Bill,
Noise level is directly related to impedance! A power amplifier should just by that be 20 dB ish quieter then a preamp. The type of noise + distortion curve shown in the BT review is the classic noise level falls until it reaches the low point where the distortion rises. As the rise occurs at a point above the actual operating point the data is not very useful.
Now when measuring audio power amplifier distortion into a resistive load you not only have lower noise but again a not very useful measurement. My experience has been the conversion from electrical power to acoustic power is not of any use with resistors!
.
Actual loudspeakers seem to have reactive loads. Every time with a reactive load like an actual loudspeaker system for a test load the distortions I measure rise by very significant values.
I have never measured an audio power amplifier into a real load that had anywhere near the static measurements occasionally shown.
The best resistors I have looked at power levels above 1/4 watt can barely do -160 dB on distortion. So the trade off becomes noise vs distortion, not very surprising.
As to detecting signals buried in noise there is very good data on those thresholds. Surprise...not, the human perception process peaks for that in the midrange. Higher frequency harmonic distortion is less perceptable, but should not be confused with rising high frequency distortion that is a sign of other problems such as intermodulation or even wrap around distortion.
Noise level is directly related to impedance! A power amplifier should just by that be 20 dB ish quieter then a preamp. The type of noise + distortion curve shown in the BT review is the classic noise level falls until it reaches the low point where the distortion rises. As the rise occurs at a point above the actual operating point the data is not very useful.
Now when measuring audio power amplifier distortion into a resistive load you not only have lower noise but again a not very useful measurement. My experience has been the conversion from electrical power to acoustic power is not of any use with resistors!
. Actual loudspeakers seem to have reactive loads. Every time with a reactive load like an actual loudspeaker system for a test load the distortions I measure rise by very significant values.
I have never measured an audio power amplifier into a real load that had anywhere near the static measurements occasionally shown.
The best resistors I have looked at power levels above 1/4 watt can barely do -160 dB on distortion. So the trade off becomes noise vs distortion, not very surprising.
As to detecting signals buried in noise there is very good data on those thresholds. Surprise...not, the human perception process peaks for that in the midrange. Higher frequency harmonic distortion is less perceptable, but should not be confused with rising high frequency distortion that is a sign of other problems such as intermodulation or even wrap around distortion.
Last edited:
Bill,
In pro audio the original audio power amplifier gain has been a voltage gain of 20. So with a CD output of 2 volts and an 8 ohm loudspeaker that would provide 200 watts. Probably enough to damage most home loudspeakers.
As pro preamps can provide 10 volts or more that should be enough to smoke almost all transducers. However for marketing purposes some amplifiers have higher gain so they sound louder in demonstrations. Some folks prefer to have all of the amplifiers in their systems clip at some fixed input voltage often around 1.4 VACRMS so the gain increases as the output power goes up. (Holy VU meter hangover batman!)
Obviously a preamplifier is required for sources such as vinyl records. However CDs, radio tuners and tape decks often really don't need gain in a control selector system.
In pro audio the original audio power amplifier gain has been a voltage gain of 20. So with a CD output of 2 volts and an 8 ohm loudspeaker that would provide 200 watts. Probably enough to damage most home loudspeakers.
As pro preamps can provide 10 volts or more that should be enough to smoke almost all transducers. However for marketing purposes some amplifiers have higher gain so they sound louder in demonstrations. Some folks prefer to have all of the amplifiers in their systems clip at some fixed input voltage often around 1.4 VACRMS so the gain increases as the output power goes up. (Holy VU meter hangover batman!)
Obviously a preamplifier is required for sources such as vinyl records. However CDs, radio tuners and tape decks often really don't need gain in a control selector system.
Volti Audio Razz loudspeaker Measurements | Stereophile.com
This is not the way I would do a speaker. Flat to +/-10dB above 1kHz is not my idea of good...
This is not the way I would do a speaker. Flat to +/-10dB above 1kHz is not my idea of good...
Although there is some very good stuff going on using the purifi woofer in the multi-way forum here. It's expensive but you can get some very impressive performance from a 6.5" woofer in the right cabinet. Handy for those of us without room for larger speakers. And I am secretly jealous of those with room in their house for Fridge sized cabinets and 15" drive units whilst remaining married.
Edit: some people have dedication well beyond what I could get away with.
Edit: some people have dedication well beyond what I could get away with.
Attachments
Last edited:
Relatively easy to achieve with decent opamps these days. Regrettably, some say opamps are no good ‘for the sound’.
Yes you can even get the noise down further if you do a full Self on it an parallel up the opamps, but that is getting silly. I like silly in a hobby.
Just about every balanced DAC out there? Sony/Philips set the standard for CD output at 2 volts unbalanced, which is 4 volts balanced.
I have measured countless DACs with 4 volt output and some that go way above that (pro DACs). Here is a budget $200 balanced DAC:Schiit Modius Balanced DAC Review | Audio Science Review (ASR) Forum
I test preamps by simply using unity gain (4 volts in, 4 volts out). In a pass-through mode like this, there better not be a lot more distortion.
You can't replicate JA's results because there is insufficient documentation on the parameters used for his tests. No THD+N measurement for example is valid without bandwidth being mentioned. The wider the bandwidth, the more noise and distortion products are captured. Here is my graph with bandwidth clearly stipulated on top:
There is no such mention in JA's tests.
He also does silly things like use 600 ohm as output load. That is unusually harsh load and doesn't resemble any power amp input impedance that I know. You want to complain about something, you should complain about JA's tests, not mine.
Ever thought of emailing him and asking? He's a nice guy and I'm sure would be happy to share more details with you.