Head-Fi.org › Forums › Equipment Forums › Sound Science › Damping factor. Why it is not always as important as some make it out to be.
New Posts  All Forums:Forum Nav:

Damping factor. Why it is not always as important as some make it out to be. - Page 7

post #91 of 128

TIM is no problem for a solid state amp even with crazy amounts of negative feedback provided it is fast enough (slew rate).

 

Btw, I read that the "otala amp" was eventually improved by increasing negative feedback to 30 dB in order to reduce THD.

 

Happy easter.

post #92 of 128
Quote:
Originally Posted by xnor View Post

TIM is no problem for a solid state amp even with crazy amounts of negative feedback provided it is fast enough (slew rate).

Btw, I read that the "otala amp" was eventually improved by increasing negative feedback to 30 dB in order to reduce THD.

Happy easter.

i see. thanks for the info! much appreciated!

yea, i'm not sure how much was used on harman kardon designs but i do know they used his designs and they are very nice clean sounding amps. i only knew they fellowed the ultra high bandwidth rule,so most their amps were dual-mono design using separate toroidal power supplies for each channel and were also high current designs is only thing i know of them.
post #93 of 128
Thread Starter 

Was just at my friends studio today who has one of my friend jacks amps. The ATP TA-25 all triode tube zero feedbackclass A1 limited amp. This is a very fine sounding amp & I prefered it to the Bryston amps my studio friend has. Slightly more lively sounding & closer to the sound of the real piano he had in his studio when a recording of the same piano was played back in the studio.

 

Both amps sounded very good but the tube amp ever so slightly better overall. This is in spite of a slight 120hz hum problem the he was having. Not sure if if it was due to an outside source or the A.C. filiments but we were able to get it down quite a bit though reposoitioning of cables. This amp is incredably sensitive to outside interferance but we got it down to where it was only barely audible.

 

Even with this hum the sound was exquisite & was not a factor when listening to music. Bare in mind this was with horn speakers (Altec Lansing Voice Of The Theater A7-500s). These speakers were located in the studio, not the controll room though the studio in this case can double as the control room to as he had a direct link to the controll room computer & mixer via keyboard, mouse & monitor in the studio itself. Just about everything that can be done in the control room can be done in the studio except patching in external processors & source switchingthrough the patch bay..

 

Jack is gone now bless his heart but I had the the pleasure of knowing him when he was alive as well as my studio friend at that time. At that time my studio friend was just building his studio.It is currently one of the most advanced small studios in the world. Much of the work he does is to help music students achieve thier goals both musically & professionally as well as make incredable sounding recordings for classical & jazz musicians.


Edited by germanium - 4/8/12 at 8:42pm
post #94 of 128
Quote:
Originally Posted by RexAeterna View Post


yes, you are right. i was gonna ask why at first but then i read up on it for a bit. i spoke too soon on my part. so what i really meant Dr Matti Otala believed in ''minimum'' negative feedback,but was also more focused on T.I.M and I.I.M distortion than THD cause he believed that T.I.M and I.I.M was up to 10x more audible to the human ear than THD and was one of the more important aspects of an amplifier's output stages. i don't know much, but that is what i read so far, so it's my fault on my part earlier. i also forgot to tell every here happy easter and hope you guys are enjoying yourselves.



Happy belated Easter.biggrin.gif

Since you cannot share your Easter Eggs, maybe you can share some links to those articles?

 

Matti Otala has been massively misquoted by many people.

Otala was not anti-feedback, he was trying to reduce or minimize T.I.M.

Keep in mind that feedback is used to keep DC offset under control too.

He recommended increasing bandwidth in Open Loop gain, this has the side effect of reducing reducing Global feedback in the audio bandwidth.

He also recommended improving the Open Loop linearity of the power amp input stage, i.e. adding Local Feedback to the input stage, you can argue he was reducing Global Feedback and replacing it with Local feedback within the amplifier.

 

I have a whole pile of Matti Otala's papers, I can scan them if you are interested. I wrote a research paper on Otala's work in 4th year University.
 

