Well Gerhard, you are right and Bob is partially right. I easily fixed it, without adding resistors in the gates, how?
john curl said:
Make the capacitive load at the emitter smaller, or the emitter resistor.
A ferrite bead on the base would help, too, but this is frowned upon
by audio types. Shifting the serial resonance frequency of the base circuit upwards
to a frequency where the transistor runs out of gain will help, too.
(decoupling for RF with a small & very close capacitor).
But if you are unlucky and have a fast transistor, this will only move
the problem to a higher frequency.
I have formulated this in
<http://www.hoffmann-hochfrequenz.de/downloads/Hoffmann_VHF_Quarzoszillatoren_Teil2.pdf>
part 1 is interesting, too, but the articles are about VHF oscillators.
Text is both german & english, intertwined.
Gerhard
john curl said:
Hi John.
Well then why don't you share your secret faux pas with us?
It did occur to me that the different caps they put on the rail might have had higher winding inductance than the original, but the conclusion would still be that the rail was designed without enough regard for damping.
Cheers,
Bob
john curl said:
I've had more good transactions on e-bay than bad. I haven't bought anything in a while, partly because I can't justify/afford the time to cruise the joint looking for things and partly because I'd prefer not to start a PayPal account...which seems to be required these days. (My wife has one. I grumble about it, but could use hers if push came to shove.)
Unless Demian pops up with a really groovy idea, I'll probably go back to my fallback plan, which is to track down an AP System 1. Maybe Nelson will sell me one of his. As I understand it, the man owns a small herd of the confounded things. The money, per se, doesn't bother me nearly so much as the idea that I will get so little use out of it. If I could fiddle every day or even every week, I'd feel better about buying something, but plunking a couple of grand down for a toy I only get to play with once per blue moon curdles the Scottish portions of my genetic makeup.
You can either have money or time, but not both at once. The seeming exception to this...isn't. Assuming you retire with enough money to live comfortably, your health blindsides you and gobbles both your time and your money. Currently, I'm in the best financial condition I've ever been in, but have virtually zero time. My health is (fingers crossed) fine. I fully anticipate that my circumstances will change once I retire--like The One Hoss Shay, I'll find that everything falls apart at once:
The poor old chaise in a heap or mound,
As if it had been to the mill and ground!
You see, of course, if you’re not a dunce,
How it went to pieces all at once, –
All at once, and nothing first, –
Just as bubbles do when they burst.
(Oliver Wendell Holmes)
Grey
Bob Cordell said:
Agreed Bob.
Even though the dominant pole might be at a very LF the amp can
still have extremely low HF (20kHz) distortion characteristics and
be characterised as 'fast'.
Now, what if we were to get such an amp and resistively load the
VAS so the pole shifts nearer to or above 20kHz. This should
bring the distortion flatter over the audio band and less tilted up
at high frequencies.
One thing I notice about the Dart Zeel and the Ayre MXR open loop
amps that both got rave class A reviews - is their distortion,
although comparatively high, is almost entirely independent of
frequency.
Very few amplifiers exhibit this characteristic.
cheers
Terry
Another common element between the DartZeel and MX-R is that they both blew up in test. Maybe this means that in order to get a class A rating, the amp must blow up in test?
I found it quite entertaining to read the "my dog ate my homework" manufacturer's comments in both cases.
I found it quite entertaining to read the "my dog ate my homework" manufacturer's comments in both cases.
Terry Demol said:
Indirectly related: For years I've disliked current sources. I had a gut sense that there were frequency-related compliance problems, just based on the sound. In many posts over the years, I have recommended against current sources; use a simple resistor if at all possible. And, no, differentials do not always require a CCS, regardless of what some say. Through trial and error, I had homed in on a current source that was least offensive and tended to use it in cases where I absolutely needed a CCS.
Imagine my gratitude and relief a year, perhaps two ago when someone (Walt Jung, I think) went and methodically tested a dozen CCS topologies against each other. Now, he was looking for something a little different, but he published graphs and--mirabile dictu--the ones that showed the flattest response regarding compliance or noise or whatever it was he was looking for, were the same ones that I had found to be best sounding. The flattest of them all was the one I'd decided was the best of the group, sound-wise.
He did no listening tests, just bench tests.
I'll have to see if I can dredge that up so as to refresh my memory exactly what it was he was testing for. My point being that sometimes "flat" whatever (distortion, noise, etc.) is better sounding than "lower" that then rises at one frequency extreme or the other.
Just another case where "better" specifications do not necessarily mean better sound. And resistors are pretty bloody flat when it comes to response.
