C5 can't charge "slowly" ... it has to equal "T" below.
t = -log((V-Vc)/V)R*C
C5 in the Sym is (V =12V supply) (Vc=6 volt at cap) R x C = that is only
54 milliseconds. Q5 only needs 10uA ! So 99.99% of the "action" is the R/C.
Calculating the 1Meg/1u or the 1Meg/.22u is more difficult as the initial
output voltage could be non repeatable .... or would respond to the servo's
correction as it is deviating from zero.
The servo R/C should be .6 sec , change of 1 volt at output.
Edit - things to try , different types of integrator caps and removing D7-8.
(reverse leakage in diodes , extreme low voltage ESR/leakage in caps)
OS
Last edited:
What type of capacitor is best for the integrator? I've been using Metallized Polyester. Is there a preferred composition for this type of circuit?
jwilhelm, I'm going to do a little experiment this evening when I'm home from work and see how low I can get the DC offset before inserting the servo and see. I suspect the design has just so much native DC offset that the servo just simply takes time to reign it in to an acceptable value. That's the nature of the servo, it's supposed to be slow and only correct for static DC offsets and to compensate for things like temperature, component drift and such.
That's my suspicion as well as any modification to the servo itself hasn't really had any effect on mine so far and it seems to increase with temperature. I haven't had the time to sit down and really try to figure anything out. I'd like to hook the scope on the output of the servo and watch it's operation as it's happening. If it's already at it's maximum output long before the voltage stabilizes you know it's swamped by something else.
It does in LTSpice, which is good enough for me; if you have a better suggestion op-amp wise, I'm all ears 🙂 Or better said, this opamp gave the best results THD wise of almost all I tried that came with LTspice and it fitted the description in the datasheet. My opinion is that a good opamp for audio definitely doesn't need to be marketed as such (audio opamp).
Last edited:
Well, that's a big assumption. But considering that LT spice comes with a large library of their own (Linear Technology) opamp models I'd guess they would put some accuracy in their models and make their sim results valid enough. Ofcourse this is in Sim-World (tm) and obviously not in reality which most likely will end up less than perfect.
This opamp just did best for what I needed it to do (drive a relatively low impedance of 560 ohm, configured as unity gain buffer with a 2Vpp signal max)
First, I wonder whether you know what 10ppb really means.
Second, THD figures derived from macro models, which is apparently the case here, are meaningless. Only THD figures derived from device-level models may be more or less accurate.
Second, THD figures derived from macro models, which is apparently the case here, are meaningless. Only THD figures derived from device-level models may be more or less accurate.
10 Parts Per Billion Sir 🙄 It's okay what you are doing though, not that I don't see it.
Sir Stuart,
It would seem somewhat ridiculous to think that anything short of a straight wire and even then I would question that with wire purity that you could ever in any instance reach 10ppb THD in an audio circuit. It just seems statistically improbable that this level of purity of a reproduced audio signal would have to disregard all the parasitic elements along any audio chain. My question would be what is the actual lowest level distortion anyone has actually measured in a truly active circuit and not a simulation? Even in the best Scott Worcer discrete opamp line drive what were the actual measured distortion numbers? Even the low ppm values often seen about power amplifier THD reaching into the single digit ppm range seem to be wishful thinking.
It would seem somewhat ridiculous to think that anything short of a straight wire and even then I would question that with wire purity that you could ever in any instance reach 10ppb THD in an audio circuit. It just seems statistically improbable that this level of purity of a reproduced audio signal would have to disregard all the parasitic elements along any audio chain. My question would be what is the actual lowest level distortion anyone has actually measured in a truly active circuit and not a simulation? Even in the best Scott Worcer discrete opamp line drive what were the actual measured distortion numbers? Even the low ppm values often seen about power amplifier THD reaching into the single digit ppm range seem to be wishful thinking.
Hi Kindhornman,
Halcro claims 0.5ppm for their dm88 amp and Nat Semi 0.3ppm for their LME49710 op-amp. I think this is about the bottom limit.
Cheers, E.
Halcro claims 0.5ppm for their dm88 amp and Nat Semi 0.3ppm for their LME49710 op-amp. I think this is about the bottom limit.
Cheers, E.
Thank you Edmond,
Even the Halcro values seem rather low for an amplifier of that power even baring the cost no object design. Not saying that it isn't a great sounding amplifier,I've never heard it. At some point it still seems silly as even most better than average loudspeakers can't even get close to less than 1% distortion on the best days.
Even the Halcro values seem rather low for an amplifier of that power even baring the cost no object design. Not saying that it isn't a great sounding amplifier,I've never heard it. At some point it still seems silly as even most better than average loudspeakers can't even get close to less than 1% distortion on the best days.
> it still seems silly......
Maybe, maybe not. Point is that the distortion from an amp is quite different than from a loudspeaker. Apart from this, let's not forget that many people (including me) can't resist the temptation or challenge to make the best amp ever (or whatsoever), even if it makes no sense practically or scientifically.
Cheers, E.
Maybe, maybe not. Point is that the distortion from an amp is quite different than from a loudspeaker. Apart from this, let's not forget that many people (including me) can't resist the temptation or challenge to make the best amp ever (or whatsoever), even if it makes no sense practically or scientifically.
Cheers, E.
I would agree with that. Most guys buy expensive equipment so they can "measure" the differences since they are so minute it is near impossible to "hear" them. I'm sure this is one of the things that drives the popularity of Ltspice.
Blessings, Terry
Edmond,
I know you are right, the challenge is always to know when to leave well enough alone sometimes. I guess I am just as anal as those working on the electronics side on the loudspeaker side. Little changes such as adhesives and motor design always seem to have those little things that you want to improve upon, slight changes in cone chemistry can make more difference than some may imagine but it never seems to stop. There always seems to be something else to chase.
I know you are right, the challenge is always to know when to leave well enough alone sometimes. I guess I am just as anal as those working on the electronics side on the loudspeaker side. Little changes such as adhesives and motor design always seem to have those little things that you want to improve upon, slight changes in cone chemistry can make more difference than some may imagine but it never seems to stop. There always seems to be something else to chase.
Well, the test instruments which measure the amps may point to the answere..... circuits which can measure real oscillator purity of -140dB readily exist. And, instruments which can measure to -150dB exist (directly with spec of 1dB error of harmonic). So, I would think from those that -160dB is about the real bottom of it all.
THx-RNMarsh
Okay, that's the measurement floor of the best instruments there are. But what is the lowest measured distortion of the best (op-) amp?
Cheers, E.
Cheers, E.