Quote:
Originally Posted by xnor View Post

TIM is no problem for a solid state amp even with crazy amounts of negative feedback provided it is fast enough (slew rate).

 

Btw, I read that the "otala amp" was eventually improved by increasing negative feedback to 30 dB in order to reduce THD.

 

Happy easter.


T.I.M. and slew rate (or S.I.D.   Slewing Induced Distortion) are two slightly different things.

As I said, reducing T.I.M. has more to do with improving linearity of the Power Amp's input stage.

Reducing S.I.D. has more to do with speeding up the dominant pole which has more to do with the second voltage gain stage in most solid state power amps.

Noted amp designer Douglas Self doesn't seem to adhere to Otala's thinking about T.I.M., but he does recommend increasing linearity of the Input Stage

 

To complicate things even more, feedback varies with frequency.    So 30 dB of feedback at what frequency? Oye Vay!

 

Dual mono construction is nice, but it has nothing to do with T.I.M. or S.I.D.

Yes, Otala also appears to be the guy who also recommended a good solid state amp should have enough output current to drive 4 ohm and even 2 ohm loads properly.
 

 

post #95 of 128
i would love to read your papers! be nice learning experience.

only thing i read on him was mainly forum like this i found in google

http://www.diyaudio.com/forums/solid-state/125541-matti-otala-amplifier-milestone-dead-alive.html

and this

http://home.online.no/~tsandstr/OtalaStory.htm

i found out bout him when through this web page

http://www.thevintageknob.org/harman_kardon-hk770.html

cause i was trying to find info on that power amp and the h/k 700 series set cause i acquired them through a trade with a local. was originally for the preamp but later on he called me and asked if i wanted the matching power amp and found them very impressive sounding and thought it personally sounded like a very accurate transparent set. i have more powerful power amps but i find the h/k 700 series one of the most accurate i ever heard so far, so it got curious when i read Dr Matti Otala was the amp designer.

even my local tech who been working on amps for past 30 years told me that harman kardon did make some very accurate gear in the past and they measure out very well objectively.
post #96 of 128

hmmm

 

 

Quote:
In electronics, impedance matching is the practice of designing the input impedance of an electrical load (or the output impedance of its corresponding signal source) to maximize the power transfer or minimize reflections from the load.

 

mostly unless this is a dynamic system it will only match at a certain desired frequency (characteristic impedance), same as with load lines, the impedance is designed to match at the frequency of interest for transmission without reflections

 

@ Germanium

 

Pass labs Alephs use feedback, besides which one are you talking about? an aleph is an active current source (best you look at the patent) that uses feedback from the load to dynamically control the bias, sound like no feedback?. how about degeneration (imo just another type of local feedback)? how about local feedback? there are very very few amps that use no feedback, thats a marketing term. perhaps you really mean no global feedback?

 

there are also quite a number of low feedback solid state amps coming about lately, mostly mosfet or power jfet designs and Nelson is definitely at the forefront, but he doesnt limit himself to that, i'd kill for some of those SITs

 

also, horns are being used for other reasons these days as well and many used with active or digital XO's and not just tube amps. they are being used to improve directivity and phase, even as part of the crossover by smoothing out the response of compression drivers. check out the unity horn concept by Tom Danley for a wacky point source horn speaker. (you should see some of the fringe DIY versions!

images?q=tbn:ANd9GcTFra66IGIK471FMfFg4MGdD8kecvlRjSsUFJQRKXMXxSy1MlFBPQ

 

myself I prefer high damping factor and i'm not sure how much of an effect it has on well damped modern headphones (unless its too high), but it should be remembered that transducers work in the current domain. i'm no damping expert, i'm only recently getting into building loudspeakers, having focused on building headamps, dacs etc up till the last 12 months or so (DIY, not for profit) but I do know that generally i prefer high damping factor, not much point striving for utterly ridiculously high Df, as it'll be swamped by cables and if they exist, crossovers, but a reasonable level is preferred in my experience. high feedback shouldnt be used to cover up less than proficient or unstable design, but I see nothing wrong with judicious use of it.

