"It is very well known that increasing the reservoir caps will increase both the level and order of the harmonics on the supply line. That gives a lot of scope to have an audible effect, depending on several factors. Nothing 'magical' here..."
I think this related to the charging currents in the reservoir caps. Localized decoupling should reduce the HF content on the rails due to signal current.
I think this related to the charging currents in the reservoir caps. Localized decoupling should reduce the HF content on the rails due to signal current.
But why, other than the fact that "they" use that for a test? If there's a reason to do it, there's plenty of op-amps out there that will, but the reason to do that is not clear given the likely applications.
I am concerned about cable capacitance and input filters of the receiving equipment. I just could not bring myself to design a pre that had K ohms of OP Z.
A 'bad cable may have a few hundred pF of capacitance, an input BW filter 270 Ohms and 1 nF (ok, this is quite stiff, but not uncommon). Feeding this from a high Z?
But, each to his own.
A 'bad cable may have a few hundred pF of capacitance, an input BW filter 270 Ohms and 1 nF (ok, this is quite stiff, but not uncommon). Feeding this from a high Z?
But, each to his own.
Add the light loading buffer OPS on an IC amp any way you wish -- then you will get less of the distortion from loading and more of the best potential from IC amps.
[And, If a low noise design, the fb resistors will be low values -- more reason why a stock IC amp's distortion will be higher. So, it is obvious that the external buffer OPS is useful/needed for lowest thd.]
Re large reservior caps.... generally not needed in low/line level circuits. A good regulator or source follower (C-Multiplier) type will take care of ripple. But with regulators, minimal output C seems to work best.
THx-RNMarsh
Last edited:
I once bought a good tube preamp (Audio Research) and I saw it had several K Ohms in series with output jack. I jumpered it out for just the reason you state. Months later the HV supply crapped out.... couldnt take the loading current, I suspect. I never bothered to fix it and gave it to Kavi who would have it fixed. No high output Z, pls!
In addition, a high source Z contributes to EMI/RFI pickup on cables/grounds.
THx-RNMarsh
Last edited:
This is not about the opamp output Z, but the receiver input Z. By default opamp output Z is almost always very low, and often determined by the build-out series R to ensure stability with cap loads. The Zout is almost always < 100 ohms.
A 5m interlink with 500pF/m represents an impedance of some 3.3 k ohms at 20kHz. Piece of cake.
I agree with John that often the feedback network represent a significant part or even most of the opamp load impedance.
But observe: in line amp applications, there's really no reason to go so low with the feedback R, noise is not an issue here.
It is, of course, in say phono pre's. But the noise-critical first stage handles at most a few 10's of mV signals, and even with a 100 ohm feedback R that doesn't really tax the output stage that much - 100mV across 100 ohms is just 1 mA....
So, another dozen of pages of posts on non-issues
We're making progress???
Jan
It is, Jan. You will add 20 - 30dB gain of the power amplifier behind the (noisy high R feedback) opamp pre output and you will seriously degrade power amps S/N, in case you have a very good power amplifier. Believe me, it makes a difference, and it makes a sound difference.
Oh no, but which opamp would work well to 100R + 100R network, except for AD811 or ADA4898 etc.?
"Standard" opamps like OPA134 will be happy with 10k + 10k and then you are in noise issues.
Usual definition: sum of the currents through the two OP transistors constant, and no one ever turning off.
John, I tried to make it clear that the values were not real, there just to illustrate the point.
Actual values are different, the trouble is I caouldn't find the actual files which isslustrated my actual attempts.
BTW, I thought about you yesterday. In one of your posts God knows how long ago, you mentioned in passing that some of the old op amps, specifically LF 356, was capable of solid results. Obviously, I agree, I used it as my general test bed, but I also did AD 829, AC 818 and even the old OPA 37. All did well, but the winner is - OPA 37. That op amp has its issues, but overall it can provide some surprisingly good results.
Nah, that's nothing, -120 dB?
You must be joking, Dado. That's wild man!
The only question is what does it do in the 30-80 Ohms region, where many headphones actually are?
- Status
- Not open for further replies.