"IC op-amps have a nasty gritty distortion that once you have heard and identified you don't want to go back to. The buggers can even be heard in the audio path when used for DC servos. I found a way around that too but that information is for sale only."
Seriously ?
I have never heard a IC Op-Amp giving out any sort of nasty gritty distortion . Maybe that's because i have been developing high-end products , but no , i use common parts like NE5532 and TL072 , TL081 , LM4558 .
They do not give nasty distortion if you design them not to hang on the supply ,
that is a bad thing and usually misdesigned .
Discrete op amps can be a bad thing when speed and efficiency is required badly , because the more capacitance is present in the leads are present , the worser the ability to speed .
Well all of us here know it . Depending on the parts used , JFET input opamps sound good .
But , discrete op amps can be really expensive quick compared to IC , even though they only have so few parts because IC's are just cheap to get around with .
Have any tried to design a OPA2604-like Discrete Op Amp ? Care to give on the figures , cost , BOM ?
Seriously ?
I have never heard a IC Op-Amp giving out any sort of nasty gritty distortion . Maybe that's because i have been developing high-end products , but no , i use common parts like NE5532 and TL072 , TL081 , LM4558 .
They do not give nasty distortion if you design them not to hang on the supply ,
that is a bad thing and usually misdesigned .
Discrete op amps can be a bad thing when speed and efficiency is required badly , because the more capacitance is present in the leads are present , the worser the ability to speed .
Well all of us here know it . Depending on the parts used , JFET input opamps sound good .
But , discrete op amps can be really expensive quick compared to IC , even though they only have so few parts because IC's are just cheap to get around with .
Have any tried to design a OPA2604-like Discrete Op Amp ? Care to give on the figures , cost , BOM ?
Agree 100%.
There are only a few things that sounds worse than opamp @unity gain.
then I suppose you really don't want anyone looking at my http://www.diyaudio.com/forums/solid-state/45794-high-loop-gain-composite-op-amp-circuits.html
or to look up Jung's substrate bootstrap, or Dimitiri's
Supply Bootstrapping Reduces Distortion In Op-Amp Circuits | New operational amplifiers optimized for high-performance audio and ultrasound applications combine extremely low total harmonic distortion plus noise (THD+N), -130 dB, with large output vo
"IC op-amps have a nasty gritty distortion that once you have heard and identified you don't want to go back to. The buggers can even be heard in the audio path when used for DC servos. I found a way around that too but that information is for sale only."
I can absolutely agree that this may be a problem on an opamp audio circuit IF you do just one of the things below
1. run a non-unity gain stable opamp at unity gain
2. run a decompensated opamp without a comp cap, or too small a comp cap for the gain setting
3. fail to fit an isolation resistor in series with the output when driving a real world load - like a cable for example, or a capacitive load
4. fail to ensure that the junction of the feedback network is located physically very close to the op-amp feedback input pin (usually the inverting input)
5. fail to locate the input filter as physically close as possible to the op-amp input (usually the non-inverting input)
6. fail to decouple the supply rails adequately
7. you use an opamp not characterized for wide band audio usage that has an inadequate slew rate
8. etc
I've done all of these things, and yes, if just one of them occurs, the sound can be truly awful. OTOH, if you follow the engineering rules and guidelines associated with high performance opamps, you will get great performance and great sound.
That said, probably 99% of the CD and SACD players on the planet use opamps of some description in their signal chains (whether stand alone or integrated into the converter) . . . but I seriously doubt that 99% of them sound crap. Keep in mind that almost all music is mixed on desks using op-amps, passed through filters and what have you. Douglas Self mentions '100's' in a typical signal chain. Again, if opamps were as bad as some claim, there would be no good music available. Just about everything would sound bleached, gritty and truly terrible. But that's not the case, is it?
As for affecting the sound when used in servo applications, well, that's nonsense. Read Robert Cordell's book for a bit of enlightenment on that subject.
I don't suspect I will be able to convince opamp detractors that they are by and large wonderful devices for audio, but that's ok. Am I against discrete designs? No, they are great, but so are opamps.
I can absolutely agree that this may be a problem on an opamp audio circuit IF you do just one of the things below
1. run a non-unity gain stable opamp at unity gain
2. run a decompensated opamp without a comp cap, or too small a comp cap for the gain setting
3. fail to fit an isolation resistor in series with the output when driving a real world load - like a cable for example, or a capacitive load
4. fail to ensure that the junction of the feedback network is located physically very close to the op-amp feedback input pin (usually the inverting input)
5. fail to locate the input filter as physically close as possible to the op-amp input (usually the non-inverting input)
6. fail to decouple the supply rails adequately
7. you use an opamp not characterized for wide band audio usage that has an inadequate slew rate
8. etc
I've done all of these things, and yes, if just one of them occurs, the sound can be truly awful. OTOH, if you follow the engineering rules and guidelines associated with high performance opamps, you will get great performance and great sound.
