Home |
Search |
Today's Posts |
#1
|
|||
|
|||
A little feedback worse than none at all?
Hi gang. I'm working on the design of an amplifier to be used in a
tube mixing console. (Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!). Anyway, I cooked up a usable circuit in short order: two triodes in cascade, with negative feedback from the plate of triode 2 to the cathode of triode 1. I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. My distortion analyzer is out of order at the moment, so I can't look at the distortion products of my breadboarded circuit. I guess I'm just wondering if Baxandall's assertion (which apparently arose from experiments with a FET amplifier) also applies to tubes. Your comments are very much appreciated. Why not just use, say, a 12AX7 (with an open-loop gain of 1600 in this particular circuit)? Well, I tried it, but it clips with a much lower input voltage (0.14VRMS versus 0.67) and has a much lower maximum output swing than the 12AU7 (27 VMRS versus 45). |
#2
|
|||
|
|||
"NewYorkDave" wrote in message om Hi gang. I'm working on the design of an amplifier to be used in a tube mixing console. (Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!). Anyway, I cooked up a usable circuit in short order: two triodes in cascade, with negative feedback from the plate of triode 2 to the cathode of triode 1. I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. My distortion analyzer is out of order at the moment, so I can't look at the distortion products of my breadboarded circuit. I guess I'm just wondering if Baxandall's assertion (which apparently arose from experiments with a FET amplifier) also applies to tubes. Your comments are very much appreciated. Why not just use, say, a 12AX7 (with an open-loop gain of 1600 in this particular circuit)? Well, I tried it, but it clips with a much lower input voltage (0.14VRMS versus 0.67) and has a much lower maximum output swing than the 12AU7 (27 VMRS versus 45). Perhaps you might try cascading multiple gain stages with global NFB? Maybe a 12AX7 for the high initial gain and a 'U7 for the output swing? Just a thought. |
#3
|
|||
|
|||
"NewYorkDave" wrote in message
om... Hi gang. I'm working on the design of an amplifier to be used in a tube mixing console. (Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!). Actually, let's not ignore it - it's often a mistake to give someone engineering advice if you don't understand why they're trying to do what they're trying to do. If someone wants a tube mixing console, it is presumably because they want a certain sound or behavior which they associate with tube mixing consoles. But obviously, not all tube circuits sound the same - if they did, you wouldn't be worried about what topology to choose. So, unless the problem is just that their (your?) studio is too cold, you really need to know what kind of mixing console they're dreaming of; find its schematic, or at least its topology; and build that. But your post suggests that you are attempting to build with the simple goal of reducing distortion. With due respect to the fact that this is cross-posted to rec.audio.tubes, I would submit that starting with a tube design is not the most straightforward way to achieve low distortion, though it may be a good way to achieve euphonic distortion. There must be some additional goal, and you need to find it (if you haven't already) or tell us what it is (if you have). My distortion analyzer is out of order at the moment, so I can't look at the distortion products of my breadboarded circuit. Do you have a sound card on a computer? Send a sinewave to the circuit, sample the results, do an FFT. If not, can you build a reasonably sharp notch filter - even a passive RLC one, with three components? If so, build one; send a sinewave at the circuit, adjust the frequency for maximum null, look at the results. In either case, if you don't have a low-distortion sinewave handy, you won't get a perfect picture - but it will certainly be good enough to tell you whether your high-order distortion products are decreasing or increasing. Just compare with and without the circuit. But again, low distortion is not necessarily the goal here. We don't know what the goal is, yet. |
#4
|
|||
|
|||
NewYorkDave wrote: Hi gang. I'm working on the design of an amplifier to be used in a tube mixing console. (Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!). Anyway, I cooked up a usable circuit in short order: two triodes in cascade, with negative feedback from the plate of triode 2 to the cathode of triode 1. I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. My distortion analyzer is out of order at the moment, so I can't look at the distortion products of my breadboarded circuit. I guess I'm just wondering if Baxandall's assertion (which apparently arose from experiments with a FET amplifier) also applies to tubes. Your comments are very much appreciated. I have seen the very article in Wireless World to which you refer. What Mr Baxandall said was all too true. But the thd without FB has to be quite high before you'd worry about the increase in high order artifacts after applying a modest, ie, less than 14 dB of NFB. Where you have only 1% of thd you won't have a problem with thd increase after FB application. Why not just use, say, a 12AX7 (with an open-loop gain of 1600 in this particular circuit)? Well, I tried it, but it clips with a much lower input voltage (0.14VRMS versus 0.67) and has a much lower maximum output swing than the 12AU7 (27 VMRS versus 45). The earlier clipping isn't caused by the FB, its due to the load imposed on the second 1/2 of the 12AX7. The 12AX7 operates ideally with only 0.7mA of anode current, and if it clips at 27 vrms, then the load including the DC supply resistor, external load on the amp, plus the feedback resistor will total 54k. This is a low load for 12AX7, and they like loads up around 150k. But you could have a 12AU7 cathode follower, and take the FB from the follower cathode to the cathode of the V1, and then you get an extremely low output resistance, and a large output voltage capability, and very low thd. On the other hand, why use any loop FB at all? triodes like 6CG7/6SN7 are very linear at 1 v output, which may be all you use in a mixer. It does depend on the gain you want, and the circuit you propose to use, and its use. 12AU7 isn't quite as linear, but has a warmer sound, imho, and your use of moderate, and trimable FB to establish a fixed known gain shouldn't affect the sound much. Patrick Turner. |
#5
|
|||
|
|||
Walter Harley wrote: "NewYorkDave" wrote in message om... Hi gang. I'm working on the design of an amplifier to be used in a tube mixing console. (Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!). Actually, let's not ignore it - it's often a mistake to give someone engineering advice if you don't understand why they're trying to do what they're trying to do. If someone wants a tube mixing console, it is presumably because they want a certain sound or behavior which they associate with tube mixing consoles. But obviously, not all tube circuits sound the same - if they did, you wouldn't be worried about what topology to choose. So, unless the problem is just that their (your?) studio is too cold, you really need to know what kind of mixing console they're dreaming of; find its schematic, or at least its topology; and build that. But your post suggests that you are attempting to build with the simple goal of reducing distortion. With due respect to the fact that this is cross-posted to rec.audio.tubes, I would submit that starting with a tube design is not the most straightforward way to achieve low distortion, though it may be a good way to achieve euphonic distortion. There must be some additional goal, and you need to find it (if you haven't already) or tell us what it is (if you have). My distortion analyzer is out of order at the moment, so I can't look at the distortion products of my breadboarded circuit. Do you have a sound card on a computer? Send a sinewave to the circuit, sample the results, do an FFT. If not, can you build a reasonably sharp notch filter - even a passive RLC one, with three components? If so, build one; send a sinewave at the circuit, adjust the frequency for maximum null, look at the results. In either case, if you don't have a low-distortion sinewave handy, you won't get a perfect picture - but it will certainly be good enough to tell you whether your high-order distortion products are decreasing or increasing. Just compare with and without the circuit. But again, low distortion is not necessarily the goal here. We don't know what the goal is, yet. With all due respect to your very informative comments, one should realise that a 6SN7, or a humble 12AX7 might make 5% of mainly 2H distortion at about 65 vrms output from a typical circuit, optimally loaded with resistances, or a lot less if loaded by constant current sources. The rate of decrease in the thd as the output voltage is reduced isn't linear so that by 1 volt output, it may be only 0.04%, and nearly all 2H. If you have 2.83 volts applied to an 8 ohm speaker, then that's a watt, and maybe you get 90 dB SPL. The distortion voltage would be 1 mV. This would be inaudible at 4 metres away, even if it was played some how without the main undistorted signal present. It represents an SNR of 68 dB. The sound of tubes isn't always one of euphony, and pleasantness caused by the 2H, since it is inaudible, so some other hard to define effect is going on to render the music more pleasant to listen to. One can plug in various brands of the same tube type, and get a different sonic signature each time. Its more aparent with less FB. For example, I prefer Telefunken NOS 6CG7 ahead of NOS Mullard, ahead of EH6CG7, in that order. They measure somewhat similarly. Building a tube audio system is the hard way to go about audio, but many ppl have heard this inexplicable difference to the sound, which they like, so its worth doing, for them. The original poster was aking about if he should use NFB, and yes, its a reliable way to trim different channels to get the same gains. It may/may not have effects on the thd, and maybe the sound, but I have found a little FB doesn't ruin the sound, ie, it won't change tp that of a typical SS set up. Some folks spend the bucks on Burr Brown opamps, some spend them on tubes, and either devices offer low enough thd, depending on the use. There are many ppl fed up with the hard sound of totally digital and SS systems, and they add some tubes into the chain to warm the sound. It sounds like snake oil, I know, but I have heard the effects with my own ears. Totally analog tubed studio systems can offer truly stunning sound, especially if they don't compress anything, of fiddle with the sound in a chain of 50 cascaded opamps. I know of one small studio here where they do use a tube compressor, and the do a lot of fiddling, since the musicians scream loud for it, because they loathe the accurate sound of themselves on a recording, and beg the studio to titivate their sound this way and that to make it better than reality, more appealing, so they'll sell more CDs. Music is a aural fantasy. Anything goes in the world of pop, which is the greater part of recording these days. But to hear the essence of a Stradivari, I'd prefer a tubed system. But you are right about goals. We dunno what the guy really wants to do with his mixer. Patrick Turner. |
#6
|
|||
|
|||
"Patrick Turner" wrote...
