Wavebourn, thanks for the big info pak.
I believe it is going to be very fruitful.
As you will see, I'm not all that up on the software packages for handing such. I downloaded the free trial version of WinRAR, and unzipped it. It becomes 3 .tif's and 5 .pdf's.
I am happy to redistribute this. If I do, I think I will do it with just the zip function built into windows, or even just file by file.
I can send up to 10 meg at once with gmail.
brian.zenmaster@gmail.com
The .pdf's work fine. But I need something better for the .tif's. They are so big that my computer almost stops cold.
They are excellent and well annotated schematics, and I will derive much from them.
Just to get started, may I ask just a few questions about them?
What exactly is this material for, a recording console? What make? When was it made? Have you used it?
I see the B+ and B- notation. This is usually seen on Asian schematics.
I see opamp symbols, do you know which ones they are using?
This thread has taken off. I'm glad there are people who have such console experience. Are these analog consoles still used? You need to have multitrack tape decks to go with them. Are these consoles now museum pieces?
I think a good place to start is the power supplies. What voltage rails do they have, and how do they use them. I mean, are they used bare, or do they send higher voltages and have some additional drops.
Do they have linear regulation on each channel? ( more heat )
Do they have feed foward diodes and extra capacitance. These diodes would not really help here, that is more a technique used to get the full rail voltage back in poweramps that have incomplete filtering.
The 48 volts, does that have any channel isolation or regulation? Do they send out exactly 48 volts from the power supplies, or more?
Does the microphone cable get exactly 48.0 volts, or a bit less due to regulation. Do they use the 48 volts for anything else?
I believe it is going to be very fruitful.
As you will see, I'm not all that up on the software packages for handing such. I downloaded the free trial version of WinRAR, and unzipped it. It becomes 3 .tif's and 5 .pdf's.
I am happy to redistribute this. If I do, I think I will do it with just the zip function built into windows, or even just file by file.
I can send up to 10 meg at once with gmail.
brian.zenmaster@gmail.com
The .pdf's work fine. But I need something better for the .tif's. They are so big that my computer almost stops cold.
They are excellent and well annotated schematics, and I will derive much from them.
Just to get started, may I ask just a few questions about them?
What exactly is this material for, a recording console? What make? When was it made? Have you used it?
I see the B+ and B- notation. This is usually seen on Asian schematics.
I see opamp symbols, do you know which ones they are using?
This thread has taken off. I'm glad there are people who have such console experience. Are these analog consoles still used? You need to have multitrack tape decks to go with them. Are these consoles now museum pieces?
I think a good place to start is the power supplies. What voltage rails do they have, and how do they use them. I mean, are they used bare, or do they send higher voltages and have some additional drops.
Do they have linear regulation on each channel? ( more heat )
Do they have feed foward diodes and extra capacitance. These diodes would not really help here, that is more a technique used to get the full rail voltage back in poweramps that have incomplete filtering.
The 48 volts, does that have any channel isolation or regulation? Do they send out exactly 48 volts from the power supplies, or more?
Does the microphone cable get exactly 48.0 volts, or a bit less due to regulation. Do they use the 48 volts for anything else?
Anatech,
<<
BTW, the total tail current will not change at all for the sake of argument. It just seesaws back and forth. The gate is a much higher impedance at low audio frequencies and therefore will show a higher signal on it's gates. Not a problem at all in op amps or discrete circuits.
>>
Opamp designs focus on minimizing changes in tail current. But they also try to minimize the seesawing back and forth too. This is minimized by trying to get the highest possible second stage gain.
But still, you will have some of this seesawing. It actually depends on the output loading.
So you have tail current variations. You have 2nd stage loading currents. You have the seesawing current variations.
It is quite complex, because you have inverting mode and noninverting mode. Then for each you have the condition of high gain, near unity gain, and unity gain.
Each of these six cases will have a different mix of these current variations.
Likewise, how low of an impedance you are driving at the output will effect how big the 2nd stage loading variations are.
So there are quite a few combinations. I'm am sure this is a factor confusing text treatments of such matters.
But also, these various cases should each be understood because such understanding should influence overall signal flow architecture, and the choice of circuit types for each postion in the circuit.
<<
BTW, the total tail current will not change at all for the sake of argument. It just seesaws back and forth. The gate is a much higher impedance at low audio frequencies and therefore will show a higher signal on it's gates. Not a problem at all in op amps or discrete circuits.
>>
Opamp designs focus on minimizing changes in tail current. But they also try to minimize the seesawing back and forth too. This is minimized by trying to get the highest possible second stage gain.
But still, you will have some of this seesawing. It actually depends on the output loading.
So you have tail current variations. You have 2nd stage loading currents. You have the seesawing current variations.
It is quite complex, because you have inverting mode and noninverting mode. Then for each you have the condition of high gain, near unity gain, and unity gain.
Each of these six cases will have a different mix of these current variations.
Likewise, how low of an impedance you are driving at the output will effect how big the 2nd stage loading variations are.
So there are quite a few combinations. I'm am sure this is a factor confusing text treatments of such matters.
But also, these various cases should each be understood because such understanding should influence overall signal flow architecture, and the choice of circuit types for each postion in the circuit.
Continued:
So we have current variations in these input devices.
The output distortion of an opamp circuit is largely determined by the accuracy of the input summation of the two signals. Those being the input, and and the feedback signals.
Remember, the more primitive noninverting circuit form only has one input transistor, and takes the feedback via the emitter.
So, there is an offset difference in this summation, or really this differencing.
The older circuit form doesn't really work DC coupled becase of this offset voltage.
So going to the full diffpair circuit is a way to correct for this offset.
With the full diffpair you can go DC coupled.
So, if this offset voltage is changing during the signal swing, you have error in the differencing and you have the largest source of opamp distortion.
So we have current variations in these input devices.
The output distortion of an opamp circuit is largely determined by the accuracy of the input summation of the two signals. Those being the input, and and the feedback signals.
Remember, the more primitive noninverting circuit form only has one input transistor, and takes the feedback via the emitter.
So, there is an offset difference in this summation, or really this differencing.
The older circuit form doesn't really work DC coupled becase of this offset voltage.
So going to the full diffpair circuit is a way to correct for this offset.
With the full diffpair you can go DC coupled.
So, if this offset voltage is changing during the signal swing, you have error in the differencing and you have the largest source of opamp distortion.
Continued:
So for a given condition, you will have this input transistor current variation.
Is it better to go BJT or FET?
Well, how much Vbe or Vgs varies for the projected level of current variation, will determine how much distortion you have.
This means that you get the lowest distortion with the device of higher transconductance. This would be a BJT, and a high biased one at that.
So for a given condition, you will have this input transistor current variation.
Is it better to go BJT or FET?
Well, how much Vbe or Vgs varies for the projected level of current variation, will determine how much distortion you have.
This means that you get the lowest distortion with the device of higher transconductance. This would be a BJT, and a high biased one at that.
Hi zenmasterbrian,
You can easily confuse yourself by drilling too deep into semiconductor behavior if you wish. Jfets perform the voltage to current conversion internally unlike BJT's where that conversion occurs outside the device, and therefore may result in an uncompensated distortion before the signal even enters the amplifier. Keep things on a macro level if you ever want a discussion to go anywhere. It's like a meeting. Keep on track.
The very best way to look at these comparisons would be to construct an amplifier or signal stage and play. Experiment and allow that to guide your studies, because a theoretical discussion can wind on for years. When you observe a behavior you can't explain easily, ask and you will get mostly meaningful answers. Allow reality to impinge on theory and look at the other factors that may modify what you need to do in real life.
If you study op amp circuitry you will see things done on chip that simply don't work well using discrete components. This is an example where you need to factor reality and imperfect components into your studies.
-Chris
You can easily confuse yourself by drilling too deep into semiconductor behavior if you wish. Jfets perform the voltage to current conversion internally unlike BJT's where that conversion occurs outside the device, and therefore may result in an uncompensated distortion before the signal even enters the amplifier. Keep things on a macro level if you ever want a discussion to go anywhere. It's like a meeting. Keep on track.
The very best way to look at these comparisons would be to construct an amplifier or signal stage and play. Experiment and allow that to guide your studies, because a theoretical discussion can wind on for years. When you observe a behavior you can't explain easily, ask and you will get mostly meaningful answers. Allow reality to impinge on theory and look at the other factors that may modify what you need to do in real life.
If you study op amp circuitry you will see things done on chip that simply don't work well using discrete components. This is an example where you need to factor reality and imperfect components into your studies.
-Chris
Continued:
So the argument for JFET inputs pertaining to distortion is not of less distortion, but of different distortion.
The argument is that it can be lower order and even order, and hence more listenable.
I'm not sure that this really does pan out.
But even so, the harmonic distortion is much much higher.
Also, this means much higher levels of intermodulation products.
While lower order and even order harmonic distortion may be more desireable, this does not apply to intermodulation distortion.
So the argument for JFET inputs pertaining to distortion is not of less distortion, but of different distortion.
The argument is that it can be lower order and even order, and hence more listenable.
I'm not sure that this really does pan out.
But even so, the harmonic distortion is much much higher.
Also, this means much higher levels of intermodulation products.
While lower order and even order harmonic distortion may be more desireable, this does not apply to intermodulation distortion.
Continued:
So D. Self and G. R. Slone reject the JFETs for different distortion option.
But Slone likes a double diffpair option. Use two complementary diffpairs.
This substantially lower odd order distortion. On most opamp like power amp circuits, the 3rd order distortion does start to increase radically above a couple of khz.
The mirroring option cuts this way way down, while giving only modest increases in second order distortion.
So D. Self and G. R. Slone reject the JFETs for different distortion option.
But Slone likes a double diffpair option. Use two complementary diffpairs.
This substantially lower odd order distortion. On most opamp like power amp circuits, the 3rd order distortion does start to increase radically above a couple of khz.
The mirroring option cuts this way way down, while giving only modest increases in second order distortion.
I agree with Anatech, you're over-analysing things. Get yourself a breadboard, a few chips, and RightMark analyser for your PC.
Continued:
Anatech, who is Mr. Curl? An author?
<<
Could have been the technology of the time for Mr. Self and Mr Slone. I wasn't there, so I have no idea why they don't like them.
>>
Self and Slone are contemporary.
<<
My personal least favorite. This topology did allow for higher tail currents so that DC offset voltage could be kept down. I like fets even more for this in a straight differential. Cascoded of course!
>>
Do you have a particular text or article which analyzes the circuit variations you are describing? Do you have a schematic? One where this is actually used?
Anatech, who is Mr. Curl? An author?
<<
Could have been the technology of the time for Mr. Self and Mr Slone. I wasn't there, so I have no idea why they don't like them.
>>
Self and Slone are contemporary.
<<
My personal least favorite. This topology did allow for higher tail currents so that DC offset voltage could be kept down. I like fets even more for this in a straight differential. Cascoded of course!
>>
Do you have a particular text or article which analyzes the circuit variations you are describing? Do you have a schematic? One where this is actually used?
Continued:
<<
Discrete circuits. Simpler that most op amps, or more complicated. Most do not require separate supplies as they operate in class A typically.
>>
You are referring not only to something that is not only class A, but to something that has single ended output, like a single emitter follower.
The use of such circuits was common. Of course tube circuits would use a cathode follower. This was common
If the emitter follower is the output stage of an opamp being used closed loop, then the distortion can be low.
But if used open loop, and driving a resistive load, there is still distortion due to Vbe variation with the device current.
If instead the emitter follower is driving a current source, and not much else. Then it can have very low distortion. But you will see, this starts to resemble the opamp, just the diffpair common node becomes the output.
If you use these single ended output ciruits, the distortion they make is odd order. If they drive resistive loads, the distortion is greater than monolithic opamps.
<<
Discrete circuits. Simpler that most op amps, or more complicated. Most do not require separate supplies as they operate in class A typically.
>>
You are referring not only to something that is not only class A, but to something that has single ended output, like a single emitter follower.
The use of such circuits was common. Of course tube circuits would use a cathode follower. This was common
If the emitter follower is the output stage of an opamp being used closed loop, then the distortion can be low.
But if used open loop, and driving a resistive load, there is still distortion due to Vbe variation with the device current.
If instead the emitter follower is driving a current source, and not much else. Then it can have very low distortion. But you will see, this starts to resemble the opamp, just the diffpair common node becomes the output.
If you use these single ended output ciruits, the distortion they make is odd order. If they drive resistive loads, the distortion is greater than monolithic opamps.
anatech and pinkmouse,
I think this is turning out to be one of the best threads yet.
I don't know how old you guys are, but I was playing with oscilloscopes and transistor and tube circuits 30 years ago.
I suggest that you guys get the G. R. Slone book, and then the D. Self book and read them. The stage by stage analysis they give, in the quest for lowest noise and distortion, is hundreds of pages long.
The distortion levels of modern opamp circuits, monolithic and discrete, are very very low. This is the result of a design science that has evolved over many decades.
A while back we had a Jim Williams quote. He is one of the recognized leaders in such design science.
Another excellent book in my hands right now is Intuitive Operational Amplifiers, by Thomas M. Frederiksen, 3rd ed.
Frederiksen talks about some of these very same issues. He is a leader in the design of monolithic opamps. These chips are made to perform. Its not subjectivism, its a merciless competition in the semiconductor industry.
D. Self has been involved in very agressive recording console designs, trying to maximize SNR and minimize distortion.
I believe that much could be learned from such.
I think this is turning out to be one of the best threads yet.
I don't know how old you guys are, but I was playing with oscilloscopes and transistor and tube circuits 30 years ago.
I suggest that you guys get the G. R. Slone book, and then the D. Self book and read them. The stage by stage analysis they give, in the quest for lowest noise and distortion, is hundreds of pages long.
The distortion levels of modern opamp circuits, monolithic and discrete, are very very low. This is the result of a design science that has evolved over many decades.
A while back we had a Jim Williams quote. He is one of the recognized leaders in such design science.
Another excellent book in my hands right now is Intuitive Operational Amplifiers, by Thomas M. Frederiksen, 3rd ed.
Frederiksen talks about some of these very same issues. He is a leader in the design of monolithic opamps. These chips are made to perform. Its not subjectivism, its a merciless competition in the semiconductor industry.
D. Self has been involved in very agressive recording console designs, trying to maximize SNR and minimize distortion.
I believe that much could be learned from such.
Hi zenmasterbrian,
-Chris
I think you just hurt John's feelings. He actually designs and builds very high end preamplifiers and power amplifiers. His lower end designs can be seen by studying a Parasound amplifier. His contributions here have been noted. Please research this thread on the Blowtorch preamplifier. Certainly pushing the limits of technology. Understand that nothing is given, no schematics. Technology is discussed and is an excellent read. It may take you a while, but I recommend it highly.
-Chris
Hi zenmasterbrian,
I've read a fair amount, but we've learned so much in the last little while even.
You need to play with these things in your hands to get a good feel. Really, start building stuff. An author can talk about things for ages, but it's all hearsay until you do it yourself. You wouldn't let a 16 year old on the road just passing a book test, would you?
One great book you really ought to get is "Valve Amplifiers" by Morgan Jones. Yes, it deals with vacuum tubes. He does very well at explaining real world issues and how they may impact design. I believe it has a bearing on audio in general.
-Chris
I've been servicing audio electronic for that long at least. I was a bad boy when it came to things I tried as a young lad.
I've read a fair amount, but we've learned so much in the last little while even.
You need to play with these things in your hands to get a good feel. Really, start building stuff. An author can talk about things for ages, but it's all hearsay until you do it yourself. You wouldn't let a 16 year old on the road just passing a book test, would you?
One great book you really ought to get is "Valve Amplifiers" by Morgan Jones. Yes, it deals with vacuum tubes. He does very well at explaining real world issues and how they may impact design. I believe it has a bearing on audio in general.
-Chris
Anatech, I've built tube amps and solid state amps, and serviced equipment of all types from the 1930's to the present.
Most of my early electronics experiences were with tube hifi's and TV's
I've done my share of tinkering. But I've also designed circuits that have to meet rigourous specifications for both the commercial market, and defense contracts.
I was in charge of microwave transistor modelling for a major defense and commercial microwave company. I was the principle software developer for a commercialized SPICE simulation program. I ran the technology department for an integrated circuits start up.
I didn't start this thread to talk about stuff like that. I started it because I think there can be something learned from the design of the best recording consoles, about how to get the lowest feasible noise and distortion. It takes more than tinkering to do that.
Mostly, I've been encouraged to think this way by D. Self.
Pinkmouse,
Take a look at Slone's discussion of using the second and complementary diffpairs. Then see how he shows how it really reduces odd order distortion to miniscule levels. I like his approach.
You can also see where he argues against JFETs for their distortion properties.
I threw the JFETs issue out as a semi-strawman. I don't really go along with it. But I still think there could be some situation where it might be good.
What would really help here is an easy way to draw sketches and post them, and even be able to use mathematical notation.
Most of my early electronics experiences were with tube hifi's and TV's
I've done my share of tinkering. But I've also designed circuits that have to meet rigourous specifications for both the commercial market, and defense contracts.
I was in charge of microwave transistor modelling for a major defense and commercial microwave company. I was the principle software developer for a commercialized SPICE simulation program. I ran the technology department for an integrated circuits start up.
I didn't start this thread to talk about stuff like that. I started it because I think there can be something learned from the design of the best recording consoles, about how to get the lowest feasible noise and distortion. It takes more than tinkering to do that.
Mostly, I've been encouraged to think this way by D. Self.