 


Edited by qusp - 4/9/12 at 1:57pm
post #97 of 128
Thread Starter 

Yes I was mistaken on passs labs Aleph being zero feedback The first version had more feedback than later versions  & better specs but many prefered the later versions in spite of poorer specs. I was going by memory & My memory was not good on this one. Sorry.

 

This of coarse does not negate what I was saying about transformer coupled class A tube amps with zero feedback matching the load with thier output impedance nor that all transformer coupled tube amps must be reasonably impedance matched to the speaker before adding feedback. Yes feedback lowers the output impedance as the speaker sees it but the tube still needs it's matching load as it sees it to perform best even with feedback. The speaker may see the output impedance as very low but the tubes see the load as if they are matched when properly setup.

post #98 of 128
Quote:
Originally Posted by germanium View Post

Yes I was mistaken on passs labs Aleph being zero feedback The first version had more feedback than later versions  & better specs but many prefered the later versions in spite of poorer specs. I was going by memory & My memory was not good on this one. Sorry.

 

This of coarse does not negate what I was saying about transformer coupled class A tube amps with zero feedback matching the load with thier output impedance nor that all transformer coupled tube amps must be reasonably impedance matched to the speaker before adding feedback. Yes feedback lowers the output impedance as the speaker sees it but the tube still needs it's matching load as it sees it to perform best even with feedback. The speaker may see the output impedance as very low but the tubes see the load as if they are matched when properly setup.



The problem I have with these statements are that they are way too vague.
Define " matching the load with their output impedance".
post #99 of 128
Quote:
Originally Posted by RexAeterna View Post

i would love to read your papers! be nice learning experience.

only thing i read on him was mainly forum like this i found in google

http://www.diyaudio.com/forums/solid-state/125541-matti-otala-amplifier-milestone-dead-alive.html

and this

http://home.online.no/~tsandstr/OtalaStory.htm

i found out bout him when through this web page

http://www.thevintageknob.org/harman_kardon-hk770.html

cause i was trying to find info on that power amp and the h/k 700 series set cause i acquired them through a trade with a local. was originally for the preamp but later on he called me and asked if i wanted the matching power amp and found them very impressive sounding and thought it personally sounded like a very accurate transparent set. i have more powerful power amps but i find the h/k 700 series one of the most accurate i ever heard so far, so it got curious when i read Dr Matti Otala was the amp designer.

even my local tech who been working on amps for past 30 years told me that harman kardon did make some very accurate gear in the past and they measure out very well objectively.



Sure,

I'll send you a PM and you can read all about the exciting adventures of TIM and his sworn enemy SID.

The HK 770! I Had one of those way back in the late 80s with the matching pre amp!
My first taste of hi end amps!
post #100 of 128
should of kept the h/k 770! very good performing amp. might be 65wpc conservatively but thanks to it's ultra wide bandwidth big dual power supply design it does not roll off the low and sub-bass like other amps do. it has tons of current reserves and very quick to pull out all the low-level detail needed across the full frequency spectrum. i like it so much i got it serviced by local tech and brought up to spec and selling off/trading my other amps. just keeping my yamaha m-45 as a back up amp and dedicated headamp(and cause i love yamaha 1980's power amps of course and find them very accurate as well).
post #101 of 128
Thread Starter 
Quote:
Originally Posted by Chris J View Post


The problem I have with these statements are that they are way too vague.
Define " matching the load with their output impedance".


I don't see what is so vague about what I said. It seems very clear to me especially if you read the whole context & not just little snippets like you seem to be doing or at least by what you are quoting.

post #102 of 128
Quote:
Originally Posted by qusp View Post

hmmm

 

 

 

mostly unless this is a dynamic system it will only match at a certain desired frequency (characteristic impedance), same as with load lines, the impedance is designed to match at the frequency of interest for transmission without reflections

 

@ Germanium

 

Pass labs Alephs use feedback, besides which one are you talking about? an aleph is an active current source (best you look at the patent) that uses feedback from the load to dynamically control the bias, sound like no feedback?. how about degeneration (imo just another type of local feedback)? how about local feedback? there are very very few amps that use no feedback, thats a marketing term. perhaps you really mean no global feedback?