Grey
Gerhard, if you were available and I needed someone to troubleshoot a high frequency circuit. I would hire you.
GRollins said:
I do in my phono stage purely for "zen" reasons. I would never bother with the controls and rigor necessary to actually try and convince myself that there really was an audible difference. In casual listening I can never repeatably tell any differences in most things, including those some consider obvious.
andy_c said:
Same thing happened when I visited a friend reviewing Quicksilver monoblocks for TAS.
SY, I don't remember that one. What was the story with it?
Regarding the darTZeel, here's a classic quote from Stereophile:
"The darTZeel NHB-108 is like a Shelby Cobra—a handbuilt, high-performance design that makes demands on its always gratified owner, rather than a plug'n'play, never-very-gratifying Ford Taurus."
IOW, "The effing thing blew up, but we will polish that particular turd with some lame automotive analogy". Of course, the more expensive the gear, the less critical they become. Like in the case of the MX-R, blaming the high level of measured high-order harmonics on a failure of the FFT rather than placing the blame where it belongs. Anyone making value judgments of design techniques based on the assumption of reviewer credibility is pretty naive in my view.
Regarding the darTZeel, here's a classic quote from Stereophile:
"The darTZeel NHB-108 is like a Shelby Cobra—a handbuilt, high-performance design that makes demands on its always gratified owner, rather than a plug'n'play, never-very-gratifying Ford Taurus."
IOW, "The effing thing blew up, but we will polish that particular turd with some lame automotive analogy". Of course, the more expensive the gear, the less critical they become. Like in the case of the MX-R, blaming the high level of measured high-order harmonics on a failure of the FFT rather than placing the blame where it belongs. Anyone making value judgments of design techniques based on the assumption of reviewer credibility is pretty naive in my view.
My amp just 'oscillated', it is called protection, which drains away the subtle nuances from high end products. Many designers won't use it. Parasound can't afford NOT to.
You know, fellow engineers, Jocko and I call each other and just LAUGH. We find it pretty funny. I'm having a good time.
I'm glad you're having fun John 🙂. I thought the protection of the JC-1 either did nothing or shut it down? I wasn't aware of any V/I limiting that kicks in and out with an oscillation as some amps do.
john curl said:
I can better that.
A "hi-end" or audiophile amplifier reviewed by one of the big name magazines with a speaker protection circuit that was designed (by accident I assume) to self destruct the first time the circuit was powered on. This resulting in the protecton circuit no longer offering protection.
I reworked the circuit to fix the customers amplifier and offered it for free (my big mistake) to the manufacturer.
The manufacturer got a little offside about the whole thing.
GRollins said:
My feeling is from most of the designs I have seeen that most of the designers do not consider how the current souce/sink will perform dynamically above the audio range and sometimes even inside of it.
Things scroll by here really fast. . .
I have been working with sampling scopes for some time. They are the only way to look at an HDMI data stream. The sample rate isn't high- 100K to 1 MHz but the bandwidth is 20 GHz (and even 100 GHz) but only on repetitive signals. With averaging the signals are very sharp. Still I much prefer analog scopes when I need to look at an analog signal. I use a Tek 7854 which is a hybrid- 400 MHz analog mainframe with an early digitizer (not very fast but 12 bit).
As for a front end for a soundcard I can see combining several techniques to do it. Even the best soundcard doesn't have enough range to handle the full dynamic range expected of a distortion analyzer. I think it would need to be able to handle 300V RMS and autorange down to 10 mV or less with a readout (possibly available over USB). I would also use something like AP's transformer isolation on the output to remove ground loops. These are bigger problems than they look to be if the performance is to be state of the art. And for that last 40 dB of measurement a passive notch filter for the front end.
I have been working with sampling scopes for some time. They are the only way to look at an HDMI data stream. The sample rate isn't high- 100K to 1 MHz but the bandwidth is 20 GHz (and even 100 GHz) but only on repetitive signals. With averaging the signals are very sharp. Still I much prefer analog scopes when I need to look at an analog signal. I use a Tek 7854 which is a hybrid- 400 MHz analog mainframe with an early digitizer (not very fast but 12 bit).
As for a front end for a soundcard I can see combining several techniques to do it. Even the best soundcard doesn't have enough range to handle the full dynamic range expected of a distortion analyzer. I think it would need to be able to handle 300V RMS and autorange down to 10 mV or less with a readout (possibly available over USB). I would also use something like AP's transformer isolation on the output to remove ground loops. These are bigger problems than they look to be if the performance is to be state of the art. And for that last 40 dB of measurement a passive notch filter for the front end.
- Status
- Not open for further replies.