Onvinyl said:
input waveform signal source must be connected in circuit, parameters to be filled in
Transient analysis box:
X axis : F
Y axis : harm(variable), variable = voltage, current, or sum of more voltages, currents
time range: in case of 1kHz input for example 10m, but analyze only 1m (1 last period od 10). In FFT parameters set Tmin = 9m, Tmax = 10m. Here the circuit was fast enough, so I analyzed 1m - 2m, 2nd period.
maximum time step: proper choice. Try 100n for 1m analyzed
number of points: 16384, you can try less to speed up, distortin must not rise.
Avoid RC time delay circuits in signal path.
.... as you see, it is easy. Just play several years with that and you will easily get reliable results
😉
One can easily make a mistake. Here is a "harmonic distortion" from pure RC circuit. Of course there is no harmonic distortion in fact. Just problem of inherent MC dist. an. parameters vs circuit parameters.
When one simulates, he has to know behaviour of real circuits, math, FFT, everything. Otherwise there is a danger of misleading "pure hobby" result.
When one simulates, he has to know behaviour of real circuits, math, FFT, everything. Otherwise there is a danger of misleading "pure hobby" result.
Attachments
Justcallmedad said:
This would be fine, but you often do not know how much.
Hereby the same circuit, but "another" approach (not delay). Now the result is OK.
Sorry to be boring, I finish with this. Only wanted to point at possible sources of mistakes. It is not a problem of a simulator, but the user.
P.S.: sorry for different amplitude of sine source.
Attachments
Richard,
you will have to wait way more than only one RC time constant, more like 10 or 100 (depending how low you want your residual error to be).
A way to improve this (as 100+ cycles increases sim time very much) is to use a generator with shaped fade-in, the best shape is normally raised cosine of a few periods (contains the lowest amount of "false" spectral products -- including DC component -- of any fade-in contour for a given fade-in time).
- Klaus
you will have to wait way more than only one RC time constant, more like 10 or 100 (depending how low you want your residual error to be).
A way to improve this (as 100+ cycles increases sim time very much) is to use a generator with shaped fade-in, the best shape is normally raised cosine of a few periods (contains the lowest amount of "false" spectral products -- including DC component -- of any fade-in contour for a given fade-in time).
- Klaus
EUVL said:
Patrick, In this case the output at least in HF is connected to +rail via the mosfet drain gate parasitic capacitance (often high) which creates a peak at high frequency and limits bandwidth.
The solution would be a few leds replacing R13/R16 and 2k/3k resistor at gates input.
Richard,
you will have to wait way more than only one RC time constant, more like 10 or 100 (depending how low you want your residual error to be).
KSTR said:
Agree with you, Blackman-Harris window should be better in this case.
Richard was quite right here, one of the possibilities is to insert same RC circuit after the sine source and to analyze vs. voltage behind that RC. The problem is, that we often do not know the real "delay" (phase shift). In case we are deep in flat area of freq. response, increase of analysis cycles usually helps.
john curl said:
This was originally for the software industry but, after adapting, it fits for audio engineering as well.
-----------------------
The Klingon Audio Engineer decalogue
- Audio specifications are for the weak and timid!
- This amp is a piece of GAGH! I need 2SK389/2SJ109 if I am to do battle with this design!
- Distortion?! - I will show you distortions when I distort your skull!
- What is this talk of 'measurements'? Klingons do not make 'measurements'. Our amps 'escapes' leaving a bloody trail of designers and audiophiles people in its wake.
- Simulating? Klingons do not 'simulate'. Our amps does not coddle the weak!
- I have challenged the entire Stereophile team to a Bat-Leth contest. They will not concern us again!
- A TRUE Klingon engineer does not 'experiment' or 'breadboard' his amp! Let that human vermin deal with this!
- By publishing this review you have challenged the honor of my family. Prepare to die!
- You question the worthiness of my design? I should kill you where you stand!