 

there are also quite a number of low feedback solid state amps coming about lately, mostly mosfet or power jfet designs and Nelson is definitely at the forefront.

 

myself I prefer high damping factor and i'm not sure how much of an effect it has on well damped modern headphones (unless its too high), but it should be remembered that transducers work in the current domain.



Impedance matching has no application in low frequency transmission (i.e. audio frequencies).

Impedance matching refers to output impedance = load impedance.

In a power amp you want output impedance to be much lower than load impedance for most efficient power transfer and accurate voltage transfer.

In actual practice it is impossible to build an amp without feedback, local feedback is inherent to all transistors and tubes, for example, without feedback a tube or transistor would have infinite bandwidth.

I agree, degeneration is Local Feedback.

I'm sure the "no feedback in this amp" thing is another usless marketing term.

Anyone have any examples of a "No Feedback" amp?

 

Generally speaking, loudspeakers are designed to be driven by a voltage source.

The voltage is applied to the speaker, which causes current to flow, the change in current creates a varying magnetic field, which makes the voice coil move.

 



 

Quote:
Originally Posted by germanium View Post

Yes I was mistaken on passs labs Aleph being zero feedback The first version had more feedback than later versions  & better specs but many prefered the later versions in spite of poorer specs. I was going by memory & My memory was not good on this one. Sorry.

 

This of coarse does not negate what I was saying about transformer coupled class A tube amps with zero feedback matching the load with thier output impedance nor that all transformer coupled tube amps must be reasonably impedance matched to the speaker before adding feedback. Yes feedback lowers the output impedance as the speaker sees it but the tube still needs it's matching load as it sees it to perform best even with feedback. The speaker may see the output impedance as very low but the tubes see the load as if they are matched when properly setup.



OK, lemme try again:

 

match tube output impedance to transformer primary impedance

 

Transformer output impedance (secondary impedance) is less than load impedance.

 

Right?

post #103 of 128
Thread Starter 
Quote:
Originally Posted by Chris J View Post



Impedance matching has no application in low frequency transmission (i.e. audio frequencies).

Impedance matching refers to output impedance = load impedance.

In a power amp you want output impedance to be much lower than load impedance for most efficient power transfer and accurate voltage transfer.

In actual practice it is impossible to build an amp without feedback, local feedback is inherent to all transistors and tubes, for example, without feedback a tube or transistor would have infinite bandwidth.

I agree, degeneration is Local Feedback.

I'm sure the "no feedback in this amp" thing is another usless marketing term.

Anyone have any examples of a "No Feedback" amp?

 

Generally speaking, loudspeakers are designed to be driven by a voltage source.

The voltage is applied to the speaker, which causes current to flow, the change in current creates a varying magnetic field, which makes the voice coil move.

 



 



OK, lemme try again:

 

match tube output impedance to transformer primary impedance

 

Transformer output impedance (secondary impedance) is less than load impedance.

 

Right?



Without feedback in transformer coupled tube amps output impedance should match. Once you add feedback output impedance goes down as far as the speaker sees the source but the tube still sees the load as matched otherwise performance suffers. Transformer secondary output impedance should match load without feedback. It is ok that it doesn't after feedback is added so long as the tube still sees the load as matched from it's side. Note that you cannot match the load to the plate resistance of the tube without matching the output impedance to the speaker impedance before feedback is added. The plate load on the transformer primary is dependant on the turns ratio of the primary to secondary & when the turns ratio is correct the output impedance of the secondary will match the load impedance without feedback.

 

When I say no feedback I am refering to no loop feedback local or glabal. In tubes there is a type of feeback that is not loop feedback & is based on cathode resistors. This type is used in the cathodyne inverter & by nature of the output tubes used in my friends amp the power tube as they have direct heated AC filiment cathodes much like you would find in many transmitter tubes. This type of feedback is not loop feedback & there is no delay with it unlike loop feedback the goes back to the input grid. Most tubes & transistors have an internal feedback mechanism but this is not considered feedback in the same sense as added external loop feedback. So when I say that my friends amp has no feedback I'm talking of loop feedback of any kind local or global hense it is an accurate statement & not just marketing.

