Very good. Thank you. Now then this leads into what JC has been talking about regarding higher amplitude of odds.. esp 7th etc. And related OL BW etc. So, this needs to be looked at carefully and contouring the OLG et al carefully to avoid increasing the more annoying odds. And, I was going to ask about compensation's contribution but you answered it already.
Is this the basis of the general comments that low OL BW circuits could produce the annoying higher order harmonics within the mid-upper audio freq range?
THx-RNMarsh
Is this the basis of the general comments that low OL BW circuits could produce the annoying higher order harmonics within the mid-upper audio freq range?
THx-RNMarsh
Last edited:
Huhh? How does this work then? Isn't the harmonic content primarily depending on the actual transfer function, with the loop gain only determining how much it will be suppressed when you close the loop?
If you have a linear circuit, there will be no 7th (or any other), whatever the loop gain contour.
Jan
Last edited:
Nothing, this line of reasoning has no merit what so ever. I share PMA's frustration, I have demonstrated at least twice that the pole in the DC open-loop gain has nothing to do with anything important. In fact you can take a tiny amount of positive feedback to the input offset of just about any amplifier and give it a 10Hz open-loop BW INCLUDING the Otala amp and have NO effect on the performance.
Interesting Keantoken!
Richard Marsh, thanks for your keen eye in noting that PMA's example uses a 10K load.
PMA while I respect your engineering and measurement capabilities, I am not using the AD825 just because it is a 'fashion' device. It has several features, and several drawbacks.
Let me point them out:
Features:
Good solid IC topology, like the AD797.
High open loop bandwidth (can't hurt)
Came well recommended by Scott Wurcer and Walt Jung.
Tried in a new prototype power with great sonic success.
High slew rate, (better than 100V/us that I like, even in preamps).
Drawbacks: Very noisy, this makes it useless for low level applications.
Limited output current (typical of many other IC's)
Limited voltage (I usually use +/- 28V) Here I can only use +/- 15V
Difficult to use package. (for me)
So what is your problem PMA? These two products that I tried the AD825 in are not in production, and there is no advertisement associated with them, so how can I use their inclusion as a fashion statement?
Be competitive, not slanderous, PMA.
Richard Marsh, thanks for your keen eye in noting that PMA's example uses a 10K load.
PMA while I respect your engineering and measurement capabilities, I am not using the AD825 just because it is a 'fashion' device. It has several features, and several drawbacks.
Let me point them out:
Features:
Good solid IC topology, like the AD797.
High open loop bandwidth (can't hurt)
Came well recommended by Scott Wurcer and Walt Jung.
Tried in a new prototype power with great sonic success.
High slew rate, (better than 100V/us that I like, even in preamps).
Drawbacks: Very noisy, this makes it useless for low level applications.
Limited output current (typical of many other IC's)
Limited voltage (I usually use +/- 28V) Here I can only use +/- 15V
Difficult to use package. (for me)
So what is your problem PMA? These two products that I tried the AD825 in are not in production, and there is no advertisement associated with them, so how can I use their inclusion as a fashion statement?
Be competitive, not slanderous, PMA.
Last edited:
No. Not good. Tales of mystery & imagination, to put it friendly.
The distortion at the output of an opamp depends only on the
nonlinearities inside and the available loop gain to correct for
them. Especially there is no such thing as a high pass filter effect.
All that happens is that you lose the ability to correct with rising
frequency because the loop gain usually drops with 20 dB/decade.
But for the quality of correction, all that counts is the absolute value
of the loop gain.
An opamp with, say, 140 dB and a 10 Hz OL corner will deliver the
same results as one with a 1KHz OL corner and 100 dB open loop gain.
Above 1 KHz. Below 1 KHz ist will be better and it will have the
better PSSR at 60 Hz. (All else being equal).
Get a text book or read the section in D.Self, he has elaborated
that nicely.
And I don't believe in those wonderful 7th. harmonics that are
needed to make it sound ooooooh so good. Every distortion
is bad. I leave the sound design to the musicians.
Gerhard
The distortion at the output of an opamp depends only on the
nonlinearities inside and the available loop gain to correct for
them. Especially there is no such thing as a high pass filter effect.
All that happens is that you lose the ability to correct with rising
frequency because the loop gain usually drops with 20 dB/decade.
But for the quality of correction, all that counts is the absolute value
of the loop gain.
An opamp with, say, 140 dB and a 10 Hz OL corner will deliver the
same results as one with a 1KHz OL corner and 100 dB open loop gain.
Above 1 KHz. Below 1 KHz ist will be better and it will have the
better PSSR at 60 Hz. (All else being equal).
Get a text book or read the section in D.Self, he has elaborated
that nicely.
And I don't believe in those wonderful 7th. harmonics that are
needed to make it sound ooooooh so good. Every distortion
is bad. I leave the sound design to the musicians.
Gerhard
Now you know why I changed my sig, it's time for some folks to accept input without taking it as a personal attack, at least SOME new input.
Well as there has been a decided lack of interest in the opamp listening test I will, unless there is a sudden rush, probably reveal the devices later tomorrow.
Electroj, were you able to get the file from the Onedrive link ?
Electroj, were you able to get the file from the Onedrive link ?
I think John's tongue in cheek response says it all, he is playing this and laughing all the way.
OK, just to be completely serious:
I design circuits with high open loop bandwidth, when I can, because experience with the listening reactions of others tells me that this is a good way to go.
With today's IC's, some have high open loop bandwidth, (for some reason not seen on the simplified schematic) and that makes them more interesting in that way. What else do we get? Well, all else being equal, we get higher slew rate. The only real test is to listen, or put them into a given design, and note the results.
If you don't believe in audio differences at the raw casual listening level, then use what measures best and let it go.
I design circuits with high open loop bandwidth, when I can, because experience with the listening reactions of others tells me that this is a good way to go.
With today's IC's, some have high open loop bandwidth, (for some reason not seen on the simplified schematic) and that makes them more interesting in that way. What else do we get? Well, all else being equal, we get higher slew rate. The only real test is to listen, or put them into a given design, and note the results.
If you don't believe in audio differences at the raw casual listening level, then use what measures best and let it go.
- Status
- Not open for further replies.