|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
76.90.38.219
In Reply to: RE: More from Lynn Olsen posted by Naz on September 01, 2014 at 22:26:57
Lynn covers that in the paragraph preceding the first I quoted.
"Most commercial tube amps only have 1 or 2dB of headroom in the driver circuit, so the entire amp clips at once. This leads to longer recovery times and exaggerating the audibility of clipping. I prefer 3 to 6dB of headroom, so the driver can retain its (voltage) linearity even when the output stage is deep into clipping. "
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Follow Ups:
while there is a connection between the two, ensuring that the load line of the driver is kept as horizontal as possible including, and factoring in the load it must drive, right up to the point the OP stage overloads is most important.
It just so happens that designing in plenty of headroom helps the cause but all the headroom in the world won't fix an inherently non-linear driver. I know you know this, just clarifying my point.
Naz
"I prefer 3 to 6dB of headroom, so the driver can retain its (voltage) linearity"
I believe that's true even if it's swinging into an area where linearity degrades. Most important is the linearity at reduced swing ("the first watt"), and of course, even a tube driven into a somewhat less linear region on peaks is to be preferred over one that actually clips.
--------------------------
Buy Chinese. Bury freedom.
Those of us into tube amps, who use EFFICIENT speakers, are listening 99 percent of the time to 50 milliWatts to 500 milliWatts of amplifier power, seldom over 3/4s of a Watt on well-wired speakers like, say, my Altecs.
Why not design for THAT power level to be optimum, since we use it 99 percent of the time. If the first Watt isn't pristine, who cares about power above that... and designing overhead for "clipping". You are focussing and are emphasizing the things that are not important. I think even John Camille figured that one out.
Get larger, more efficient speakers for tube amps or go GOOD solid state, it exists, just costs some money. I am sure a Spectral, or Boulder, or the big MBL solid state amps will absolutely KILL in performance, high powered tube amps on inefficient speaker loads.
Jeff Medwin
Once upon a time I had a girlfriend who taught aerobics out of our house. One lady (client) used to bring her boyfriend with her (she didn't drive) and we listened to my stereo drank/smoked/BS'ed.
One day he told me about a friend of his and mentioned he was into audio. Then he told me what the guy owned....well I flat out begged to get a chance to listen. He said it was nice and all....but it didn't connect like mine did. Given most of my stuff came from a thrift store....and most of this guy's stuff ran 5 figures per item....well I found that hard to believe. He had good stuff that sounded like it had been picked to compliment each other.
The day finally came and I have to admit I was quite surprised. Oh it was clean and powerfull. But it was souless.
I keep hearing that there is good affordable solid state out there now. I just haven't had the pleasure. Maybe I fell in love with false artifacts of simple tube circuits and HE speakers. I knew the soundstage I loved hearing wasn't a true representation of real live music...played in front of me in my room....yet when it wasn't there on recordings that I knew had it....I just couldn't be satisfied.
If there is any single area I think many miss the boat on it is speakers. Sometimes it comes down to an epihany. Jeff I think you must of had one with Dennis. I had one (with another inmate) with speakers. This guy had some ribbon tweeters...crazy specs like 110db/watt from 1Khz to 30Khz....yet rated for something stupid like 2,000 watts....yes 2,000 watts....like rolling stones in a football stadium stuff.
We walked away thinking our speakers were toys. We learned what it was to play a HE driver at a very small fraction of its rated power...and to operate it over a range where it is a non-reactive linear load.
So yes, 1 good watt can be all you need....cept for real sub-bass:)
"designing overhead for 'clipping'. You are focussing and are emphasizing the things that are not important."
No, I'm not. Music has a huge dynamic range, and every system runs the risk of clipping occasionally. Once that occurs, the integrity of information in the "first watt" can only be regained following the recovery time. Ignore this principle during design, and the quality of your system will suffer for it.
--------------------------
Buy Chinese. Bury freedom.
I agree with what you are saying but my system gain prevents the driver tube or the output tube from ever being clipped.
2vrms max from my CD player (no active preamp, passive AVC), and a driver tube with a max mu of 17, that gives me no more than 95.88 volts peak to peak to the grid of the output tube.
2 * 1.41 = 2.82 * 17 = 47.94 * 2 = 95.88
With my output tube biased at -53 (SET 6b4g) or -80 (SET 300b) I don't see how the output tube can ever be driven to clipping. (I use the same driver for either output stage)
And yes, I only get a few watts out of the 300b. I just don't have the voltage gain to get any more.
I love the way the system sounds and it gets as loud as I need it to be.
I should point out that the tube amp(s) are only playing down to 200Hz.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Well, there are exceptions to everything. The design philosophy I'm describing is not invalidated by that. I also don't believe it's at odds with "first watt" transparency, as implied by drlomu. Just trying to keep it all straight... ;)
--------------------------
Buy Chinese. Bury freedom.
TK, I'm not trying to invalidate any of the good points you have made.
