2Ohm, for example, is okay for me. But 0.5Ohm is outrageous. Sorry.
In different words, theoretically?SY said:
If someone do this in practice that circuit will preform near ideal only with one speaker model in that specific position in specific room.
It is interesting that from the beginning of hifi audio amplifier era everybody(well maybe not everybody, just those who wants that
) wants to make ideal audio amplifier for all those nasty behaved speakers. There are few no compromise amps in the world that approaches the ideal of voltage source, at wide band, with high speed response to input signal and all that without global feedback of any kind, and in 90% of cases there is still no true feeling of music touching listener through that high tech, no compromise amps. Why?
I don´t want to name brands that belong to that 90% or that in 10% region.
So, I believe that amp design is much more that fight with real speaker load traps and various distorsions. That is only a good base, or vice versa in some other approaches.
As I see, very few minds in the world know the secret of delivering the music to people.
Keith O. Johnson is certainly one of them.
Edmond Stuart said:
assume a virtually perfect sound reproduction, but 0.5 ohms @3khz.
5hz-50khz +/- 0.05 dB
108dB 2.83v 1m
internal reflections -60dB
signal -60dB after 1us of no signal
etc. etc.
do you
A. add a resistor and destroy the sound
B. add a transformer and destroy the sound
C. market it for the few
D. give up
janneman said:
I've only scanned the article so far. Looks really good though.
I did build Colin's Low Noise Measurement Preamp which uses the LSK389B JFET at the input and it had really low noise. I measured the noise of a Sony Bluray player at -115 dBV (same as specs). The input noise of the preamp was -132 dBV, measured flat with the input grounded.
I also built a preamp with 18dB of gain and no feedback, and it uses paralleled LSK170C's at the input. It's very quiet.
Uses Scott's AD797 as the measurement amp, a nice touch I thought
Jan Didden
Who Scott? Just curious.
One more question. Will anyone tell me why JFET noise is specified as nV/sqrt(Hz)? I guess this is to take the change in noise versus frequency into account. I'm just not an experienced audio engineer.
Hello Johnloudb, The Scott that was referred to with the AD797 and you asked about is Scott Wurcer, a regular contributor of information and commentary here, and the designer of the AD797 opamp for Analog Devices. If you go to Analogs website and look around you will find his work related bio.
Peace,
Dave
Peace,
Dave
Johnloudb said:
Not FET's in particular, physicists like to speak of equal power per freqency band normalized to a one Hz band i.e. volts squared per Hz. To make that plain volts it becomes volts/rt-Hz. For equal bins in an FFT or constant IF BW on a spectrum analyser you get a flat spectrum (white noise).
PMA, thanks.
Try measuring the impedance curve with your speaker in several different realistic spots in the room. Note how it hardly changes at all. You have to do something pathological, like put the speaker facing a surface a cm away, before you see any effect at all.
To understand how an amp will react to bad speaker loads, something like the Power Cube used by Audio Critic will do a nice job.
And 0R5 minima in the midrange are NOT typical, John. There's the WATT, which had a problematic crossover, and there's... there's... hmmm..... Now, if you're designing an amp to drive those, yes, the 0R5 load is an important parameter. But for a better engineered speaker, that's a lot less of an issue.
Try measuring the impedance curve with your speaker in several different realistic spots in the room. Note how it hardly changes at all. You have to do something pathological, like put the speaker facing a surface a cm away, before you see any effect at all.
To understand how an amp will react to bad speaker loads, something like the Power Cube used by Audio Critic will do a nice job.
And 0R5 minima in the midrange are NOT typical, John. There's the WATT, which had a problematic crossover, and there's... there's... hmmm..... Now, if you're designing an amp to drive those, yes, the 0R5 load is an important parameter. But for a better engineered speaker, that's a lot less of an issue.
Joshua_G said:
Hi Joshua,
No. Otala's method for measuring IIM involves feeding a large 60 Hz signal into the output of the amplifier under test through a resistor, like 8 ohms. I just followed Otala's test approach. I don't know of an IIM test that would employ a loudspeaker in place of the resistor in the Otala IIM test.
Cheers,
Bob
- Status
- Not open for further replies.