- Our users will know fear and cower before our amps! Ship it! Ship it and let them flee like the dogs they are!
john curl said:I might also say that this controversy started when Bob Cordell challenged me to produce a SPICE model of my JC-1 power amp, months ago, even after 100's, if not 1000's had already been produced and successfully used, as well as independently measured by 'Stereophile'. The absurdity of his demand got me to respond jokingly: "Spice, I don't need no stinkin' Spice" which started the Spice thread, that I occasionally read with interest.
Hi John,
It is a shame that a professional like yourself has to resort to fabrication in demonizing someone who does not always see things his way. Below is what I actually said (BJT vs MOSFET thread, page 58, post 1449).
Been to Bosnia lately? Been under fire?
Stick to the truth and you will get a lot more respect and credibility.
Cheers,
Bob
Bob Cordell said:
Bob,
Your post has a rather 'the guilty flee when no man pursueth' flavor to it.
Why on earth should John--or anyone else for that matter--bother simulating a finished design that's already shipping? And why should you accuse him of overreacting when you're requesting he do a simulation of questionable value after the fact?
I find your reaction to be a little on the touchy side.
Shall we run the tape and see if there was sniper fire when you got off the plane in Bosnia?
Grey
Your post has a rather 'the guilty flee when no man pursueth' flavor to it.
Why on earth should John--or anyone else for that matter--bother simulating a finished design that's already shipping? And why should you accuse him of overreacting when you're requesting he do a simulation of questionable value after the fact?
I find your reaction to be a little on the touchy side.
Shall we run the tape and see if there was sniper fire when you got off the plane in Bosnia?
Grey
Justcallmedad said:
Funny how that works. Get rid of the feedback in the Wilson current mirror and it lowers the high-order distortion.... Now where have I heard that before....hmmm....let me think....
Justcallmedad said:
Yep. Well worth the cost if you are trying to build the best.
Justcallmedad said:
Personally I would rather use the Hitachi lateral parts (even with their imperfections) than any vertical devices. I think John Curl agrees with me on this, at least in the case of the Blowtorch.
As far as BJT's are concerned, I have no experience. I stopped using folded cascodes and switched to current mirrors instead. You will just have to try it for yourself.
Justcallmedad said:
Listening test are the *only* way to know which sounds better!
The difficulty is to make a fair test, where there is only *one* variable that changes. As you pointed out, the folded cascode requires a top-notch power supply. But I would put the same power supply on both circuits to keep things even.
Same for the PCB material. It's not necessarily important to use the very best materials, but they must be the same for each test. And you must make sure that they have both equal amounts of break-in time. Et cetera, et cetera.
It isn't easy to do a truly fair comparison, as there are so many variables to keep track of.
Good luck!
I agree, Grey. First, he IMPLIES that my amp is unstable. Did he measure it? NO! But he expects me to satisfy his concerns, rather than any objective evaluation. By the way, I tried to follow up with John Atkinson, where his mysterious oscillation came from. He did not know, and could not REPEAT the oscillation. How about that? Why did you not ask him, yourself, Bob, before implying that I did anything incompletely or inaccurately?
Actually, at least to me, 'So it sounds like your amp is not unconditionally stable into a capacitive load. That's OK. It is a reasonable choice,' comes across as damning with faint praise at a bare minimum, but more likely accusatory in tone...nothing implied about it.
You'd better watch your step, John. Next thing you know he'll fire off a letter to Stereophile like he did with Charles.
Bob,
I, for one, am still waiting for the specs to back up your various claims in the letter you wrote to Stereophile dunning Charles. Surely you had rock-solid information in hand before writing a letter like that...didn't you?
Grey
You'd better watch your step, John. Next thing you know he'll fire off a letter to Stereophile like he did with Charles.
Bob,
I, for one, am still waiting for the specs to back up your various claims in the letter you wrote to Stereophile dunning Charles. Surely you had rock-solid information in hand before writing a letter like that...didn't you?
Grey
Fascinating, this thread has just transversed through an odyssey that's melded old school intuition with current technology to a satisfying convergence, only to be slam-dunked back to an unnecessary soap opera that both John and Bob have demonstrated being technically far above.
I guess this is a part of human nature conversing that is hard to escape in our daily life.
Let me quote Eleanor Roosevelt:
“Great minds discuss ideas.