With all due respect to your very informative comments, one should realise that a 6SN7, or a humble 12AX7 might make 5% of mainly 2H distortion at about 65 vrms output from a typical circuit, optimally loaded with resistances, or a lot less if loaded by constant current sources. The rate of decrease in the thd as the output voltage is reduced isn't linear so that by 1 volt output, it may be only 0.04%, and nearly all 2H. Quite so, and I agree with your further comments on the inaudibility of said distortion. I will make the observation that in a typical recording console these days, one expects to pass reasonably clean signals of at least +18dBu Vrms, which would be about 17 volts peak to peak. So your estimate of 1V is a bit of a straw man, I fear - I expect we are closer to the 5% end than the 0.04%, and that is quite audible. Nonetheless, the only point I'm trying to make - and I think from your comments that you would agree - is that not all tube circuits sound alike. Since as you say "building a tube audio system is the hard way to go about audio," there must be a reason why the OP wants tubes. If he simply sets about building something that has tubes in it - or, if he applies novel design techniques - there is no reason to expect he will get anything like the sound he or his client has in mind. There are many ppl fed up with the hard sound of totally digital and SS systems, and they add some tubes into the chain to warm the sound. It sounds like snake oil, I know, but I have heard the effects with my own ears. I would not dream of getting into a tubes vs. semiconductors argument; it has been repeatedly shown that that is unproductive for all involved! Personally, on my hifi I prefer to hear something as close as possible to what the mixing engineer was hearing, so I opt for a system that reproduces with maximal accuracy. But that is a personal choice; and it is achievable through a variety of technologies. It should be beyond argument that, if one makes the opposite choice (wishing one's stereo to further modify the sound), any technology can supply the necessary inaccuracies, and different technologies will supply different inaccuracies. Either way, we are talking here about a mixing desk rather than a hi-fi, and as you say it is commonplace (and to my ears often desirable) to introduce intentional inaccuracies during audio production, whether through distortion in the microphones, preamps, mixing desk, analog tape, or other processing gear. I'd be glad to build a tube console if I could find a client to pay for it. I'd start by finding the schematic of a classic tube console. -walter |
#7
|
|||
|
|||
"Walter Harley" "Patrick Turner" The rate of decrease in the thd as the output voltage is reduced isn't linear so that by 1 volt output, it may be only 0.04%, and nearly all 2H. Quite so, and I agree with your further comments on the inaudibility of said distortion. * A figure 0.04% THD ( mainly 2H) would be a very good *overall spec* for an audio reproduction system - ie from mic input to home loudspeaker terminals. But for an single stage out of of twenty or so such stages it is totally unacceptable - if all were the same the THD would then build up to a total of 1% or more and be very audible as muddy sound - or more precisely innummerable intermodulation products. All the pre-amp and line level stages that make up the recording chain need to be as free of non-linearity as possible so the end result is no worse than the touted 0.04% THD "inaudible" goal. ............ Phil |
#8
|
|||
|
|||
I read in sci.electronics.design that NewYorkDave
wrote (in b514d126.0311141449.7ac28720@po sting.google.com) about 'A little feedback worse than none at all?', on Fri, 14 Nov 2003: I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. With only 4 dB of feedback, you won't get predictable gain. You need more feedback. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. ANY amount of feedback causes this; the question is whether it matters. With tubes and FETs, it usually doesn't; with the more aggressive non- linearities of bipolar transistors, it often does. You get a low THD, say 0.3%, but it is composed of every harmonic from 2nd to infinity. It's not so much the harmonic distortion that matters but the consequent complex intermodulation distortion. You overcome this with bipolars by paying a lot of attention to minimising the open-loop non-linearity, but that's a good principle to adopt with any sort of active devices. [snip] Why not just use, say, a 12AX7 (with an open-loop gain of 1600 in this particular circuit)? Well, I tried it, but it clips with a much lower input voltage (0.14VRMS versus 0.67) and has a much lower maximum output swing than the 12AU7 (27 VMRS versus 45). But the 12AX7 with more feedback is potentially a better solution, if there isn't some special reason why you need more than 27 V RMS output. More feedback will allow you to apply 0.67 V at the input, or even more. If you choose plate resistors more suitable for 12AX7 you can probably get more than 27 V RMS output anyway. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#9
|
|||
|
|||
John Woodgate wrote:
I read in sci.electronics.design that NewYorkDave wrote (in b514d126.0311141449.7ac28720@po sting.google.com) about 'A little feedback worse than none at all?', on Fri, 14 Nov 2003: I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. With only 4 dB of feedback, you won't get predictable gain. You need more feedback. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. ANY amount of feedback causes this; Ahmmm, a bit misleading john. If the feedback is large enough, essentially, all distortion will be reduced in practise. The deal is that with *low* feedback, a pure square law device will have its input mixed with its output causing generation of 3rd harmonics, and higher, which were not present in the original amp. However, if the FB is large enough, the resulting distortion can easily go to 0.01%. Sure, if the original device was *pure* square, than 0.01% 3rd, is still larger than zero. The reality however, is that all practical devices have odd harmonics as well, such that large feedback will result is a net total reduction of the original distortion at all harmonics of relevance. Kevin Aylward http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#11
|
|||
|
|||
