Hi everyone!
I am wondering if there are certain parameters that will make popular opamps like NE5532, LM4562, TL072, OPA2134, and others, to give the best performance as far as THD and sound quality?
When I studied datasheets for those opamps it seems that the lowest THD is achieved when R load is 2k.
Another parameter that I noticed was the voltage. For example, OPA2134 gives better THD performance when powered with dual 18V power supply.
One more parameter could be the gain. For example G=1, 2, ..., 100
The question that I have is, on individual basis, what would be best circuits that will allow those opamps to give the best sound they can produce?
Any comments are appreciated!
Thanks,
Pavlo
I am wondering if there are certain parameters that will make popular opamps like NE5532, LM4562, TL072, OPA2134, and others, to give the best performance as far as THD and sound quality?
When I studied datasheets for those opamps it seems that the lowest THD is achieved when R load is 2k.
Another parameter that I noticed was the voltage. For example, OPA2134 gives better THD performance when powered with dual 18V power supply.
One more parameter could be the gain. For example G=1, 2, ..., 100
The question that I have is, on individual basis, what would be best circuits that will allow those opamps to give the best sound they can produce?
Any comments are appreciated!
Thanks,
Pavlo
Your question is impossible to answer because you don't specify what you mean by 'best sound'.
You mention THD, are you looking for lowest THD?
There are general factors, like lower load impedance generally increases THD, higher gain increases THD, higher signal levels will increase THD, lower supply with a certain signal level will increase THD.
Some opamps have lower THD than others, all else being the same.
Jan
You mention THD, are you looking for lowest THD?
There are general factors, like lower load impedance generally increases THD, higher gain increases THD, higher signal levels will increase THD, lower supply with a certain signal level will increase THD.
Some opamps have lower THD than others, all else being the same.
Jan
In my experience for the most subjectively pleasing sound out of an opamp, it needs the lowest noise power rail. Opamps don't in general draw constant current vs signal so to keep power rail noise as low as possible, arrange for the lowest impedance power and load them as lightly as practically possible.
I would not equal best sound with lowest THD.
THD is just one of the metrics that are used to judge quality.
Personally I find running op-amps in class AB a bad idea since this inevitably creates crossover distortion and it is entirely up to GNFB to reduce this ugly kind of distortion, and GNFB becomes less effective in the higher frequency range.
Thus, running the op-amp in class A seems to be a really good idea.
THD is just one of the metrics that are used to judge quality.
Personally I find running op-amps in class AB a bad idea since this inevitably creates crossover distortion and it is entirely up to GNFB to reduce this ugly kind of distortion, and GNFB becomes less effective in the higher frequency range.
Thus, running the op-amp in class A seems to be a really good idea.
Agree. I design discrete op-amp with single-ended class A output. Slew rate is decent and it compensated at gain 1x. A friend made pcb and sell it online. The design is free. All person who used it pleased with the sound.
I already made simulation op-amp with buffer single-ended class A. I use 2 or 3 parallel of BC550C.
So far out of all four opamps mentioned in the first post, my favorite is OPA2134. I like the way it sounds the most. NE5532 sounded the muddy, but I wonder if there is a way to make it sound nice?
Maybe this will be stupid question, but is there a way to get NE5532 into class A operation trying to improve on sound?
Maybe this will be stupid question, but is there a way to get NE5532 into class A operation trying to improve on sound?
I tried only with OPA627, but that has different output than the NE5534. I found the following for you:
https://www.diyaudio.com/community/threads/class-a-biasing-of-opamps.6983/
https://tangentsoft.net/audio/opamp-bias.html
So for example, will the NE5532 give its best performance / best sound when loaded with 2k load, powered with +-17V, and with gain = 1 or 2 ??
What about input signal voltage or current output?
What about input signal voltage or current output?
Found cool thread on similar topic - https://www.diyaudio.com/community/...st-possible-thd-n-really-the-best-way.367692/
I like have certain harmonic profile distortion like class A output first than make it THD lowest as possible. THD must low in all audio frequency and all level before clipping.
