Today I assisted a friend who was expressing some concern about the amount of hiss produced by his new Sony receiver. This is a brand new model, 6 channel AC3/DTS amp (not sure of the model, sorry). We found that hiss was quite noticable on the output of the amp under normal listening conditions.
We had the amp set up and were listening to a DVD. With no 'absolute level' to work by, we set the volume level to a typical sort of movie level (hard to believe I'm a scientist, I know ;-)), and I measured the noise output at the speakers. I found in 2 channel mode 13-17mV rms at the speaker. This level was acceptable at the listening position. In AFD mode, which seems to be required for AC3 discs etc, the noise measured 28-31mV (hiss!) In the Sony 'movie' mode, the noise was 39-45mV (very noticable hiss!). Incidentally, this noise did not depend on the input selected - it was the same with even nothing plugged into the amp. The noise would increase with an increase in volume knob setting. In mute mode the output was 3mV (silence is golden!).
We found that anything above 20mV or so with his speakers was objectionable.
My question to those amp designers out there is what is classed as 'acceptable' for an amp? Of course whether noise is heard at the listening position will depend on the sensitivity of the speakers, still, this Sony seems to produce quite a lot of noise. At the moment I have no measurements from any other amps.
We also found that the noise depended on a number of settings on the amp. By changing the centre channel speaker setting from a phantom type mode to 'large' as well as selecting a separate sub-woofer output, the noise went from 30mV or so down to an acceptable 20mV. Presumably due to reduced channel mixing required.
I'm just glad he doesn't own 99dB/W speakers!
BTW, the amp appears to have a switchmode power supply! (High pitched noise with ear to case) Plus small power transformer and what looks like chokes and things, plus two ferrite cores on the power cord!!!
Anyway, back to the original question - what's a good level to amp for when designing an amp? 5mV noise, 10mV, 20mV??? Now I'm curious!
We had the amp set up and were listening to a DVD. With no 'absolute level' to work by, we set the volume level to a typical sort of movie level (hard to believe I'm a scientist, I know ;-)), and I measured the noise output at the speakers. I found in 2 channel mode 13-17mV rms at the speaker. This level was acceptable at the listening position. In AFD mode, which seems to be required for AC3 discs etc, the noise measured 28-31mV (hiss!) In the Sony 'movie' mode, the noise was 39-45mV (very noticable hiss!). Incidentally, this noise did not depend on the input selected - it was the same with even nothing plugged into the amp. The noise would increase with an increase in volume knob setting. In mute mode the output was 3mV (silence is golden!).
We found that anything above 20mV or so with his speakers was objectionable.
My question to those amp designers out there is what is classed as 'acceptable' for an amp? Of course whether noise is heard at the listening position will depend on the sensitivity of the speakers, still, this Sony seems to produce quite a lot of noise. At the moment I have no measurements from any other amps.
We also found that the noise depended on a number of settings on the amp. By changing the centre channel speaker setting from a phantom type mode to 'large' as well as selecting a separate sub-woofer output, the noise went from 30mV or so down to an acceptable 20mV. Presumably due to reduced channel mixing required.
I'm just glad he doesn't own 99dB/W speakers!
BTW, the amp appears to have a switchmode power supply! (High pitched noise with ear to case) Plus small power transformer and what looks like chokes and things, plus two ferrite cores on the power cord!!!
Anyway, back to the original question - what's a good level to amp for when designing an amp? 5mV noise, 10mV, 20mV??? Now I'm curious!
Typical noise for a high quality power amp is below 100uV with inputs shorted. This is a Vrms noise measured with a 20-20k bandwidth.
Typically an A-weighted noise figure would be more interesting, but can unfortuneately not be measured by a normal Digital multimeter, but needs a weighting filter between source and meter. A typical digital multimeter will normally not measure Vrms with a 20-20k bandwitdh, maybe a 20-500Hz or so bandwidth would be more typical if the meter is not a true rms type.
It seems the noise in your amplifier is comming from the preamp or signal processing stage of the amplifier, as you measure a noise level at a few millivolts with the volume turned completely down. Anyhow the measurement value of a few millivolts AC on a Digital multimerer is not necessarily very accurate. So it seems the power amp part has an OK noise level, while the other parts look a little noisy to me.
regards Noize
Typically an A-weighted noise figure would be more interesting, but can unfortuneately not be measured by a normal Digital multimeter, but needs a weighting filter between source and meter. A typical digital multimeter will normally not measure Vrms with a 20-20k bandwitdh, maybe a 20-500Hz or so bandwidth would be more typical if the meter is not a true rms type.
It seems the noise in your amplifier is comming from the preamp or signal processing stage of the amplifier, as you measure a noise level at a few millivolts with the volume turned completely down. Anyhow the measurement value of a few millivolts AC on a Digital multimerer is not necessarily very accurate. So it seems the power amp part has an OK noise level, while the other parts look a little noisy to me.
regards Noize
Re: Re: Acceptable noise level at amp output
Hi guys, thanks for the replies.
Indeed the noise appears to be coming from the pre-amp stage. Activating mute (which I'm guess is likely to take effect before the power amp stage) reduced noise to 3mV, which is quite an improvement on the initial 15-50mV.
For the record I was using a hand-held oscilloscope. One of those Velleman (spelling?) devices, and that has an rms output reading.
Hi guys, thanks for the replies.
Indeed the noise appears to be coming from the pre-amp stage. Activating mute (which I'm guess is likely to take effect before the power amp stage) reduced noise to 3mV, which is quite an improvement on the initial 15-50mV.
For the record I was using a hand-held oscilloscope. One of those Velleman (spelling?) devices, and that has an rms output reading.
I remember reading elsewhere that the new lower end Sony 'digital' receivers have significant hiss problems in their dsp circuitry (a design or manufacturing problem) that are causing much customer dissatisfaction. Is it the STR-DA3000es or 5000, which I recall are the models supposed to be most afflicted? IAC, I think a few millivolts of hum at the output of an amplifier is far less objectionable than the same absolute level of hiss.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.