, and NFB is gone when the amplifier clips.
Also, recognize that an amplifier can clip because it runs out of voltage or because it runs out of current. In particular, if it runs out of current due to a current-limiting circuit turning on, output impedance will usually go very high, since the current-limiting circuit tends to turn the output stage into a current source. When that happens, the counter EMF from the energy stored in the loudspeaker can have its way and create a significant voltage pulse, possibly damaging a tweeter.
Thanks for your kind words about my book!
Cheers,
Bob
I was under the impression that current signal transfer is much more susceptible to interference than voltage signal transfer, thus the lengths gone to to keep the IV stage in DACs as close to the convertor as possible; so it would seem odd to make what is typically the longest cable run/loop in the system (the speaker cables) a current mode transfer.
maybe ive made an incorrect assumption there, this isnt a statement of fact.
maybe ive made an incorrect assumption there, this isnt a statement of fact.
Ah, I see, you're talking about a completely different amp concept. I would agree current drive can have a lot of advantages.
Implementations I'd seen previously required sensors on the speaker-does the book discuss something that will work with any speaker? Or must be designed as a system?
- Status
- Not open for further replies.