I was just pointing out that some valid concerns can be circumvented.
Like the good and proper concern with RC coupling having a long recovery time once grid current is present. (charged cap needing to discharge, etc)
Again, with less than enough max drive voltage to overdrive the output tube, that troublesome condition is never realized.
I have garnered a lot of negative attention while describing what is and isn't Class A operation.
In some of the text books Class A is described as not only the idle current but the load impedance and the driving voltage.
That is, limiting the drive voltage as to not drive the tube into the non-linear region of operation.
I contend that an amplifier, while under normal intended use, that allows the output tube to enter the non-linear portion of the operating curve is not a Class A amplifier.
Excuse me while I search for my flame suit. :-)
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
I don't think I've ever seen a definition of Class A that was complete in the context of modern hi-fi. Some say simply that the tube never draws grid current, others that it conducts over 360 degrees. These definitions lead to the often-heard myth that Class AB amplifiers operate in Class A for small signals. It's unfortunate that the majority of engineering texts fail to expand on these simple definitions, but most engineers probably assume that Class A operation will position the tube at the mid-point of conduction. Considering that this not only allows maximum +/- current swing, but also the most linear operation, it seems superfluous to note this in even a formal definition.
Getting back to your statement, when you say the "non-linear region of operation," are you referring to driving the tube into grid current? Or do you mean the general areas just prior to cutoff and saturation, areas through which the tube continues to conduct well, but nevertheless responds in a degraded manner with respect to linearity?
On the subject of designing for headroom (or more correctly, minimal recovery time) is your CD player actually incapable of producing more than the 5.6V PP that represents 2V RMS? That would seem to place the allowable average output level at a rather low number, voltage-wise. Not saying you're wrong, just asking. I don't remember what peak levels I saw the last time I put a scope on my CD player...
--------------------------
Buy Chinese. Bury freedom.
The general areas just prior to cutoff and saturation, areas through which the tube continues to conduct well, but nevertheless responds in a degraded manner with respect to linearity.
The manual says 2vrms and that's what I remember measuring when playing a 60Hz CD I made in Sound Forge normalized to 0db.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
"The manual says 2vrms and that's what I remember measuring when playing a 60Hz CD I made in Sound Forge normalized to 0db."
Have you looked at the output with a scope to see whether signals created from a standard music CD can exceed that level? I'll do it on mine if you don't have time, they're probably similar in terms of levels...
--------------------------
Buy Chinese. Bury freedom.
Here a screen shot of the wave form in Sound Forge and a close up showing the peak sample sitting right on the digital "0".
TK, here's the part that maybe, somehow I'm not understanding.
A digital "0" is a digital "0".
There can be no level above that without digital clipping.
If my 60Hz wave form is peaking at a digital "0" and I measure 2vrms with a multimeter, then I think I'm done.
No other non-clipped signal can be higher than that. Sine wave or music waveform.
That's the thing about digital. It has an upper limit that can't be exceeded.
I can look at it with a scope but I don't see the need. If a wave form that is peaking at digital "0" measures 2vrms at the output of the CD player with a volt meter then how can there be any level above that?
Maybe I should measure it again. I seem to remember that it measured just a little over 2vrms and at the time I thought "close enough".
I posted/hosted the wav file at the link below.
Edit, a digital 0 is represented as all 1's for the word for that sample. The D to A converter will output full voltage for a sample with the word written with all 1's.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
Edits: 09/03/14
Tre', I'm not familiar with the system or software you're using, so I have to ask - are you certain the signal in the upper screenshot represents "all ones"? If it does, then I certainly agree, the amplifier won't saturate. I don't think this inability to clip is representative of most systems, but it's a smart thing to do. The only problem might come about when a CD with a very wide dynamic range is in use, especially if much of the music is at a rather low output level. I often find myself turning up the volume so I can hear the quieter passages in more detail, but that makes the system prone to clipping on peaks. If the amplifier is designed for quick recovery, those brief periods of saturation are much less offensive. Anyway, it sounds as though you've "calibrated" your system so that none of this is a problem. If your speakers have the sensitivity to raise you off the ambient noise floor of your listening environment even at low source levels, it's a good approach. Wish I was that lucky! :)
--------------------------
Buy Chinese. Bury freedom.
"...are you certain the signal in the upper screenshot represents "all ones"?"
No, just the one sample that touches digital O. The peak, the apex of the wave form.
And as such, that's as much voltage as the D to A converter can output.
I live in a rural area and my house is back off the street. It's pretty quiet here most of the time. People from the city are always amazed.
Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"
and to expand on one of them a common misconception is that tubes are inherently non-linear at low current because that's the way curves appear on data sheets. If we magnify that region they look pretty much the same as they do at higher current.
Higher current does improve the overall performance parameters of a tube, lower rp, higher gm etc etc but properly designed any tube is capable of great performance and sound at low current. Unfortunately OP tubes need to deliver power into the load so we are limited to how far we can go.
Naz
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: