J
Jack Hammer
New member
With reference to preaps and such, what is the difference between cip and discrete in practical terms. Which is more desirable and for what reasons, under what circumstances and for which applications?
Roel
That SMART guy.
Whoa! Heavy question...
Discrete has longer lines in between components, more contacts that can loosen, maybe more current circulating...
That's about all I can think of now... hehe. Not much, but it's a start.
Discrete has longer lines in between components, more contacts that can loosen, maybe more current circulating...
That's about all I can think of now... hehe. Not much, but it's a start.
Dom Franco
New member
Basic Electronics 101
A discrete component is a Transistor, Capacitor, resistor, coil etc.
A "Chip" is an Integrated circuit, that is packaged together with several Transistors, caps, resitors...
NOW SOME HISTORY>>> (The dates are approximate)
Early 1900's....first discrete components combined into simple circuits by point to point wiring (Vacuum tubes)
1950's Printed circuits were made by depositing metal lines on an insulated board. (My Dad was one of the first technicians that created these boards he worked for a company called "Photocircuits.") (These were sort of giant chips)
1957 the Transistor was invented and began replacing tubes and soon an integretated circuit was formed by combining several components in a silicon package.
1960's The space race accelerated the development of smaller and smaller components, chips were made smaller but eventually holding 1000's of transistors.
1970's Guitarists finnally figured out that the old tube amps sounded better than transistorized amps.
1980's Guitarists finnally figured out that point to point wiring sounded better than printed circuits.
1990's Guitarists finnally figured out that discrete components sounded better than chips.
:OKAY?
Dom
A discrete component is a Transistor, Capacitor, resistor, coil etc.
A "Chip" is an Integrated circuit, that is packaged together with several Transistors, caps, resitors...
NOW SOME HISTORY>>> (The dates are approximate)
Early 1900's....first discrete components combined into simple circuits by point to point wiring (Vacuum tubes)
1950's Printed circuits were made by depositing metal lines on an insulated board. (My Dad was one of the first technicians that created these boards he worked for a company called "Photocircuits.") (These were sort of giant chips)
1957 the Transistor was invented and began replacing tubes and soon an integretated circuit was formed by combining several components in a silicon package.
1960's The space race accelerated the development of smaller and smaller components, chips were made smaller but eventually holding 1000's of transistors.
1970's Guitarists finnally figured out that the old tube amps sounded better than transistorized amps.
1980's Guitarists finnally figured out that point to point wiring sounded better than printed circuits.
1990's Guitarists finnally figured out that discrete components sounded better than chips.
:OKAY?
Dom
J
Jack Hammer
New member
Let me see now, do I have it straight?
Thank you both for your replies. So, A chip then is a miniaturized version of a circut that originally would have used transistors, resistors etc. So discrete then is the old fashioned way of making a circuit with the actual compenent parts whatever they may be whereas a chip is a siliconesque imitation of a discrete circuit. I realize that even if this is close to correct, it is a gross oversimplification. And I assume that the part of the history lesson that dealt with the epiphany had by guitarists en masse whereby tube where discovered to sound "better" was, by extension, a similar similar event as the one that is currently (pardon the pun) taking shape in the digital world whereby discrete systems are believed to sound better than chip based electronics.
If this is close, I am grateful, if not, please try and help me unless I am hoplessly lost in which case you can leave me to flounder in the sea of commercial hype used to sell electronic gear to unsuspecting and, in fact, unknowledgable, frankly idiotic at times, consumers that, becuase of their lack of basic electronic knowledge, are prone to buy into any advertising gimmick or actual or virtual gimmick and end up wasting money on, among other things, triphazors and the like.
Thank you both for your replies. So, A chip then is a miniaturized version of a circut that originally would have used transistors, resistors etc. So discrete then is the old fashioned way of making a circuit with the actual compenent parts whatever they may be whereas a chip is a siliconesque imitation of a discrete circuit. I realize that even if this is close to correct, it is a gross oversimplification. And I assume that the part of the history lesson that dealt with the epiphany had by guitarists en masse whereby tube where discovered to sound "better" was, by extension, a similar similar event as the one that is currently (pardon the pun) taking shape in the digital world whereby discrete systems are believed to sound better than chip based electronics.
If this is close, I am grateful, if not, please try and help me unless I am hoplessly lost in which case you can leave me to flounder in the sea of commercial hype used to sell electronic gear to unsuspecting and, in fact, unknowledgable, frankly idiotic at times, consumers that, becuase of their lack of basic electronic knowledge, are prone to buy into any advertising gimmick or actual or virtual gimmick and end up wasting money on, among other things, triphazors and the like.
Roel
That SMART guy.
You are partly correct... The part about discrete being all little seperate components is right, chips ARE the components put on a small piece of silicon and in a little black box. The problem here comes with resistors and capacitors. While with discrete components you have enough space, on a piece of silicon, space is quite 'expensive', and making resistors as big as they are in discrete versions is not possible. It also brings problems with heat dissipation etc... But that's about all you need to know.
Actually I never heard of the discrete circuits being better-thingy... I think it can be quite hard to get the quality and freqrespons of the quality opamps that are currently available... Oh well...
This is the part that's wrong... The tube part is right. But when you say that in the digital world the sa,e is happening you are wrong. Digital has no sound. I mean, it's all 1's and 0's that are being processed, so here the difference between the discrete and silicon versions of the circuit won't add to the sound. It really is what the circuit does that makes/breaks the sound. Discrete parts in digital would be simple calculations; and slower... Not better... Worse. So, if discrete sounds better, it would still be analog discrete circuits that sound better as analog chips...
Actually I never heard of the discrete circuits being better-thingy... I think it can be quite hard to get the quality and freqrespons of the quality opamps that are currently available... Oh well...
skippy
New member
Right. Nerdy EE junk follows- tune out if not interested.
Integrated circuit opamps have a few advantages, and a few disadvantages. The primary advantage is cost: they put most of the circuitry needed to implement your preamp into one single little black plastic package with a lot of legs. The primary disadvantage is they they are designed to be used in many, many different applications (from instrumentation through industrial controls to whatever else), and audio is only one of those- and a very small one at that.
So they aren't optimized for audio. They aren't optimized for much of anything, in fact. But they are as convenient as hell to design with, and cheap- so you can get wonderfully adequate products, and even some stellar ones.
Now with discrete components, you can use different circuit structures (like Class A amplifiers for example), and you can optimize every component in the chain for that particular audio application. You don't have to band-aid around the fact that the circuitry was really designed for something else, and you are just borrowing it.... In the case of mic preamps in particular, you can design front end stages that have orders of magnitude lower noise than you can ever hope to achieve with an integrated opamp. Similarly, you can optimize for distortion, or for bandwidth- and you can implement an output stage that doesn't require compromises on your *input* stage.
Ever notice how some low-end IC preamps get noisier with you drive a heavy load than a light (bridging) one? That's because the heavily loaded output stage heats up the chip, which changes the bias current on the input stage- they are thermally coupled, because they live on the same piece of silicon... The designer sand giveth, and the designer sand taketh away.
With opamp-based design, you are really constrained in what you can do. With discrete design, you can do much more, since every component is under your control- and sometimes that can sound better. In the hands of a good design engineer, it can sound _magnificent_, and the opamp-based stuff would be hard-pressed to match it...
There used to be a handful of companies that produced IC opamps that really were optimized for audio. They're mostly gone now, and while what is left has improved along with the improvement in semiconductor manufacturing processes, there's not much choice any more. So if you want to do something _different_ in your circuit design (like Class A), you go discrete.
There's a rage right now in rechipping older opamp-based gear with the newer, faster, quieter parts that are currently available. And that can help a lot! But even so, the best of the opamp-based mic pres really can't compete with the best of the discrete designs for input noise, distortion, and bandwidth.
Mic pres are the place where this is the most start, since it is critical that you have very high gain with an absolute minimum contribution of noise, nonlinearity, or phase/bandwidth problems. Especially noise. The difference between a good discrete design and a run-of-the-mill opamp design is stark. It is real, it can be heard, it can be measured, and it can be understood, even by a non-nerd! It's not magic, and it most definitely isn't the voices of the angels.
Do you want your whole mixer to be discrete? Hell, no- you couldn't afford it, and it'd me a maintenance nightmare... But for truly critical applications like specialized high-end mic pres, discrete designs can offer some advantages- particularly in the noise arena. You put the bucks where the bang really needs to be.
Integrated circuit opamps have a few advantages, and a few disadvantages. The primary advantage is cost: they put most of the circuitry needed to implement your preamp into one single little black plastic package with a lot of legs. The primary disadvantage is they they are designed to be used in many, many different applications (from instrumentation through industrial controls to whatever else), and audio is only one of those- and a very small one at that.
So they aren't optimized for audio. They aren't optimized for much of anything, in fact. But they are as convenient as hell to design with, and cheap- so you can get wonderfully adequate products, and even some stellar ones.
Now with discrete components, you can use different circuit structures (like Class A amplifiers for example), and you can optimize every component in the chain for that particular audio application. You don't have to band-aid around the fact that the circuitry was really designed for something else, and you are just borrowing it.... In the case of mic preamps in particular, you can design front end stages that have orders of magnitude lower noise than you can ever hope to achieve with an integrated opamp. Similarly, you can optimize for distortion, or for bandwidth- and you can implement an output stage that doesn't require compromises on your *input* stage.