John Woodgate wrote: I read in sci.electronics.design that NewYorkDave wrote (in b514d126.0311141449.7ac28720@po sting.google.com) about 'A little feedback worse than none at all?', on Fri, 14 Nov 2003: I'm using both sections of a 12AU7. The open-loop gain of the circuit is 123, and with feedback it's 76. So, there's 4.18dB of feedback, a very modest amount. I didn't like the idea of using NFB at all at the beginning, but this application really demands predictable gain that isn't unduly affected by tube aging or manufacturing spread. With only 4 dB of feedback, you won't get predictable gain. You need more feedback. Tube circuits often measure within 1/2 a dB with no FB. So to equalise gains of 3 or more channels, one can trim the feedback with a pot to equalise the gains, so maybe you have 3.5 dB in one amp, and 4.5 db in another. This will not make much difference to the sound of each channel. But if say 20 db of FB is used, there isn't any need to adjust the FB networks, as the same fixed 1% resistor networks will give gains well within 1/2 a dB. Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. ANY amount of feedback causes this; the question is whether it matters. Baxandal said that where thd was high, and FB was moderate, and open loop bandwidth restricted, then this made the thd barely a measured improvement. So lots of open BW, lots of NFB, and low open thd were needed for a best result. With tubes and FETs, it usually doesn't; with the more aggressive non- linearities of bipolar transistors, it often does. I have found j-fets produce maybe 20 times the thd of a tube at low signal voltages. Fets should be lumped in with bjts, not tubes, imho. You get a low THD, say 0.3%, but it is composed of every harmonic from 2nd to infinity. It's not so much the harmonic distortion that matters but the consequent complex intermodulation distortion. You overcome this with bipolars by paying a lot of attention to minimising the open-loop non-linearity, but that's a good principle to adopt with any sort of active devices. True, the IMD is the worst problem, but the lower the THD, the less IMD you have. [snip] Why not just use, say, a 12AX7 (with an open-loop gain of 1600 in this particular circuit)? Well, I tried it, but it clips with a much lower input voltage (0.14VRMS versus 0.67) and has a much lower maximum output swing than the 12AU7 (27 VMRS versus 45). But the 12AX7 with more feedback is potentially a better solution, if there isn't some special reason why you need more than 27 V RMS output. More feedback will allow you to apply 0.67 V at the input, or even more. If you choose plate resistors more suitable for 12AX7 you can probably get more than 27 V RMS output anyway. The loading seen by the 12AX7 restricts its output, since a typical idle current is a low 0.7 mA. So a load which a bjt might drive easily, say 1k, can only have 5vrms of output, from a 12AX7, and with far greater thd, because the RL seen by a triode should be many times its Ra. Patrick Turner. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#12
|
|||
|
|||
I read in sci.electronics.design that Patrick Turner
wrote (in u) about 'A little feedback worse than none at all?', on Sun, 16 Nov 2003: The loading seen by the 12AX7 restricts its output, since a typical idle current is a low 0.7 mA. So a load which a bjt might drive easily, say 1k, can only have 5vrms of output, from a 12AX7, and with far greater thd, because the RL seen by a triode should be many times its Ra. I wouldn't dream of driving a 1 kohm load with a 12AX7. I can't see that what you say is relevant. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#13
|
|||
|
|||
John Woodgate wrote: I read in sci.electronics.design that Patrick Turner wrote (in u) about 'A little feedback worse than none at all?', on Sun, 16 Nov 2003: The loading seen by the 12AX7 restricts its output, since a typical idle current is a low 0.7 mA. So a load which a bjt might drive easily, say 1k, can only have 5vrms of output, from a 12AX7, and with far greater thd, because the RL seen by a triode should be many times its Ra. I wouldn't dream of driving a 1 kohm load with a 12AX7. I can't see that what you say is relevant. I know you wouldn't dream of it; I was just pointing out about a low load effect. Atcually, my figures for 1 k are wrong, and the vo should only be 0.5v. It is concievable that the load to which a mixer is hooked up to is 10k, in which case the vo is 5v. If the tube used on the output is a 1/2 a 12AU7, with 5 mA of idle curent, then the maximum output into 1 k is around 3.5 v, and into 10k its 35 v. Patrick Turner. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#14
|
|||
|
|||
I apologize for not providing more detail in my original post. I'll
try to give a clearer explanation of what I'm trying to do. I have only a few minutes left in my lunch hour, but I'll try to put it to good use I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) and perhaps some of that intangible "tube warmth." However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. Contrast this to a typical console layout from 1961: http://vintageproaudio.com/photos/langevinwiring1.jpg http://vintageproaudio.com/photos/langevinwiring2.jpg The "open system" architecture of the classic tube console is not cost-effective to build today, with its constant-impedance attenuators and panpots, transformer-coupled I/O on all amplifier stages, etc., so I'm opting for a simpler high-impedance passive mixing arrangement followed by tube makeup gain, tube buffering and a transformer-coupled, balanced, Low-Z output. My circuit has 6.5dB of loss in the panpot, and about 21dB of loss in the mixing network. The output stage of the mixer is a white cathode follower driving a 2:1 stepdown transformer, so there's another 7dB or so of loss. Between the mixing network and the cathode follower, there needs to be an amplifier stage to make up all these losses, plus provide an additional 12dB or so of reserve amplification. The topology I was going to use was a two-stage feedback amplifier followed by a master volume control (potentiometer), then the cathode follower output stage. I considered using the amplifier stage direct-coupled to the cathode follower, with feedback from the CF to the cathode of the first stage, but then I would have to put my master volume control BEFORE the amplifier input, which would wreck the signal-to-noise ratio. |
#15
|
|||
|
|||
I read in sci.electronics.design that NewYorkDave
wrote (in b514d126.0311170958.2d277ae3@po sting.google.com) about 'A little feedback worse than none at all?', on Mon, 17 Nov 2003: I'm opting for a simpler high-impedance passive mixing arrangement followed by tube makeup gain, You are likely to find that not very satisfactory. An EF86 makes a good virtual-earth mixer with minimal unavoidable noise, but a 12AX7 cascode should be better, although I never tried it. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#16
|
|||
|
|||
(NewYorkDave) wrote in message . com...
