Interesting read (short) on Op-Amps.
http://nwavguy.blogspot.com/2011/08/op-amps-myths-facts.html?m=1
http://nwavguy.blogspot.com/2011/08/op-amps-myths-facts.html?m=1
Op amp myths and........
- Thread starter WOPL Sniffer
- Start date
- Joined
- Jan 14, 2011
- Messages
- 74,873
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
maybe i can revisit that but he said the opa2134 will sound the same as an ne5532 and will I swapped an opa into an ne spot and the player sounds much better. Didn't care for it before, not it is fine. Some folks just don't hear the same. Or maybe it was my desire to reaffirm the expenditure of that 3 bucks. wtf why do folks insist on what others can hear. Those guys need glasses for their ears to hear the differences some hear.
- Joined
- Jan 14, 2011
- Messages
- 74,873
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
- Joined
- Jun 11, 2010
- Messages
- 6,849
- Location
- Powhatan, Virginia, United States
- Tagline
- WassupYa Mang?
I think the effectiveness of swapping one definitely does depend on the circuit overall. So I do agree to that fact, but I think what he's missing is that a better engineered op amp design has the capability of performing better, hence the better sound. In my experience it's been sometimes hit and miss, but when the circuit itself can support the new design it really does make a huge difference in noise, clarity and soundstage. The other obvious thing with vintage gear is one shouldn't just "stop" at the op amps, since chances are good that a recap along with a swap will be the best strategy.
So yeah, it's like rolling the dice sometimes (which corresponds to why I think "rolling op amps" is an appropriate phrase).
So yeah, it's like rolling the dice sometimes (which corresponds to why I think "rolling op amps" is an appropriate phrase).
Hmmm. Too many disagree. First, many high end designers use tricks like pulling the opamp's class B output stage down so it runs class A. Not every opamp can do this! In a sense, the designer is "hacking" the opamp to get a better sonic result.
I read a wonderful article authored by a tube amp designer where he discussed the feedback problem (s). His primary point was that designers with a huge amount of open loop gain are likely to be very lazy in circuit design because the feedback will fix it. Indeed the whole concept of the opamp is that they behave almost at the theoretical "ideal" opamp model that practical limitations (like the author cited an opamp headphone amp driving 60 ohm phones) get ignored. If we know one thing about audio, there is a difference between theory and the real thing.
Do all opamps sound alike? I can't comment. Do I think a 2 triode line stage preamp sounds better than an opamp preamp? Yes. Why is is better? Which is the simple design? Which would you think best fits "less is more"?
Did any of you ever own the Dynaco PAT5? I had a PAT4 (all discrete) in college, but switched to a PAT 5 (one opamp in line stage) when I scored a big used system. The PAT 5 was brighter, and I didn't like it. Since I had sold the PAT 4, I moved on to the APT Holman preamp. Now I had more opamps in the signal path, but the sound made me happy for many years. More is more could be the lesson here, or the opamps may have been better, OR, the designer used care in designing the whole amp and didn't fall into lazy traps. I doubt we could measure or analyze these 3 preamps enough to learn what we might about opamps vis a vis audio, but I bet a bunch of you know and experienced exactly what I am describing.
I read a wonderful article authored by a tube amp designer where he discussed the feedback problem (s). His primary point was that designers with a huge amount of open loop gain are likely to be very lazy in circuit design because the feedback will fix it. Indeed the whole concept of the opamp is that they behave almost at the theoretical "ideal" opamp model that practical limitations (like the author cited an opamp headphone amp driving 60 ohm phones) get ignored. If we know one thing about audio, there is a difference between theory and the real thing.
Do all opamps sound alike? I can't comment. Do I think a 2 triode line stage preamp sounds better than an opamp preamp? Yes. Why is is better? Which is the simple design? Which would you think best fits "less is more"?
Did any of you ever own the Dynaco PAT5? I had a PAT4 (all discrete) in college, but switched to a PAT 5 (one opamp in line stage) when I scored a big used system. The PAT 5 was brighter, and I didn't like it. Since I had sold the PAT 4, I moved on to the APT Holman preamp. Now I had more opamps in the signal path, but the sound made me happy for many years. More is more could be the lesson here, or the opamps may have been better, OR, the designer used care in designing the whole amp and didn't fall into lazy traps. I doubt we could measure or analyze these 3 preamps enough to learn what we might about opamps vis a vis audio, but I bet a bunch of you know and experienced exactly what I am describing.
If all opamps sounded the same I really don't think we'd have nearly as many.
Saying all opamps sound the same is akin to saying all amps sound the same. To be sure there are many that say this same thing.
Don't agree with much that guy has to say. My experience differs greatly from his reported myths.
Saying all opamps sound the same is akin to saying all amps sound the same. To be sure there are many that say this same thing.
Don't agree with much that guy has to say. My experience differs greatly from his reported myths.
Opamps are environmentally sensitive to the surrounding circuit, and it is easy to cite one situation which creates audible differences in sound. Audio may be designed for limited bandwidth, say 1Hz to 20kHz, or it may be "wideband" DC to 1 MHz. Any opamp with limited gain bandwidth product, poor high frequency current limitations (slew rate), or time delay phase shifts is going to behave and sound different in the wideband circuit.
looks to be the same as a ne5533, which has slightly better specs.
http://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=ne5533#linkId=8&src=top
Interesting read. I agree with some of it, not all.
http://www.ti.com/sitesearch/docs/universalsearch.tsp?searchTerm=ne5533#linkId=8&src=top
Interesting read. I agree with some of it, not all.
- Joined
- Jan 14, 2011
- Messages
- 74,873
- Location
- Gillette, Wyo.
- Tagline
- Halfbiass...Electron Herder and Backass Woof
Sure Lee.