Average minds discuss events.
Small minds discuss people."
I guess this is a part of human nature conversing that is hard to escape in our daily life.
Let me quote Eleanor Roosevelt:
“Great minds discuss ideas.
Average minds discuss events.
Small minds discuss people."
Or--
"If you haven't got anything nice to say about anybody, come sit next to me."
- Alice Roosevelt Longworth (1884-1980)
Back when I was dumb enough to agree to be the first moderator here, I tried to steer things towards an informal, sort of university-ish format. To the extent that any one person can be said to have given direction to anything as big as DIY has become, I will also have to take the blame for the other informal things that sprang from that. My original intent was to allow people enough elbow room to say things about themselves that would round them out as people. The web is a huge and impersonal thing and I was hoping that if members saw each other as people rather than AI programs attempting to pass a Turing test, they might form friendships and be able to work together on larger, more intricate, more complex projects.
Did I succeed?
Well, yes...and no...
Along with the good stuff come the darker sorts of human behavior. There's never been a filter that discriminates between better and worse human traits, and all attempts to enforce arbitrary standards inevitably end up sliding to the dark side; utopias invariably become corrupted over time.
My personal experience--all the way back since I was a child and continuing to this day--is that if you don't fight back the bullies win--by default if nothing else. An attack, uncontested, invites more of the same. Eventually, you reach a point where the bad behavior becomes so ingrained that it's impossible to eradicate. Like territorial or religious conflicts, the problem acquires so much momentum that it's impossible to stop at one particular point and say with finality that henceforth no more of this will be allowed. There's never a time when attack and counterattack, revenge and counter-revenge, vendetta and counter-vendetta are precisely poised in such a way that everyone can walk away reasonably content. And so at least one side views itself as unjustly deprived of their chance to balance the books. Their feelings fester and in time burst into the open again.
I freely admit that there are people who did not grow up in the circumstances I grew up in. Their views differ from mine. I'm okay with that. All I know is that any time I attempted to turn the other cheek, I got whacked on that side, too. Generally harder than the first time, too. Yes, turning the other cheek does work for some people...I've seen it for myself. But for reasons that are not clear to me, it doesn't work for everyone.
My feeling is that John, Nelson, Charles...and, yes, you too, Scott, are worth a whole herd of the nay-sayers. My goal then...and now...is to make this place as hospitable as possible for them, whether I have any official capacity or not. I'm far more interested in what happens in the real world than I am in somebody's simulation. To the extent that those guys have done real things in the real world, they've got quite a bit more credibility than people who sit around and argue from theoretical positions--particularly positions that I learned long ago to be flawed.
But, yeah, when you leave the door open, friends enter easily...but so do the mosquitoes.
Pass me that can of bug spray, will ya?
Grey
"If you haven't got anything nice to say about anybody, come sit next to me."
- Alice Roosevelt Longworth (1884-1980)
Back when I was dumb enough to agree to be the first moderator here, I tried to steer things towards an informal, sort of university-ish format. To the extent that any one person can be said to have given direction to anything as big as DIY has become, I will also have to take the blame for the other informal things that sprang from that. My original intent was to allow people enough elbow room to say things about themselves that would round them out as people. The web is a huge and impersonal thing and I was hoping that if members saw each other as people rather than AI programs attempting to pass a Turing test, they might form friendships and be able to work together on larger, more intricate, more complex projects.
Did I succeed?
Well, yes...and no...
Along with the good stuff come the darker sorts of human behavior. There's never been a filter that discriminates between better and worse human traits, and all attempts to enforce arbitrary standards inevitably end up sliding to the dark side; utopias invariably become corrupted over time.
My personal experience--all the way back since I was a child and continuing to this day--is that if you don't fight back the bullies win--by default if nothing else. An attack, uncontested, invites more of the same. Eventually, you reach a point where the bad behavior becomes so ingrained that it's impossible to eradicate. Like territorial or religious conflicts, the problem acquires so much momentum that it's impossible to stop at one particular point and say with finality that henceforth no more of this will be allowed. There's never a time when attack and counterattack, revenge and counter-revenge, vendetta and counter-vendetta are precisely poised in such a way that everyone can walk away reasonably content. And so at least one side views itself as unjustly deprived of their chance to balance the books. Their feelings fester and in time burst into the open again.