I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) and perhaps some of that intangible "tube warmth." However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. Contrast this to a typical console layout from 1961: http://vintageproaudio.com/photos/langevinwiring1.jpg http://vintageproaudio.com/photos/langevinwiring2.jpg The "open system" architecture of the classic tube console is not cost-effective to build today, with its constant-impedance attenuators and panpots, transformer-coupled I/O on all amplifier stages, etc., so I'm opting for a simpler high-impedance passive mixing arrangement followed by tube makeup gain, tube buffering and a transformer-coupled, balanced, Low-Z output. My circuit has 6.5dB of loss in the panpot, and about 21dB of loss in the mixing network. The output stage of the mixer is a white cathode follower driving a 2:1 stepdown transformer, so there's another 7dB or so of loss. Between the mixing network and the cathode follower, there needs to be an amplifier stage to make up all these losses, plus provide an additional 12dB or so of reserve amplification. The topology I was going to use was a two-stage feedback amplifier followed by a master volume control (potentiometer), then the cathode follower output stage. I considered using the amplifier stage direct-coupled to the cathode follower, with feedback from the CF to the cathode of the first stage, but then I would have to put my master volume control BEFORE the amplifier input, which would wreck the signal-to-noise ratio. Hi Im far too tired to comment properly, but I dont like the sound of this one, I'd look at the options more. Regards, NT |
#17
|
|||
|
|||
NewYorkDave wrote: I apologize for not providing more detail in my original post. I'll try to give a clearer explanation of what I'm trying to do. I have only a few minutes left in my lunch hour, but I'll try to put it to good use I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) and perhaps some of that intangible "tube warmth." However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. Contrast this to a typical console layout from 1961: http://vintageproaudio.com/photos/langevinwiring1.jpg http://vintageproaudio.com/photos/langevinwiring2.jpg The "open system" architecture of the classic tube console is not cost-effective to build today, with its constant-impedance attenuators and panpots, transformer-coupled I/O on all amplifier stages, etc., so I'm opting for a simpler high-impedance passive mixing arrangement followed by tube makeup gain, tube buffering and a transformer-coupled, balanced, Low-Z output. My circuit has 6.5dB of loss in the panpot, and about 21dB of loss in the mixing network. The output stage of the mixer is a white cathode follower driving a 2:1 stepdown transformer, so there's another 7dB or so of loss. Between the mixing network and the cathode follower, there needs to be an amplifier stage to make up all these losses, plus provide an additional 12dB or so of reserve amplification. The topology I was going to use was a two-stage feedback amplifier followed by a master volume control (potentiometer), then the cathode follower output stage. I considered using the amplifier stage direct-coupled to the cathode follower, with feedback from the CF to the cathode of the first stage, but then I would have to put my master volume control BEFORE the amplifier input, which would wreck the signal-to-noise ratio. Why don't you draw up a schematic of the proposed tube amp mixer, and let us take a look, and make comments. The use of an amp module using two cascaded tubes with FB to replace opamps is a valid enough starting point for the amplifiers. Its one heck of a lot of work for any of us to sit down and design it all for you. Patrick Turner. |
#18
|
|||
|
|||
"Walter Harley" wrote in message ... : "Patrick Turner" wrote... : With all due respect to your very informative comments, one should realise : that : a 6SN7, or a humble 12AX7 might make 5% of mainly 2H distortion at about : 65 vrms output from a typical circuit, optimally loaded with resistances, : or a lot less if loaded by constant current sources. : The rate of decrease in the thd as the output voltage is reduced isn't : linear : so that by 1 volt output, it may be only 0.04%, and nearly all 2H. : : Quite so, and I agree with your further comments on the inaudibility of said : distortion. I will make the observation that in a typical recording console : these days, one expects to pass reasonably clean signals of at least +18dBu : Vrms, which would be about 17 volts peak to peak. So your estimate of 1V is : a bit of a straw man, I fear - I expect we are closer to the 5% end than the : 0.04%, and that is quite audible. ........................... Using 2 EF86's, preferably Valvo or Philips, in parallel, triode connected, in a mu-stage with a 6AC7, with appropriately raised filament supply, running from 650 V, expect 75 V rms output at some 3% THD, 7.5 V rms @ 0.2 % or so. Noise with a 10 mV input signal at -65 dB, dynamic range 109 dB, all not too bad, I'd say. Gain of nearly 30 *. Using a stereo 100 K potmeter at the input of this and the next stage would make a really fine input channel. Rudy................... : I would not dream of getting into a tubes vs. semiconductors argument; it : has been repeatedly shown that that is unproductive for all involved! : Personally, on my hifi I prefer to hear something as close as possible to : what the mixing engineer was hearing, so I opt for a system that reproduces : with maximal accuracy. But that is a personal choice; and it is achievable : through a variety of technologies. It should be beyond argument that, if : one makes the opposite choice (wishing one's stereo to further modify the : sound), any technology can supply the necessary inaccuracies, and different : technologies will supply different inaccuracies. : : Either way, we are talking here about a mixing desk rather than a hi-fi, and : as you say it is commonplace (and to my ears often desirable) to introduce : intentional inaccuracies during audio production, whether through distortion : in the microphones, preamps, mixing desk, analog tape, or other processing : gear. : : I'd be glad to build a tube console if I could find a client to pay for it. : I'd start by finding the schematic of a classic tube console. : : -walter : : |
#19
|
|||
|
|||
"Ruud Broens" wrote in message ... : : "Walter Harley" wrote in message : ... : : "Patrick Turner" wrote... : : With all due respect to your very informative comments, one should : realise : : that : : a 6SN7, or a humble 12AX7 might make 5% of mainly 2H distortion at about : : 65 vrms output from a typical circuit, optimally loaded with : resistances, : : or a lot less if loaded by constant current sources. : : The rate of decrease in the thd as the output voltage is reduced isn't : : linear : : so that by 1 volt output, it may be only 0.04%, and nearly all 2H. : : : : Quite so, and I agree with your further comments on the inaudibility of : said : : distortion. I will make the observation that in a typical recording : console : : these days, one expects to pass reasonably clean signals of at least : +18dBu : : Vrms, which would be about 17 volts peak to peak. So your estimate of 1V : is : : a bit of a straw man, I fear - I expect we are closer to the 5% end than : the : : 0.04%, and that is quite audible. : : .......................... : Using 2 EF86's, preferably Valvo or Philips, in parallel, triode connected, : in a mu-stage with a 6AC7, with appropriately raised filament supply, : running from 650 V, : expect 75 V rms output at some 3% THD, 7.5 V rms @ 0.2 % or so. : Noise with a 10 mV input signal at -65 dB, dynamic range 109 dB, : all not too bad, I'd say. Gain of nearly 30 *. Using a stereo 100 K potmeter : at the input of this and the next stage would make a really fine input : channel. : Rudy................... Err, make that gain of nearly 60 *. The mu-stage will be quite capable of driving a low-impedance passive eq. circuit. I have the tubes, but not the 680 V supply, but then i'd go all the way for the ultimate input strip, with a moderate step-up Sowter mu-metal canned & core input transformer i have, over 70 dB s/n should be possible @ 10 mV. Second stage tube : try a 9002. Rudy : : : I would not dream of getting into a tubes vs. semiconductors argument; it : : has been repeatedly shown that that is unproductive for all involved! : : Personally, on my hifi I prefer to hear something as close as possible to : : what the mixing engineer was hearing, so I opt for a system that : reproduces : : with maximal accuracy. But that is a personal choice; and it is : achievable : : through a variety of technologies. It should be beyond argument that, if : : one makes the opposite choice (wishing one's stereo to further modify the : : sound), any technology can supply the necessary inaccuracies, and : different : : technologies will supply different inaccuracies. : : : : Either way, we are talking here about a mixing desk rather than a hi-fi, : and : : as you say it is commonplace (and to my ears often desirable) to introduce : : intentional inaccuracies during audio production, whether through : distortion : : in the microphones, preamps, mixing desk, analog tape, or other processing : : gear. : : : : I'd be glad to build a tube console if I could find a client to pay for : it. : : I'd start by finding the schematic of a classic tube console. : : : : -walter : : : : : : |
#20
|
|||
|
|||
NewYorkDave wrote:
I apologize for not providing more detail in my original post. I'll try to give a clearer explanation of what I'm trying to do. I have only a few minutes left in my lunch hour, but I'll try to put it to good use I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) This is an old wife's tale, i.e. a complete and utter nonsense argument. Less parts do *not* mean a clearer sound. Indeed, more stages mean you can have less gain per stage allowing for more feedback at each stage which reduces the distortion tremendously. and perhaps some of that intangible "tube warmth." This where a dirty sound sounds better. However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. So bloody what. A decent modern console uses op-amps with 0.005% thd and imd. It makes no practical difference whether there are quite a few in series, especially as explianed above. A final spec of a decent console is easily 0.01%, 1db flat from 20Hz to 40khz. You are suffering from a severe "golden ears" delusion. Kevin Aylward http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#21
|
|||
|
|||
"Kevin Aylward" wrote in message ... : NewYorkDave wrote: : I apologize for not providing more detail in my original post. I'll : try to give a clearer explanation of what I'm trying to do. I have : only a few minutes left in my lunch hour, but I'll try to put it to : good use : : I record music on an 8-track multitrack recorder, and I want to build : a tube mixing console to take the place of the IC-based model I've : been using. My goal is to achieve a clearer sound (due to a simpler : circuit with fewer active stages) : : This is an old wife's tale, i.e. a complete and utter nonsense argument. : Less parts do *not* mean a clearer sound. Indeed, more stages mean you : can have less gain per stage allowing for more feedback at each stage : which reduces the distortion tremendously. Hm, Kevin, that's rather oversimplifying things there, isn't it ? Sure, with a couple of BB OPA 2604's you can build a mixing desk at least an order cheaper than with tubes and anything more than 4 channels fully tubed must be called an *investment* rather than an experiment. That said, perceived sonic qualities of a circuit and THD measurements only have a marginal correlation, therefore THD is *not* indicative for the sound quality, still a largely intangible, only a starting point. But then, many a thousand pro-tube processing and mixing gear must be bought by fools... ? Rudy : and perhaps some of that intangible : "tube warmth." : : This where a dirty sound sounds better. : : However, I'm not looking to build a "distortion : device"... The tube consoles that were used to track the records we : all love from the '50s and '60s were designed to be as linear as : possible within the limits of the technology of the day, and any : euphonic qualities added were incidental. It's not clear to me whether : that sound really came from the tubesand transformers or from the more : straightforward topologies, anyway. In a modern mixing console, a : signal may pass through tens of IC amplifiers on its way from input to : output. : : So bloody what. A decent modern console uses op-amps with 0.005% thd : and imd. It makes no practical difference whether there are quite a few : in series, especially as explianed above. A final spec of a decent : console is easily 0.01%, 1db flat from 20Hz to 40khz. : : You are suffering from a severe "golden ears" delusion. : : Kevin Aylward : : http://www.anasoft.co.uk : SuperSpice, a very affordable Mixed-Mode : Windows Simulator with Schematic Capture, : Waveform Display, FFT's and Filter Design. : : http://www.anasoft.co.uk/replicators/index.html : : Understanding, is itself an emotion, i.e. a feeling. : Emotions or feelings can only be "understood" by : consciousness. "Understanding" consciousness can : therefore only be understood by consciousness itself, : therefore the "hard problem" of consciousness, is : intrinsically unsolvable. : : Physics is proven incomplete, that is, no : understanding of the parts of a system can : explain all aspects of the whole of such system. : : |
#22
|
|||
|
|||
"Ruud Broens" wrote in message ... "Kevin Aylward" wrote in message ... : This is an old wife's tale, i.e. a complete and utter nonsense argument. : Less parts do *not* mean a clearer sound. Indeed, more stages mean you : can have less gain per stage allowing for more feedback at each stage : which reduces the distortion tremendously. Hm, Kevin, that's rather oversimplifying things there, isn't it ? Sure, with a couple of BB OPA 2604's you can build a mixing desk at least an order cheaper than with tubes and anything more than 4 channels fully tubed must be called an *investment* rather than an experiment. That said, perceived sonic qualities of a circuit and THD measurements only have a marginal correlation, ** That claim is both appallingly ambiguous AND mischievous. The PROVEN, perceived sound quality of audio circuits with flat response and inaudible levels of THD is "perfect sound". "Perfect sound " = no audible alteration in perceived sound when bypassed. therefore THD is *not* indicative for the sound quality, ** Based on a false premise - hence a false conclusion. still a largely intangible, only a starting point. ** Fools like Ruud imagine that no-one knows whatever it is they do not know. But then, many a thousand pro-tube processing and mixing gear must be bought by fools... ? ** More false reasoning based on a massive red herring. Signal processing by it very nature is out to **ALTER** sound quality - not preserve it. .............. Phil |
#23
|
|||
|
|||
Kevin Aylward wrote: NewYorkDave wrote: I apologize for not providing more detail in my original post. I'll try to give a clearer explanation of what I'm trying to do. I have only a few minutes left in my lunch hour, but I'll try to put it to good use I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) This is an old wife's tale, i.e. a complete and utter nonsense argument. Less parts do *not* mean a clearer sound. Indeed, more stages mean you can have less gain per stage allowing for more feedback at each stage which reduces the distortion tremendously. and perhaps some of that intangible "tube warmth." This where a dirty sound sounds better. However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. So bloody what. A decent modern console uses op-amps with 0.005% thd and imd. It makes no practical difference whether there are quite a few in series, especially as explianed above. A final spec of a decent console is easily 0.01%, 1db flat from 20Hz to 40khz. You are suffering from a severe "golden ears" delusion. Kevin Aylward From what you have said Kevin, what help are you offering the man to help him build a tube mixing desk? Some of the greatest recordings of all time were done using all tube gear. Patrick Turner. http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#24
|
|||
|
|||
"Patrick Turner" Kevin Aylward wrote: So bloody what. A decent modern console uses op-amps with 0.005% thd and imd. It makes no practical difference whether there are quite a few in series, especially as explianed above. A final spec of a decent console is easily 0.01%, 1db flat from 20Hz to 40khz. You are suffering from a severe "golden ears" delusion. Kevin Aylward From what you have said Kevin, what help are you offering the man to help him build a tube mixing desk? ** None of course - there is no need for anyone to offer help to a fool on a fool's mission. Some of the greatest recordings of all time were done using all tube gear. ** That is a mindlessly pompous comment about the musical performance, which if captured well was so done** DESPITE** the use of tubes in the recording chain ****NOT *** because of them. ........... Phil |
#25
|
|||
|
|||
"Walter Harley" wrote ... these days, one expects to pass reasonably clean signals of at least +18dBu Vrms, which would be about 17 volts peak to peak. So your estimate of 1V is a bit of a straw man, I fear - I expect we are closer to the 5% end than the 0.04%, and that is quite audible. Is this audible if only on brief peaks? Excuse my ignorance, but I'm from rat, and 17Vpp seems a lot to me. Is that done with op amps? What would be the typical profile of the distortion from such a console? Nonetheless, the only point I'm trying to make - and I think from your comments that you would agree - is that not all tube circuits sound alike. Since as you say "building a tube audio system is the hard way to go about audio," there must be a reason why the OP wants tubes. If he simply sets about building something that has tubes in it - or, if he applies novel design techniques - there is no reason to expect he will get anything like the sound he or his client has in mind. Exactly, almost. Unless he uses a whole heap of valves very cleverly, there are some kinds of distortion that he is unlikely to achieve. Probably get plenty noise though, but that's another issue. There are many ppl fed up with the hard sound of totally digital and SS systems, and they add some tubes into the chain to warm the sound. It sounds like snake oil, I know, but I have heard the effects with my own ears. I would not dream of getting into a tubes vs. semiconductors argument; it has been repeatedly shown that that is unproductive for all involved! Including the original argument that led to the exodus to SS? And the recent valve revival? Yes the argument grinds on unproductively most of the time, but occasionally it can amount to something. After all, we are wondering why on earth this guy wants to build a valve mixer. Personally, on my hifi I prefer to hear something as close as possible to what the mixing engineer was hearing, so I opt for a system that reproduces with maximal accuracy. Several problems. First, the circularity: you only believe you are hearing the same as the mixing engineer because you believe your system is accurate. Secondly, the poverty of expectation, that your critical mind must focus when listening to music at home upon an anonymous engineer in some distant time and place: there is no art in aspiring to mere reproduction. Thirdly, since "maximal" is not absolute, you cannot escape the need to consider what kinds of inaccuracy are more or less acceptable than others and, having taken that step, you have two major axes, quantity and quality to consider, and it becomes reasonable to expect that more good quality distortion may be better than less of an inferior kind. But that is a personal choice; and it is achievable through a variety of technologies. It should be beyond argument that, if one makes the opposite choice (wishing one's stereo to further modify the sound), any technology can supply the necessary inaccuracies, and different technologies will supply different inaccuracies. Er...that last sentence is self-contradictory surely? In *principle* it may be possible to reproduce the characteristics of a particular circuit by using another technology, and indeed DSP promises exactly that. But actually, and economically, it doesn't work like that. Hence the continuing popularity of valve guitar amplifiers despite the availability of cheap and reliable DSP emulators. A good plastic violin costs a fortune I should imagine, and is successful not because it emulates a good wooden one, but because it sounds like a good violin...there is a distinction there somewhere I'm sure... So the occasional offerings of "valve-like" mosfet amps are silly and miss the point, for me. In general they address only one or two characteristics, and not the right ones. Easier to build a valve amp anyway. Either way, we are talking here about a mixing desk rather than a hi-fi, and as you say it is commonplace (and to my ears often desirable) to introduce intentional inaccuracies during audio production, whether through distortion in the microphones, preamps, mixing desk, analog tape, or other processing gear. I'd be glad to build a tube console if I could find a client to pay for it. I'd start by finding the schematic of a classic tube console. Yes. But what for I wonder? Seems a crazy scheme. Perhaps easier to restore one. Do they come up on Ebay or were they all trashed years ago? cheers, Ian |
#26
|
|||
|
|||
Now, it should be apparent why my original post included the phrase:
"Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!" The discussion was pretty constructive up to the point that I revealed my rerasons for wanting to build a tube mixer. Now the ad-hominem attacks have begun. Without knowing anything about me, two posters have seen fit to call me "deluded" and "a fool on a fool's quest." But to the rest of the posters, I thank you for your constructive criticism and the ideas you've offered. |
#27
|
|||
|
|||
"NewYorkDave" wrote in message om... Now, it should be apparent why my original post included the phrase: "Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!" The discussion was pretty constructive up to the point that I revealed my rerasons for wanting to build a tube mixer. Now the ad-hominem attacks have begun. Without knowing anything about me, two posters have seen fit to call me "deluded" and "a fool on a fool's quest." But to the rest of the posters, I thank you for your constructive criticism and the ideas you've offered. May I apologise on behalf of this NG. We are all nice guys. If we can, we are helpful. We are polite, friendly, nice, people. Ocassionally a Stuka dive bomber wings in with somebody with a silly hat and a smile on his face... We try to ignore them .... they are nothing to do with us.... it is their idea of .....'having a good time' ??!! You're very welcome.. Stick with us.... kind regards jim |
#28
|
|||
|
|||
"jim" wrote in message ... : : "NewYorkDave" wrote in message : om... : Now, it should be apparent why my original post included the phrase: : : "Please, let's ignore for the moment the question : of WHY someone would want to design a tube mixing console in 2003!" : : The discussion was pretty constructive up to the point that I revealed : my rerasons for wanting to build a tube mixer. Now the ad-hominem : attacks have begun. Without knowing anything about me, two posters : have seen fit to call me "deluded" and "a fool on a fool's quest." But : to the rest of the posters, I thank you for your constructive : criticism and the ideas you've offered. : : May I apologise on behalf of this NG. We are all nice guys. If we can, we : are helpful. We are polite, friendly, nice, people. Ocassionally a Stuka : dive bomber wings in with somebody with a silly hat and a smile on his : face... We try to ignore them .... they are nothing to do with us.... it is : their idea of .....'having a good time' ??!! : You're very welcome.. Stick with us.... : kind regards : jim Yeah, Dave, if you actually build something along what i outlined elsewhere I could offer you 2 or 4 of these input tranny's at "a friend's price", since i have a dozen or so...Mind you, them EF86's don't come cheap..Keep us posted there.. Cheers, Rudy |
#29
|
|||
|
|||
Hi RATs!
Feedback is good for hiding problems from test equipment. Life would be simple if all we had to do was fool machines. At work, that is the way it is. At home, we are the test machines. We fool ourselves, but, not the same way we fool machines. Trying to claim that what some machine measures is the 'Ultimate Truth' is just not very helpful when we are trying to increase our private pleasures. All the leading zeroes in the Universe cannot fool anyone but those who demand final answers to the wrong questions. If you care deeply about some distortion measurement, that is your choice. If I care deeply about how my system sounds, that is my choice. It is not worth anyone's time to type insults into newsgroups. Each of us is insane in our own way. Us old farts just accept this and do whatever pleases us. The younger, testosterone poisoned members like to claim they have the straight information. There is no straight information. Thinking about it bends it to fit the mind doing the thinking. You can shoot everyone who disagrees with you, but, it won't make you any smarter, nor, get you closer to the truth. "Good and Bad, I defined these terms. Quite clear, no doubt, somehow, ah, but I was so much older then, I'm younger than that, now." - R.Z. of Hibbing, MN Happy Ears! Al Alan J. Marcy Phoenix, AZ PWC/mystic/Earhead |
#30
|
|||
|
|||
TubeGarden wrote: Hi RATs! Feedback is good for hiding problems from test equipment. Life would be simple if all we had to do was fool machines. At work, that is the way it is. At home, we are the test machines. We fool ourselves, but, not the same way we fool machines. Yup. We can come up with different results each time we fool ourselves. Stupid machines come up with the same boring answers every time. Trying to claim that what some machine measures is the 'Ultimate Truth' is just not very helpful when we are trying to increase our private pleasures. All the leading zeroes in the Universe cannot fool anyone but those who demand final answers to the wrong questions. If you care deeply about some distortion measurement, that is your choice. If I care deeply about how my system sounds, that is my choice. It is not worth anyone's time to type insults into newsgroups. Each of us is insane in our own way. Us old farts just accept this and do whatever pleases us. The younger, testosterone poisoned members like to claim they have the straight information. In a rather brilliant moment (if I dare say so myself) I managed to please both elements within me, the old fart who listens with tiring ears, and the testosterone-poisoned young buck who wanted all the right numbers to go along with the ball-shaking bass from the subs. I designed a switching arrangment to let me add about 10 dB of feedback, or not, while maintaining the same overall gain. Most days I like to listen with NFB on. Sometimes though, especially on days when I feel like listening to Van Halen, I live dangerously and turn off the NFB switch. Hey wait... what's wrong with this picture? The old fart listening with NFB, and the young buck with the hat on backwards listening "straight"! Can't take nuthin' for granted anymore. I think my wife has it straight. "Why put a switch there? If you don't like the sound, put on a different CD!" One thing we do agree on is that a nice bottle of white wine and a snuggle on the couch makes any CD or record sound way better, be damned how the switches are set. There is no straight information. Thinking about it bends it to fit the mind doing the thinking. You can shoot everyone who disagrees with you, but, it won't make you any smarter, nor, get you closer to the truth. "Good and Bad, I defined these terms. Quite clear, no doubt, somehow, ah, but I was so much older then, I'm younger than that, now." - R.Z. of Hibbing, MN Yeah, and there's that W.S. chap somewhere in England, "for there is nothing either good or bad, but thinking makes it so." Happy Ears! Al You too! Cheers, Fred -- +--------------------------------------------+ | Music: http://www3.telus.net/dogstarmusic/ | | Projects: http://dogstar.dantimax.dk | +--------------------------------------------+ |
#31
|
|||
|
|||
I dunno . . . . most SS stuff I've heard with wonderfully low THD, IM, etc.