That said, probably 99% of the CD and SACD players on the planet use opamps of some description in their signal chains (whether stand alone or integrated into the converter) . . . but I seriously doubt that 99% of them sound crap. Keep in mind that almost all music is mixed on desks using op-amps, passed through filters and what have you. Douglas Self mentions '100's' in a typical signal chain. Again, if opamps were as bad as some claim, there would be no good music available. Just about everything would sound bleached, gritty and truly terrible. But that's not the case, is it?
As for affecting the sound when used in servo applications, well, that's nonsense. Read Robert Cordell's book for a bit of enlightenment on that subject.
I don't suspect I will be able to convince opamp detractors that they are by and large wonderful devices for audio, but that's ok. Am I against discrete designs? No, they are great, but so are opamps.
Last edited:
I do insist. In one thing that i have seen with my eyes in oscilloscope. I don't know exactly the reason, but discrete are perfect in low-mid frequencies. Inject a 50Hz square wave in input and you will see exactly the same square in the output. I don't know if there is any IC that could obtain this.
So is the SSM (I think it's about 1.4nV/rtHz). What kind of mike preamp are you working on? I was knocked out by how well the SSM worked for a ribbon.
Scott Wurcer was nice enough to send me an extra pair which are currently in a prototype MM preamp.
With a 50 Hz square wave input? Quite hard to believe, unless you are trying to squeeze out of a poor ol'5532 a 10MHz BW with 40dB gain...
L.
Nearly all of them can, it's trivially easy. You might revisit your test setup.
When I tried Jung's nested feedback circuits I found they did not really sound any different to most op-amp circuits. The power supply and decoupling used has a massively larger effect on the sound than tweaking the figures with some clever feedback schemes IMO.
I don't have any problem with high feedback circuits, in fact the JE990 is a high feedback circuit. It's optimised for high gain and low noise with low input impedances, but it sounds very good at modest gain and with a 20K pot stuck in front of it.
I've built 2 or 3 highly regarded valve preamps. Although they give very impressive sound on first hearing them, somehow this becomes wearing on longer term listening. It's a mystery to me why one in particular (Rosenblit Grounded Grid), sounds like this when it's measurements are pretty much textbook.
The resistors degenerate the input pair and the inductors actually "un-degenerate" the LTP at lf. I think this is mainly done to keep the nf low. No I have not tried removing them or any other types, although just like you , the thought had crossed my mind.
One opamp type you should be carefull about is the RR input type. Analog devices has addressed this problem. Some RR input types has a crossoverrigion where the input biascurrent peaks or shifts rapidly..
Read the datasheet. If there is no graph in the datasheet of RR input types, showing input biascurrent versus input voltage type, then i would say they are no-go.
Just a little advice. Some types are very sensitive to input impedance, but so are all amps, and some more than others.
Read the datasheet. If there is no graph in the datasheet of RR input types, showing input biascurrent versus input voltage type, then i would say they are no-go.
Just a little advice. Some types are very sensitive to input impedance, but so are all amps, and some more than others.
Jung's toying with Otala inspired inner loop feedback is ill advised in my opinion (really only in one instance - he has generally advocated several "multiloop"/composite amplifier topologies)
his use of gain in the output op amp of a multiloop/composite amp allows sub regulating the input op amp supplies - since the input op amp has little dynamic current demand even simple Zener shunt gives good regulation
the psrr is then boosted by both the excess loop gain and the subregulation
to address common mode input distortions I was pointing to the Sandman "feedforwad" as one example of power supply bootstrapping
the feedforward bootstrap can improve cm distortion if the bootstrapping amp is good - say a cascoded fet - and the main amp adds gain, possibly better noise (Funasaka used a discrete fet ps buffer amp) AES E-Library Feed Forward Floating Power Supply (High Response-Speed Equalizer Circuit)
Dimitri's floating ps bootstrap circuit does even better in that the composite amp input sees only the bootstrapped U1 input - effectively "cascoding" the entire input op amp
Attachments
Last edited:
Well. You will be surprised how many good lownoise and high speed RR I/O there is on the market.
I use the AD8692 from analog, it is by far not the best opamp in the world but it is very decent and cheap.
Look at the specs and tell me if you can do it better from a +5V supply only.
G = 1, RL = 600 Ω, f = 1 kHz, VO = 1 V p-p 0.0006% - This is at +5V supply and the distortion rises to only 0.002% at 6KHz.
at RL = 1KOhm it is below 0.0008% at 6KHz.
at RL = 100KOhm it is below 0.0004% at 6KHz.
Voltage noise is 8nV/SQ(Hz) at 1KHz.
Current noise is 50fA/SQ(Hz) at 1KHz. A good candidate for MM pickup input.
Match it with a good powersupply decoupling with low noise and ripple and you can have a good and cheap MM RIAA amp.
AD8692 | Dual, Low Cost, Low Noise, CMOS Rail-to-Rail Output Operational Amplifier | Operational Amplifiers (Op Amps) | All Operational Amplifiers | Analog Devices
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.