Ever notice how some low-end IC preamps get noisier with you drive a heavy load than a light (bridging) one? That's because the heavily loaded output stage heats up the chip, which changes the bias current on the input stage- they are thermally coupled, because they live on the same piece of silicon... The designer sand giveth, and the designer sand taketh away.
With opamp-based design, you are really constrained in what you can do. With discrete design, you can do much more, since every component is under your control- and sometimes that can sound better. In the hands of a good design engineer, it can sound _magnificent_, and the opamp-based stuff would be hard-pressed to match it...
There used to be a handful of companies that produced IC opamps that really were optimized for audio. They're mostly gone now, and while what is left has improved along with the improvement in semiconductor manufacturing processes, there's not much choice any more. So if you want to do something _different_ in your circuit design (like Class A), you go discrete.
There's a rage right now in rechipping older opamp-based gear with the newer, faster, quieter parts that are currently available. And that can help a lot! But even so, the best of the opamp-based mic pres really can't compete with the best of the discrete designs for input noise, distortion, and bandwidth.
Mic pres are the place where this is the most start, since it is critical that you have very high gain with an absolute minimum contribution of noise, nonlinearity, or phase/bandwidth problems. Especially noise. The difference between a good discrete design and a run-of-the-mill opamp design is stark. It is real, it can be heard, it can be measured, and it can be understood, even by a non-nerd! It's not magic, and it most definitely isn't the voices of the angels.
Do you want your whole mixer to be discrete? Hell, no- you couldn't afford it, and it'd me a maintenance nightmare... But for truly critical applications like specialized high-end mic pres, discrete designs can offer some advantages- particularly in the noise arena. You put the bucks where the bang really needs to be.
J
Jack Hammer
New member
Excellent Reply
Skipy, if your not an electronic engineer, you should be, or a teacher, or both. Nice job. Bravo
Skipy, if your not an electronic engineer, you should be, or a teacher, or both. Nice job. Bravo
Dom Franco
New member
Electronics 201
Discrete is better for all of the above reasons, (Optimal Circuit design, Heat dissipation, and other sometimes nostalgic reasons).
But also for practical operational reasons, they can be tweeked, modified, repaired, biased etc. Whereas a chip is a throw away item!
Discrete circuits and components cost more, are more labor intensive (Soldering and wire bending) but the main reason is:
They sound better! I think?
But, Most of us can't afford an original hand wired guitar amp, with choice selected caps resitors and tubes, calibrated and biased by an audiophile electronics technician.
So many of us settle for an amplifer based on an audio chip that costs $1.49 to manfacture and ship from China and expect it to sound great because it's advertised as "TUBE TONE".... Do the originals really sound better?
Don't get me wrong, some of the new modeling amps sound awesome, but I can't afford to do a side to side comparison with an old 50's Fender deluxe amp or a 60's Vox combo.
So I'll just try to be happy searching ebay for the find of the century. A GREAT OLD FENDER TUBE AMP IN MINT CONDITION PLAYED ONLY AT CHURCH ON SUNDAYS BY A LITTLE OLD GRAMPA FOR $65.00 No reserve!
Dom
Discrete is better for all of the above reasons, (Optimal Circuit design, Heat dissipation, and other sometimes nostalgic reasons).
But also for practical operational reasons, they can be tweeked, modified, repaired, biased etc. Whereas a chip is a throw away item!
Discrete circuits and components cost more, are more labor intensive (Soldering and wire bending) but the main reason is:
They sound better! I think?
But, Most of us can't afford an original hand wired guitar amp, with choice selected caps resitors and tubes, calibrated and biased by an audiophile electronics technician.
So many of us settle for an amplifer based on an audio chip that costs $1.49 to manfacture and ship from China and expect it to sound great because it's advertised as "TUBE TONE".... Do the originals really sound better?
Don't get me wrong, some of the new modeling amps sound awesome, but I can't afford to do a side to side comparison with an old 50's Fender deluxe amp or a 60's Vox combo.
So I'll just try to be happy searching ebay for the find of the century. A GREAT OLD FENDER TUBE AMP IN MINT CONDITION PLAYED ONLY AT CHURCH ON SUNDAYS BY A LITTLE OLD GRAMPA FOR $65.00 No reserve!Dom
wes480
New member
I am confused on one thing....
for some reason whenever I read "Class A Discrete" I always think "tube".
Is that wrong? I was thinking anything that had Class A/B or whatever...was automatically tube. But...hrmm...I dunno. help me out guys.
I don't know much on the EE side of things. Still trying to digest the discrete nonsense.
for some reason whenever I read "Class A Discrete" I always think "tube".
Is that wrong? I was thinking anything that had Class A/B or whatever...was automatically tube. But...hrmm...I dunno. help me out guys.
I don't know much on the EE side of things. Still trying to digest the discrete nonsense.