doesn't sound as good to my ears as some 1% distorting tube sets . . . Simplicity is a good goal. In my experience, less complexity means less chance of failure, and sometimes lower noise. Golden ears may be a delusion, but it's often a pleasant and rather harmless one . . . From: "Kevin Aylward" Organization: AnaSoft Reply-To: "Kevin Aylward" Newsgroups: rec.audio.tubes,sci.electronics.design Date: Tue, 18 Nov 2003 07:29:20 -0000 Subject: A little feedback worse than none at all? NewYorkDave wrote: I apologize for not providing more detail in my original post. I'll try to give a clearer explanation of what I'm trying to do. I have only a few minutes left in my lunch hour, but I'll try to put it to good use I record music on an 8-track multitrack recorder, and I want to build a tube mixing console to take the place of the IC-based model I've been using. My goal is to achieve a clearer sound (due to a simpler circuit with fewer active stages) This is an old wife's tale, i.e. a complete and utter nonsense argument. Less parts do *not* mean a clearer sound. Indeed, more stages mean you can have less gain per stage allowing for more feedback at each stage which reduces the distortion tremendously. and perhaps some of that intangible "tube warmth." This where a dirty sound sounds better. However, I'm not looking to build a "distortion device"... The tube consoles that were used to track the records we all love from the '50s and '60s were designed to be as linear as possible within the limits of the technology of the day, and any euphonic qualities added were incidental. It's not clear to me whether that sound really came from the tubesand transformers or from the more straightforward topologies, anyway. In a modern mixing console, a signal may pass through tens of IC amplifiers on its way from input to output. So bloody what. A decent modern console uses op-amps with 0.005% thd and imd. It makes no practical difference whether there are quite a few in series, especially as explianed above. A final spec of a decent console is easily 0.01%, 1db flat from 20Hz to 40khz. You are suffering from a severe "golden ears" delusion. Kevin Aylward http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#32
|
|||
|
|||
Jon Yaeger wrote:
[snip] Golden ears may be a delusion, but it's often a pleasant and rather harmless one . . . Try telling that to the wives* and credit card companies of some audiophiles... Tim * Husbands, partners, choirboys, whatever. -- And the beast shall be made legion. Its numbers shall be increased a thousand thousand fold. The din of a million keyboards like unto a great storm shall cover the earth, and the followers of Mammon shall tremble. - The Book of Mozilla, 3:31 |
#33
|
|||
|
|||
John Woodgate wrote:
I read in sci.electronics.design that Kevin Aylward kevindotaylwardEXTR wrote (in et) about 'A little feedback worse than none at all?', on Sat, 15 Nov 2003: Thing is, I've come across a reference to a study done by P. J. Baxandall in the '70s, in which he showed that using less than 15dB or so of feedback can actually INCREASE the generation of higher-order distortion products. ANY amount of feedback causes this; Ahmmm, a bit misleading john. If the feedback is large enough, essentially, all distortion will be reduced in practise. No, not misleading at all. You should read more carefully. PJB was quoted as: INCREASE the generation of higher-order distortion products. Yes, and I explained that practical amps *always* have odd harmonics as well as even, and if you use *enough* feedback, the total will still be less. Note 'higher-order'. Your explanation below actually shows how intermodulation between low-order harmonics (and including the fundamental) creates higher-order ones that wouldn't be there if there were no feedback! yes, but it also shows why I am still correct. The deal is that with *low* feedback, a pure square law device will have its input mixed with its output causing generation of 3rd harmonics, and higher, which were not present in the original amp. However, if the FB is large enough, the resulting distortion can easily go to 0.01%. Sure, if the original device was *pure* square, than 0.01% 3rd, is still larger than zero. The reality however, is that all practical devices have odd harmonics as well, such that large feedback will result is a net total reduction of the original distortion at all harmonics of relevance. What is *relevant* is the nasty noise created by the high-order intermodulation distortion. Yes, but.....only if its weighted effect is more objectionable, and it wont if the feedback is large enough. ?Because some intermodulation products due to different orders add in-phase, the intermodulation products can be much larger than the harmonics. Ho hummm....*already* considered in my explanation. And, being unrelated harmonically to the fundamental, they sound much worse than even high odd-order harmonics. Ho hummm...only if there is *low* feedback. Use enough feedback, and all distortion will be less. Attempts at weighting harmonics in terms of their subjective nastiness has more or less settled on weighting by the square of the order. See IEC/EN 60268-3, although this weighting was proposed (by E R Wigan, IIRC) long before it was accepted into the standard. Naturally, weighted harmonic distortion doesn't appear in manufacturers' specifications, because the figures are often unattractively large. As I explained, none of this matters with *large* amounts of feedback. Large feedback will clobber all distortion. You wont here it if is all below 0.01%, and this is easy to do. End of story. Kevin Aylward http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#34
|
|||
|
|||
Well, I've been posting on Usenet on and off for the better part of a
decade, so I know to put on my asbestos suit before entering the building. Still, I can't help but be taken aback when I see a reasonable discussion degenerate into a flame war for no apparent reason. At any rate, I didn't mean to open up a debate on the relative merits of tube versus transistorized mixers... I was really only seeking a discussion of the effects of negative feedback on the distribution of distortion products, and how this might vary depending on the type of active device being used. |
#35
|
|||
|
|||
OTOH, until we know what you are trying to achieve, it is hard to
help. There are a heap of considerations here, concerning where and how the mixing is done, and from and to where the feedback might go. AFAIK, your original concerns were relevant in a different context and may not be applicable here. Have you considered simulation? With spice you can get a reasonably accurate comparison of topologies, with their characteristic distortion and frequency responses. Several folk hanging out on rat use circuitmaker, and there are a range of valve models available. Certainly there is no shortage of models for such as ECC8X and EF86. You do need to find one with enough sophistication to reasonably approximate the distortion. Your plan seems reasonable to me, and clear in outline. One obvious question though: if it were possible to take your kind of short-cuts to console design in the 50s, why didn't they do it? If you take your feedback from the output of the white cathode follower, you will have a ring of three, which is a classic circuit for some reason I can't remember. Have you considered mu-followers for the gain stages, or perhaps just for the second? If allowable you could achieve such by using MOSFETS for the top devices. cheers, Ian (NewYorkDave) wrote in message . com... Now, it should be apparent why my original post included the phrase: "Please, let's ignore for the moment the question of WHY someone would want to design a tube mixing console in 2003!" The discussion was pretty constructive up to the point that I revealed my rerasons for wanting to build a tube mixer. Now the ad-hominem attacks have begun. Without knowing anything about me, two posters have seen fit to call me "deluded" and "a fool on a fool's quest." But to the rest of the posters, I thank you for your constructive criticism and the ideas you've offered. |
#36
|
|||
|
|||
I read in sci.electronics.design that Kevin Aylward kevindotaylwardEXTR