Besides the competing brand differences, there are so many op amps because there are so many differing applications. There is no perfect op amp (although that ideal is being approached more and more). Since there is no perfect op amp, manufacurers optimize one or several functional parameters in each op amp type to suit a particular application. Those could be low power, low offset voltage, low bias current, rail to rail input, rail to rail output, low power supply voltage, single power rail voltage, low noise, high open loop gain, gain-bandwidth product, slew rate, output drive capability, unity gain stability and so on. You either pick the op amp type to suit the current circuit design topology or design the circuit around a particular op amp.
The NE5533/34 for example cannot be used in applications with a closed loop gain less than three so cannot be used in a unity gain buffer application. The OPA2134 can be as can many other op amp types.
Choice is good.
Besides the competing brand differences, there are so many op amps because there are so many differing applications. There is no perfect op amp (although that ideal is being approached more and more). Since there is no perfect op amp, manufacurers optimize one or several functional parameters in each op amp type to suit a particular application. Those could be low power, low offset voltage, low bias current, rail to rail input, rail to rail output, low power supply voltage, single power rail voltage, low noise, high open loop gain, gain-bandwidth product, slew rate, output drive capability, unity gain stability and so on. You either pick the op amp type to suit the current circuit design topology or design the circuit around a particular op amp.
The NE5533/34 for example cannot be used in applications with a closed loop gain less than three so cannot be used in a unity gain buffer application. The OPA2134 can be as can many other op amp types.
Choice is good.
No indeed. The LM747 suggested replacement is a horrible op amp designed in the early days of op amps. Op amps have come a long way since the LM741/747 and RC4136. Although many of these got used in early audio designs, hardly any of those designs could be called good by modern standards. The early Crown IC150 pre is a great looking and a very functional pre but the op amp choices available at the time made it sound like fingernails on a chalk board.
By the way, the original article was well thought out and written.
Another note, Burr-Brown (now owned by TI) and Analog Devices are generally accepted as the king of the hill manufacturers in op amps highly targeted for audio applications. They designed not only for specs but also for sound quality. As we all know, pure specs and SQ do not always go hand in hand.
The venerable Burr-Brown OPA627 is highly regarded in many audio circles but has some downsides that have to be considered in circuit design to apply it.
Another note, Burr-Brown (now owned by TI) and Analog Devices are generally accepted as the king of the hill manufacturers in op amps highly targeted for audio applications. They designed not only for specs but also for sound quality. As we all know, pure specs and SQ do not always go hand in hand.
The venerable Burr-Brown OPA627 is highly regarded in many audio circles but has some downsides that have to be considered in circuit design to apply it.
- Joined
- May 14, 2014
- Messages
- 2,321
- Location
- Southwest Kootenays BC
- Tagline
- No such things as bad days, just bad moments
Audio Control used the 4136 extensively in their early car audio products in the mid 80's and after. Their products were not what I would call "inexpensive", but they did develop a cult following with good marketing and engineering/technical support.
Bose uses 4558's in most of their consumer products because they say that this particular op-amp "voices" the product they way they like. Bose could afford to use other op-amps given the price commanded by their product.
I guess it's like tires... they all are round, made of rubber and roll, but there is a huge difference between various tires.
Bose uses 4558's in most of their consumer products because they say that this particular op-amp "voices" the product they way they like. Bose could afford to use other op-amps given the price commanded by their product.
I guess it's like tires... they all are round, made of rubber and roll, but there is a huge difference between various tires.
You can make an op-Amp from 5 transistors, but it would sound terrible!
So... simplicity is not the answer.
The 5 statements for the "Ideal" OP-AMP:
1. Infinite Input impedance (zero input current / voltage driven)
2. Infinite input to output Isolation
3. Infinite Gain.
4. Infinite Bandwidth
5. Zero Output Impedance.
OP-Amps today...very close, but not absolutely perfect. Does it need to be? No. Beyond a certain point... ya cant measure or hear it.
So... Best Op-AMP for audio:
1. FET input a must, with precision stable current sources on both rails for accuracy.
2. High inner stage class-A gain for highest Isolation, and Highest open loop "phase less" gain.
3. FET Output, for Lowest RDS sourcing resistance and max current capability.
4. Class A, no crossover distortions.
5. Lowest noise differential input.
6. Phase delay... minimum (a few degrees) at 100KHz gets you there.
Then, adding the gain resistors and high-quality capacitors that have no DC leakage characteristics or DC to AC bias distortion issues...and its noiseless and cleanest sounding.
BTW, the closer the Power amp meets this OP-AMP stuff, the better it will sound. All Power amps are just higher power versions of the same thing!
So... simplicity is not the answer.
The 5 statements for the "Ideal" OP-AMP:
1. Infinite Input impedance (zero input current / voltage driven)
2. Infinite input to output Isolation
3. Infinite Gain.
4. Infinite Bandwidth
5. Zero Output Impedance.
OP-Amps today...very close, but not absolutely perfect. Does it need to be? No. Beyond a certain point... ya cant measure or hear it.
So... Best Op-AMP for audio:
1. FET input a must, with precision stable current sources on both rails for accuracy.
2. High inner stage class-A gain for highest Isolation, and Highest open loop "phase less" gain.
3. FET Output, for Lowest RDS sourcing resistance and max current capability.
4. Class A, no crossover distortions.
5. Lowest noise differential input.
6. Phase delay... minimum (a few degrees) at 100KHz gets you there.
Then, adding the gain resistors and high-quality capacitors that have no DC leakage characteristics or DC to AC bias distortion issues...and its noiseless and cleanest sounding.
BTW, the closer the Power amp meets this OP-AMP stuff, the better it will sound. All Power amps are just higher power versions of the same thing!