Pinkmouse,
Take a look at Slone's discussion of using the second and complementary diffpairs. Then see how he shows how it really reduces odd order distortion to miniscule levels. I like his approach.
You can also see where he argues against JFETs for their distortion properties.
I threw the JFETs issue out as a semi-strawman. I don't really go along with it. But I still think there could be some situation where it might be good.
What would really help here is an easy way to draw sketches and post them, and even be able to use mathematical notation.
Hi zenmasterbrian,
Interestingly, the Phase Linear / SAE / early Adcom (no surprise) / Bryston inputs are Slone induced. 😉 I don't like the sound of any of them, even the Hafler DH2x0 series.
Cleaner sounding units made by Luxman / Marantz / Sansui and others use fet inputs in a standard cascoded arrangement.
There may be other factors at work, but you should try to relate your own listening experience among different amplifiers. This way you can begin to get a feel how these circuits have a characteristic sound. Then draw your own conclusions. I have improved the sound of complementary diff pairs through very careful matching of all four transistors. Even with smt parts as used in the Cyrus IIIi.
You can't analyze everything from a math standpoint until you can factor everything in. That is a very tall order. It's easier to just build the thing some times.
-Chris
Working on tube gear is relaxing. A couple of my radios are earlier. The solutions to problems are ingenious when you think of what they had to work with.
Interestingly, the Phase Linear / SAE / early Adcom (no surprise) / Bryston inputs are Slone induced. 😉 I don't like the sound of any of them, even the Hafler DH2x0 series.
Cleaner sounding units made by Luxman / Marantz / Sansui and others use fet inputs in a standard cascoded arrangement.
There may be other factors at work, but you should try to relate your own listening experience among different amplifiers. This way you can begin to get a feel how these circuits have a characteristic sound. Then draw your own conclusions. I have improved the sound of complementary diff pairs through very careful matching of all four transistors. Even with smt parts as used in the Cyrus IIIi.
You can't analyze everything from a math standpoint until you can factor everything in. That is a very tall order. It's easier to just build the thing some times.
-Chris
<<
Working on tube gear is relaxing.
>>
Very very true.
<<
The solutions to problems are ingenious when you think of what they had to work with.
>>
Yes. Bob Moog says this too.
<<
Interestingly, the Phase Linear / SAE / early Adcom (no surprise) / Bryston inputs are Slone induced. I don't like the sound of any of them, even the Hafler DH2x0 series.
>>
You mean G. R. Slone? You mean his book influenced them? "High Power Audio Amplifier Construction Manual". That would seem unlikely.
Do you mean they use that type of mirroring of the diff pair approach?
Do you have schematics to any of these?
<<
Cleaner sounding units made by Luxman / Marantz / Sansui and others use fet inputs in a standard cascoded arrangement.
>>
Schematics to any of these?
Sources of distortion, the nonlinearities can be identified and analyzed. Lots of things sound nice. Some people like some types of distortions. Circuits can be made to add it.
G. R. Slone and D. Self argue against the subjectivist camp.
Working on tube gear is relaxing.
>>
Very very true.
<<
The solutions to problems are ingenious when you think of what they had to work with.
>>
Yes. Bob Moog says this too.
<<
Interestingly, the Phase Linear / SAE / early Adcom (no surprise) / Bryston inputs are Slone induced. I don't like the sound of any of them, even the Hafler DH2x0 series.
>>
You mean G. R. Slone? You mean his book influenced them? "High Power Audio Amplifier Construction Manual". That would seem unlikely.
Do you mean they use that type of mirroring of the diff pair approach?
Do you have schematics to any of these?
<<
Cleaner sounding units made by Luxman / Marantz / Sansui and others use fet inputs in a standard cascoded arrangement.
>>
Schematics to any of these?
Sources of distortion, the nonlinearities can be identified and analyzed. Lots of things sound nice. Some people like some types of distortions. Circuits can be made to add it.
G. R. Slone and D. Self argue against the subjectivist camp.
Hi zenmasterbrian,
Schematics are available on the web. www.sharefx.com comes to mind. What you need to do is find equipment to listen to and get the manuals for them. Correlate what you hear. That is the very best advice I can give you.
-Chris
I do too, but at some point it either sounds good or it doesn't. The final arbitrator.
No, but they use the configuration, complementary differential pairs.
Schematics are available on the web. www.sharefx.com comes to mind. What you need to do is find equipment to listen to and get the manuals for them. Correlate what you hear. That is the very best advice I can give you.
-Chris
- Status
- Not open for further replies.