I freely admit that there are people who did not grow up in the circumstances I grew up in. Their views differ from mine. I'm okay with that. All I know is that any time I attempted to turn the other cheek, I got whacked on that side, too. Generally harder than the first time, too. Yes, turning the other cheek does work for some people...I've seen it for myself. But for reasons that are not clear to me, it doesn't work for everyone.
My feeling is that John, Nelson, Charles...and, yes, you too, Scott, are worth a whole herd of the nay-sayers. My goal then...and now...is to make this place as hospitable as possible for them, whether I have any official capacity or not. I'm far more interested in what happens in the real world than I am in somebody's simulation. To the extent that those guys have done real things in the real world, they've got quite a bit more credibility than people who sit around and argue from theoretical positions--particularly positions that I learned long ago to be flawed.
But, yeah, when you leave the door open, friends enter easily...but so do the mosquitoes.
Pass me that can of bug spray, will ya?
Grey
john curl said:
We own one of the same Audio Precision Dual Domain System One analyzers that JA used back then. (He recently switched to the Audio Precision 2722, which we also have.) I have found that under certain rare and specific circumstances, the System One can cause the DUT to oscillate. I have been able to correct this by changing the grounding scheme of the amp OR by changing the grounding scheme of the Audio Precision. Please note that the oscillation *never* occurred in the real world, only when connected to the System One.
I suspect that the "oscillation" in your amplifier was an artifact produced by the grounding peculiarities of the System One in that particular configuration.
Charles,
Okay...so for those of us who have larceny in our hearts and intend to knock over old ladies in order to steal their bingo money, then scramble to eBay to bid on an AP1...
What do you change in an AP1 to keep it from giving circuits fits?
Grey
Okay...so for those of us who have larceny in our hearts and intend to knock over old ladies in order to steal their bingo money, then scramble to eBay to bid on an AP1...
What do you change in an AP1 to keep it from giving circuits fits?
Grey
GRollins said:
Sorry, it's been too long and I can't remember for sure. The basic problem is that you don't just have a box called the "System One". You also have a computer, a monitor, a keyboard, dummy loads, and (if you're smart) an oscilloscope. Plus the DUT. And a Variac and a DVM and whatever else you've got hooked up on your bench.
So you end up with a great big, complex system with plenty of potential for unwanted interactions. Just look at the "waterfall" plots on Stereophile. Many of them have a big sharp ridge at 15.75 kHz. This is because the CRT monitor creates a big external magnetic field from the horizontal deflection yokes that is picked up in the external wiring. When we test phono stages, we have to turn the CRT off to get valid results. Even line level gear can get funny results if the DUT is too close to the CRT monitor.
I think it turned out to be that the leads to the dummy loads were zip-tied to a metal shelf that also had the computer on it. When we physically moved the cabling, the problem went away. But this problem only showed up with one specific product. So it was some strange interaction. Remember, with the System One, the generator and the analyzer are in the same box, so there is plenty of opportunity for troubles to arise. The fact that I've only seen this particular problem once is a testament to the designers at Audio Precision.
But there are several options with regard to how to connect the DUT. Single-ended versus balanced, floating versus grounded. And the grounding connection has an internal fuse that is all-too-easily blown that has no external indicator. So sometimes you *think* the DUT is grounded to the System One when it actually isn't...
Charles Hansen said:
Not only with one specific product. The same can be found when testing simple opamp in a high noise gain circuit, and it is device dependent! With some you have it, with other not. Same test board, same wiring, just change IC. Demodulation, detection?
GRollins said:
Grey, do you have a problem with Comprehension of Written English? Bob didn't 'request' anything. He asked wheteher it was done or not. (That's the sentence with the question mark at the end).
We can expect John to turn a simple question into a 'challenge' and 'absurdity'; it's just the way he muddles the issues.
But I always thought you to think for yourself. I apologize for my obvious mistake.
Jan Didden
- Status
- Not open for further replies.