 

Impedance matching applies to all frequencies including audio frequencies though the reasons for it are different when dealing with audio frequencies as opposed to radio frequencies. You will always get the greatest power possable when impedances match, even with transistors that is if they didn't burn up first WHICH I GUARANTEE THEY WILL . Tubes on the other hand WILL NOT BURN UP GUARANTEED into a matched load

 

 

You first statement is not true for either transistors or tubes. Transistors if you tried to match the output impedance you would burn up the transistor because of too much current otherwise it would also be fact for transistors that they would deliver thier best power into a matched impedance. The transistor would try to deliver too much power, much more than it could handle in a matched impedance scenerio. Practicalities prevent us from matching the output impedance in transistor amps due to the output impedance being so low.

 

This will be my last explanation of this because I have made it very clear all along the way what I'm talking about.

 


Edited by germanium - 4/10/12 at 10:23pm
post #104 of 128
Quote:
Originally Posted by Chris J View Post



Impedance matching has no application in low frequency transmission (i.e. audio frequencies).

Impedance matching refers to output impedance = load impedance.

 

i'm aware of that, thats what I was inferring when using RF terms, Erno Borbely often used terminated lines in his audio amps too including using BNC connectors, but this was probably more about being a perfectionist than anything. the effects are still there, but due to the reflections being less of an issue with audio its hardly worth bothering about IMO

 

 

 

In a power amp you want output impedance to be much lower than load impedance for most efficient power transfer and accurate voltage transfer.

In actual practice it is impossible to build an amp without feedback, local feedback is inherent to all transistors and tubes, for example, without feedback a tube or transistor would have infinite bandwidth.

I agree, degeneration is Local Feedback.

I'm sure the "no feedback in this amp" thing is another usless marketing term.

Anyone have any examples of a "No Feedback" amp?

 

yeah i'm doubtful too, but I dont know much about tubes at all, so I didnt rule it out and just said 'very very rare' as in i've never seen one, it would seem to me to be a ticking time bomb if it did exist. I was under the impression that in general tubes were very wide band devices, so would have to have some at all times

 

Generally speaking, loudspeakers are designed to be driven by a voltage source.

The voltage is applied to the speaker, which causes current to flow, the change in current creates a varying magnetic field, which makes the voice coil move.

 

Generally speaking yes, but they operate in the current domain and there is definitely a section of audio using current buffers to drive horn speakers. not to my taste but its definitely a reality and I respect Nelson and other engineers at DIYA who are playing with it and swear by it enough to think there must be something to it, just another flavor I guess



 



OK, lemme try again:

 

match tube output impedance to transformer primary impedance

 

Transformer output impedance (secondary impedance) is less than load impedance.

 

Right?

 

 

hes talking about current transfer not voltage transfer as I understand it



aha, so you do mean global feedback and the cathode resistor is the degeneration. same thing. so no such thing as no feedback. without some form of feedback the device would not have any..well..feedback.. or reference to be able to refer its adjustments to


Edited by qusp - 4/11/12 at 8:28am
post #105 of 128

My comments in italics

 


Impedance matching has no application in low frequency transmission (i.e. audio frequencies).

Impedance matching refers to output impedance = load impedance.

 

i'm aware of that, thats what I was inferring when using RF terms, Erno Borbely often used terminated lines in his audio amps too including using BNC connectors, but this was probably more about being a perfectionist than anything. the effects are still there, but due to the reflections being less of an issue with audio its hardly worth bothering about IMO

 

sounds like "Ernie" uses this between the pre-amp and the power amp?  Since we are not worried about power efficiency between pre-amp and power amp I guess we can let Erno be a perfectionist!

I got curious and pulled out one of my reference textbooks:   "Ben Duncan" High Performance Audio Power Amplifiers.   Ben says that at audio frequncies and at reasonable cable lengths, cables look like the respective sums of resistance, inductance and capacitance.   As he points out, most loudspeakers are looking for a low output impedance voltage source (the power amp) to drive them.