You could experiment with a classA biassed output stage like this. I've not tried this with NE5532 but have with various other opamps (AD744, LT1028, OPA2209) :
As Jan points out above opamps will have lower THD for lower signal swing. That's because the amp is operating away from the power supply rails. This also means that you tend to get lower THD if you run the opamp at the highest supply voltage it is specified for. That's especially true for the older opamps. Modern parts tend to care less.
Lighter load (higher resistance) -> lower THD.
Some opamps have lower THD when run as inverting amps as the common-mode voltage is zero (or very close to) for that configuration.
Tom
Lighter load (higher resistance) -> lower THD.
Some opamps have lower THD when run as inverting amps as the common-mode voltage is zero (or very close to) for that configuration.
Tom
Lower signal voltages and lower output current (= higher load resistance) gives lower THD.
There is nothing magically or optimum about 2k load; lower load gives lower distortion.
Higher supply gives lower load.
Jan
Found a few interesting opamps on TI website:
OPA1612 - bipolar input transistors, 0.000015% THD, even lower than LM4562
OPA1656 - FET input transistors, 0.00003 THD, Voffset = 1mV - I'm thinking to use this one for servo.
OPA1612 - bipolar input transistors, 0.000015% THD, even lower than LM4562
OPA1656 - FET input transistors, 0.00003 THD, Voffset = 1mV - I'm thinking to use this one for servo.
A very long time ago when 5532 was about as good as they made, Walt Jung wrote a book about using op-amps for audio. Y-all should have a copy? I assume newer parts make most of that history, but a few ideas are:
1. An inverting amp avoids input CMRR distortion because the inputs are fixed at ~ground voltage. However, an inverting amp is slightly noisier than a non-inverting amp, and the input impedance is relatively limited.
2. A class AB output can be operated in class-A mode by adding an external DC bias, ie a resistor to one of the supply rails, probably negative. This has to be less than half the output current limit and as an extra load, it compromises load induced performance loss.
3. Lower value feedback and bias resistors generate less noise, but again, extra output loading compromises load induced performance loss. High value feedback resistors require a high frequency bypass cap to avoid stability issues from stray capacitance. Low value resistors require larger coupling capacitors which are generally inferior to smaller capacitors.
4. Local supply decoupling is important for many op-amps since their PSRR is not great at high frequencies.
5. High supply voltages means the signal is well away from the rails, however a limited slew rate is a problem for larger signal voltage, not so much for small signals.
6. Externally compensated op-amps are hard to find today, or de-compensated. You should understand why such have a minimum gain use. This still applies to most chip power amps.
7. An open input is noisy for a non-inverting circuit but open makes an inverting circuit go quiet. A non-inverting input is more vulnerable to RF rectification and requires an RF filter.
nuff 4 now
1. An inverting amp avoids input CMRR distortion because the inputs are fixed at ~ground voltage. However, an inverting amp is slightly noisier than a non-inverting amp, and the input impedance is relatively limited.
2. A class AB output can be operated in class-A mode by adding an external DC bias, ie a resistor to one of the supply rails, probably negative. This has to be less than half the output current limit and as an extra load, it compromises load induced performance loss.
3. Lower value feedback and bias resistors generate less noise, but again, extra output loading compromises load induced performance loss. High value feedback resistors require a high frequency bypass cap to avoid stability issues from stray capacitance. Low value resistors require larger coupling capacitors which are generally inferior to smaller capacitors.
4. Local supply decoupling is important for many op-amps since their PSRR is not great at high frequencies.
5. High supply voltages means the signal is well away from the rails, however a limited slew rate is a problem for larger signal voltage, not so much for small signals.
6. Externally compensated op-amps are hard to find today, or de-compensated. You should understand why such have a minimum gain use. This still applies to most chip power amps.
7. An open input is noisy for a non-inverting circuit but open makes an inverting circuit go quiet. A non-inverting input is more vulnerable to RF rectification and requires an RF filter.
nuff 4 now