wrote (in ..net) about 'A little feedback worse than none at all?', on Tue, 18 Nov 2003: As I explained, none of this matters with *large* amounts of feedback. Large feedback will clobber all distortion. You wont here it if is all below 0.01%, and this is easy to do. End of story. This is simply not true, as has been found time after time by people who thought that 60 dB of feedback would cure all ills. When you look at the IM distortion, that 0.01% is a sick joke. -- Regards, John Woodgate, OOO - Own Opinions Only. http://www.jmwa.demon.co.uk Interested in professional sound reinforcement and distribution? Then go to http://www.isce.org.uk PLEASE do NOT copy news posts to me by E-MAIL! |
#37
|
|||
|
|||
NewYorkDave wrote: Well, I've been posting on Usenet on and off for the better part of a decade, so I know to put on my asbestos suit before entering the building. Still, I can't help but be taken aback when I see a reasonable discussion degenerate into a flame war for no apparent reason. At any rate, I didn't mean to open up a debate on the relative merits of tube versus transistorized mixers... I was really only seeking a discussion of the effects of negative feedback on the distribution of distortion products, and how this might vary depending on the type of active device being used. I haven't been involved with this thread much, but decided I'd jump in here. An exact answer would depend on more variables than can be readily covered in a usenet discussion. But for an approximate answer, the images shown at the link below might be helpful: http://www.dogstar.dantimax.dk/tubestuf/ampspecs.htm [scroll about 1/2-way down. The green bars ("clean") are with NFB, and the red ones ("classic") without.] This is with or without about 6 dB of global feedback around the driver, UL PP output stage, and output transformer. As you can see by comparing the two images, there is a significant difference in how the reduction in harmonic products is distributed *based on output power*. At the 12 watt level, the most significant reduction was to the 6th harmonic. However, at the 54 watt level, the greatest drop is in the 4th harmonic. Also note how the other differences are distributed. Note also that these graphs are done with a fixed resistance as the output load. Fluctuations in speaker impedance would greatly change the harmonic distribution, and probably accentuate the relative differences with or without NFB also. What's the audible difference between the two settings? I find that at low listening levels, the "clean" setting has a noticeably tighter bass, and more restrained highs. But the difference is not that great. At higher levels, the NFB setting sounds increasingly dry compared to the "classic" mode. The classic mode sounds punchier, richer, more interesting to listen to, but also more tiring over a long time. There seems to be a point somewhere in the middle, where the "euphonious" nature of the "classic" mode reaches a peak. I'd hazard a guess at around the -10 dB point. At high levels, the classic mode has a decidedly "rangey" sound to it, and the relative dryness of the "clean" setting is a welcome relief. But the threshold of clipping distortion is much more well-defined. In the "classic" mode you can tolerate a fair amount of clipping on the largest peaks, without noticing unless you listen really closely. Not so in the "clean" mode, when it falls apart it does so quite noticably. It's pretty easy to set the volume control for maximum undistorted output! Cheers, Fred -- +--------------------------------------------+ | Music: http://www3.telus.net/dogstarmusic/ | | Projects, Vacuum Tubes & other stuff: | | http://www.dogstar.dantimax.dk | +--------------------------------------------+ |
#38
|
|||
|
|||
Whoa !
Long Post! ) -- -- .................................................. ........................ Choky Prodanovic Aleksandar YU "don't use force, "don't use force, use a larger hammer" use a larger tube - Choky and IST" - ZM .................................................. ........................... "TubeGarden" wrote in message ... Hi RATs! Feedback is good for hiding problems from test equipment. Life would be simple if all we had to do was fool machines. At work, that is the way it is. At home, we are the test machines. We fool ourselves, but, not the same way we fool machines. Trying to claim that what some machine measures is the 'Ultimate Truth' is just not very helpful when we are trying to increase our private pleasures. All the leading zeroes in the Universe cannot fool anyone but those who demand final answers to the wrong questions. If you care deeply about some distortion measurement, that is your choice. If I care deeply about how my system sounds, that is my choice. It is not worth anyone's time to type insults into newsgroups. Each of us is insane in our own way. Us old farts just accept this and do whatever pleases us. The younger, testosterone poisoned members like to claim they have the straight information. There is no straight information. Thinking about it bends it to fit the mind doing the thinking. You can shoot everyone who disagrees with you, but, it won't make you any smarter, nor, get you closer to the truth. "Good and Bad, I defined these terms. Quite clear, no doubt, somehow, ah, but I was so much older then, I'm younger than that, now." - R.Z. of Hibbing, MN Happy Ears! Al Alan J. Marcy Phoenix, AZ PWC/mystic/Earhead |
#39
|
|||
|
|||
John Woodgate wrote:
I read in sci.electronics.design that Kevin Aylward kevindotaylwardEXTR wrote (in .net) about 'A little feedback worse than none at all?', on Tue, 18 Nov 2003: As I explained, none of this matters with *large* amounts of feedback. Large feedback will clobber all distortion. You wont here it if is all below 0.01%, and this is easy to do. End of story. This is simply not true, Yes it is. as has been found time after time by people who thought that 60 dB of feedback would cure all ills. It most certainly does, if it is applied correctly. That means that there is 60 db, i.e. if it is slew limiting than the feedback aint 60db, it goes to zero, hence violates the assumption of 60db of feedback. When you look at the IM distortion, that 0.01% is a sick joke. Rubbish. Complete and utter crap. What drugs are you on? For starters, consider the FM distortion of speakers. i.e. consider a 1khz signal riding on a speaker going back and forth at 100 Hz. 0.05 to 0.1% plus dopplar induced distortion is easily achievable. And this is really nasty stuff. Its a spread right across the spectrum. Adding in the standard 1% to 10% of normal speaker THD/IMD, and there is no chance whatsoever that your claim is supported. There is no way that one can audiable detect 0.01% THD/IMD levels from correctly designed amps. If it was frequncy shifts, maybe, but not distortion. Been there done, it, wrote the book. Kevin Aylward http://www.anasoft.co.uk SuperSpice, a very affordable Mixed-Mode Windows Simulator with Schematic Capture, Waveform Display, FFT's and Filter Design. http://www.anasoft.co.uk/replicators/index.html Understanding, is itself an emotion, i.e. a feeling. Emotions or feelings can only be "understood" by consciousness. "Understanding" consciousness can therefore only be understood by consciousness itself, therefore the "hard problem" of consciousness, is intrinsically unsolvable. Physics is proven incomplete, that is, no understanding of the parts of a system can explain all aspects of the whole of such system. |
#40
|
|||
|
|||
The PROVEN, perceived sound quality of audio circuits with flat response and inaudible levels of THD is "perfect sound". "Perfect sound " = no audible alteration in perceived sound when bypassed. therefore THD is *not* indicative for the sound quality, ** Based on a false premise - hence a false conclusion. Maybe my ears are defective, but I find that: (close to 0% distortion) = best sound, A little 2nd harmonic up to -40dB down and no more than -50dB 3rd sounds pleasant, More than -50dB of 3rd = yuck. It -might- be more meaningful to compare amps by looking at specs specifying how much 2nd H, how much 3rd H, etc listed separately. An amp with say -40dB 2nd and -50dB 3rd is likely to sound better than one with -50dB 2nd and -40dB 3rd. But both will have the same THD spec (yes?). But then, many a thousand pro-tube processing and mixing gear must be bought by fools... ? Maybe the issue is that tubes are said to permit larger transient sounds thru without hard clipping, vs op-amp designs. But that might not mean much as the recording device will likely clip that transient anyway. |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Recommendations for 5 channel power amp | High End Audio | |||
Need Advise on Feedback / Feedback Eliminators | Pro Audio | |||
Speaker feedback w/PC connection | Tech | |||
Passive RIAA VS feedback RIAA preamp | Vacuum Tubes | |||
Followers and feedback | Tech |