 

In a power amp you want output impedance to be much lower than load impedance for most efficient power transfer and accurate voltage transfer.

In actual practice it is impossible to build an amp without feedback, local feedback is inherent to all transistors and tubes, for example, without feedback a tube or transistor would have infinite bandwidth.

I agree, degeneration is Local Feedback.

I'm sure the "no feedback in this amp" thing is another usless marketing term.

Anyone have any examples of a "No Feedback" amp?

 

yeah i'm doubtful too, but I dont know much about tubes at all, so I didnt rule it out and just said 'very very rare' as in i've never seen one, it would seem to me to be a ticking time bomb if it did exist. I was under the impression that in general tubes were very wide band devices, so would have to have some at all times

 

 

I'm really not an expert in tube power amps either, but I know more than a little about solid state amps.

I think I have to eat my words, Ben Duncan refers to single ended, transformer coupled amps running in Class A with low "overall" (global) feedback  I guess the next step may be NO global feedback.   He also comments on Class A amps without feedback..

I agree, in general tubes are very wide bandwidth. In tube power amps the bandwidth is limited by the output transformer.  Of course there are guys like Julius Futterman and his  vacuum tube OTL power amps.

 

Generally speaking, loudspeakers are designed to be driven by a voltage source.

The voltage is applied to the speaker, which causes current to flow, the change in current creates a varying magnetic field, which makes the voice coil move.

 

Generally speaking yes, but they operate in the current domain and there is definitely a section of audio using current buffers to drive horn speakers. not to my taste but its definitely a reality and I respect Nelson and other engineers at DIYA who are playing with it and swear by it enough to think there must be something to it, just another flavor I guess



I agree, Nelson knows his amps!

Back to my Ben Duncan textbook!

"Ben says" loudspeakers are normally designed to be driven by a low impedance voltage source.

What's up with the current sources?  Can you send me a link?   Is this a current buffer between an SET and a loudspeaker to make the SET look like a high output current source to the speaker?  Is it a solid state amp with a voltage gain of 1?



OK, lemme try again:

 

match tube output impedance to transformer primary impedance

 

Transformer output impedance (secondary impedance) is less than load impedance.

 

Right?

 

 

hes talking about current transfer not voltage transfer as I understand it

 

I think I may have (finally!) figured it out, we were discussing this from different points of view.

 

Since tubes output high voltage & low current, then we use a transformer to step the voltage down and step the current up to a drive the loudspeaker.

At the output tubes, the transformer shows the tubes a load of (for example) 2,000 ohms or so.  Or you could say the transformer turns an 8 ohm load into a 2,000 ohm load.

From the loudspeakers point of view (and from the point of view of a power engineer like me) the loudspeaker is driven by a fairly low impedance source. From the loudspeakers point of view the loudspeaker is driven by a voltage source with a fairly low output impedance.  Generally speaking, transfomers usually have a source impedance between 5-10% of the load impedance. So lets say maybe 0.8 ohms for an 8 ohm tap. 

The contradictions are ironed out............I think.   So it is optimizing voltage and current, and therefore power transfer.
 

Quote:
Originally Posted by qusp View Post



aha, so you do mean global feedback and the cathode resistor is the degeneration. same thing. so no such thing as no feedback. without some form of feedback the device would not have any..well..feedback.. or reference to be able to refer its adjustments to

 

I agree, it's the same thing.
As I define it, or as the standard definition goes.......Global Feedback is often called Loop Feedback, from output back to input of the complete power amp (or preamp or whatever).

 

The cathode resistor his friend Jack used is often called degeneration and often referred to as Local feedback.

I have a vacuum tube pre-amp which has no loop feedback and has no degeneration at Cathode. It is the Counterpoint SA-5.1.   But since it does not have infinite bandwidth, there is parasitic capaciance between grid and plate limiting it's bandwidth.  The parasitic capacitance is the feedback.  I have the schematic if you are interested.

So no such thing as "no feedback". It just may not be apparent.

 


Edited by Chris J - 4/11/12 at 5:50pm
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Damping factor. Why it is not always as important as some make it out to be.