PMA said:
The purpose of a PhD thesis is not to set the world on fire with new and wondrous advances. The purpose is to have the student demonstrate an understanding of the process, and finish before the advisors die of old age.
The reason a "brilliant idea" is rejected for a PhD thesis has to do with what can be reasonably done by the student. PhD advisors try to keep the candidate from setting themselves up for failure.
The bottom line is...go with the advisors, do the work, get the degree.....then set the world on fire..
John
jneutron said:
Or just don't waste your time on all that, PhD is not strictly required for setting the world on fire 🙂 . Quite often people spend their most productive years doing these mostly pointless exercises... .
Alex
x-pro said:
I will not debate the decision to get a PhD or not.
I pointed out why advisors will reject thesis topics.
As an aside, I find it disturbing to think that two or three years post masters might be "their most productive years"...
John
x-pro said:
Well, ya gotta be sensitive to us old guys...sheesh, I'd hate ta think my most productive years wuz three decades ago..
Cheers, John
jneutron said:
Well, ya gotta be sensitive to us old guys...sheesh, I'd hate ta think my most productive years wuz three decades ago..
Cheers, John
John,
I certainly was more creative as an engineer 30 years ago, now I have to compensate decline in creativity with experience.
Cheers,
Alex
NO, don't tell me they broke you..x-pro said:
all they're doing is tossing marchmallows at you...Cheers, John
ps..it's unfortunate that the entire purpose of an engineering degree is to limit the creative excursions..to that which is taught..It has it's good points, and bad..
I'm surprised nobody's identified a "pin 1 problem" on the output of devices yet... but I'm patient..
jneutron said:NO, don't tell me they broke you..
John,
at the time (1978) it was the USSR after all. The only way for an electronics engineer to do some real stuff there was to work for the military. I had (and still have) a serious allergy to that field. Only after I came to the UK 17 years ago I had a chance to do some proper electronics design for living.
However all that is off-topic 🙂 and I should stop now
Cheers
Alex
john curl said:
John is largely correct. I also learned about triple beat many decades ago. One place where it was used extensively was in cable TV technology. My contribution was to come up with an attractive set of three in-band tone frequencies (9.0, 10.05 and 20 kHz) to best test audio equipment for high-frequency IM without resort to a spectrum analyzer.
The sine-square test was a good test but it required a spectrum analyzer, which was expensive at the time.
I demonstrated in my paper that the MIM test results were highly correlated to tests using the sine-sqare tests for TIM in cases where the amplifier was not into hard slew rate limiting (i.e., soft TIM). John is right that the sine-square test, especially the DIM-100 version, creates much higher peak slew rates. The price one pays for achieving such high slew rates is that much of the signal lies out of band. That is fine for many test situations, but my objective was a fully in-band test.
Given the more economical availability of spectrum analyzers these days, my favorite test now is the 19+20 kHz CCIF IM test with full spectral analysis.
Cheers,
Bob
Unfortunately, due to different practices of producers, one has often to solve the pin 1 XLR problem individually.
PMA said:
Actually, when I speak of pin one problem, I refer to the generic sensitivity of the system to ground loop currents.
I see much effort (some of it misdirected) in fixing amplifier input sensitivity to ground currents, but have yet to see anybody worry about the output of a preamp (remember, this is a preamp thread) and it's sensitivity to ground loop currents...
Reminds me of the starting scene of mash, where everybody's looking the wrong way for the helicoptor...
Cheers, John
Regarding sine-square TIM test, in case that the circuit has decent step response without input RC filter, this step response is same for small and large amplitudes, i.e. it is not slew rate limited, has not large FB and dominant pole comp., this alltogether fulfilled, there is no reason for TIM as known from the seventies.
I rather prefer to use fast rise input step and fast scope, to explore step response in whole range of amplitudes.
There may be another issue, not seen with step. It is crossconduction in output stage at frequencies of hundreds kHz or some MHz, near to full amplitude. This might not occur with square of lower frequency.
I rather prefer to use fast rise input step and fast scope, to explore step response in whole range of amplitudes.
There may be another issue, not seen with step. It is crossconduction in output stage at frequencies of hundreds kHz or some MHz, near to full amplitude. This might not occur with square of lower frequency.
When running the sine/square test, what quadrants are being exercised in the amplifier?
John
John
Sorry, that was not actually directed at you..😉PMA said:
It was more of a global statement.
John
I did not take it personally. Just to support importance of the issue. I have a lot of measurements on this.
PMA said:
The cross-conduction issue with power amplifiers is a good point. One symptom of this is increased power supply current for a given high output level as frequency is increased.
In fact, some amplifiers exhibit this behavior at 20 kHz near full power. It is usually a result of the output stage circuit not being able to turn off the output transistors quickly enough. It was originally described as dynamic crossover distortion by, I think, Bongiorno. Amplifiers that exhibit this behavior at frequencies as low as 20 kHz will, however, almost inevitably exhibit significantly elevated THD-20 as well.
Cheers,
Bob
Bob Cordell said:
For sure, in case it is at as low frequency as 20kHz (it would be a poor design according to up-to-date SOA).
The same (similar) phenomenon we may see in a discrete diamond buffer, at frequencies of units or tens of MHz. A cap between bases of output devices may help a bit.
- Status
- Not open for further replies.