• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Silent but audible noise (hiss) in loudspeaker and SNR

Zek

Major Contributor
Joined
Feb 17, 2018
Messages
1,581
Likes
2,146
Can you try with your phone only connected to the amp?
Good idea - but the amp doesn't have a headphone connector.
I'm thinking about a battery driven source - but do not have any idea at the moment. It should be a device with low noise floor I think.
Looks like you didn't understand @Chrise36 - he asked if you had a mobile phone to try.;)
 

Chrise36

Major Contributor
Joined
Dec 23, 2018
Messages
1,043
Likes
359
So stereoplay says the dac is very clean but the amp seems not well designed. I read they messed with the power supply of the amp (to improve it).
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
Signal to noise is max signal with signal notched out and often in that condition noise floors are higher.

No, S/N is a measurement of idle/residual noise with no signal applied and the inputs shorted vs the full output. As were are discussing amplifiers here, no signal 'notching' is performed in S/N tests.

AP says it best:

SNR essentially characterizes the ratio between the full scale output of a device and its idle noise.
Dynamic range characterizes the ratio between the full scale output of a device and the spurious noise products created when a device is producing a very low level signal.


See here:
 
OP
C

chris256

Member
Joined
Jan 5, 2023
Messages
59
Likes
14
Using a phone would need an adapter cable to RCA which I do not have at the moment. Therefore I got a another idea: The headphone amplifier does have a symmetric output interface (fixed output). I used this and connected the A 200 (digital in to headphone amplifier, out via XLR to A 200). This way, I couldn't hear any hiss / noise in the tweeter / midrange. Seems the interface parameters preamplifier / power amplifier are not "optimal".

what does the damping factor switch do
Tested it - but it has no effect. But I wouldn't have expected anything to be honest at this point.
 
Last edited:

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
All those completely different values are confusing. Why isn't there a standard?

It all comes down to two parameters- the signal and the noise.

Bear with me. :)

The Signal
For preamps:
Manufacturers will specify the signal component as a rating in volts. At that rating, they also specify a few other parameters for their spec sheets. Like frequency repsonse and THD. As such they want attractive numbers.

That output voltage will likely be well below what the preamp is capable of. For instance, I have plenty of preamplifiers where they specify say 1.5V output for all their specs, but those preamplifiers are more than capable of 12V output. Reviewers should test against specifications as well as test for maximum capability, however they are inherently lazy and only push to the limit and present those tests results, rarely comparing or holding to account those numbers against rated specifications.

So we have variable levels, depending on whether they are being measured with respect to their rated voltages (the signal) or what they actually can produce.

The Noise
For preamps and power amps:
Noise is measured either with or without a shorted input, sometimes loaded with a particular resistance. That can vary.
Noise can be measured with the volume/gain controls wide open or at some other arbitary position.
Noise is measured over a specified bandwidth which must be stated- it cannot be assumed to be anything. Is it 20-20k or 20-200k? Widen the bandwidth and the noise increases. Shrink it and the noise decreases.
Noise can also have weighting filters applied such a A-weighting which takes out high and low frequencies, or a hum and noise filter can be utilized (say an HPF@400Hz)

All those variations of both signal and noise mean wildly different numbers can come from the various reviewers and test sites.

There are defacto standards for presenting S/N specifications, but you rarely see it in practice.
 

AnalogSteph

Major Contributor
Joined
Nov 6, 2018
Messages
3,334
Likes
3,278
Location
.de
Stereo:
S/N at 50mW/5W 67/63 dB 87/85 db (8 Ohm / 4 Ohm)
63 dB re: 50 mW / 4 ohms (or 316-ish µV) is knot grate. That would have been a rather average receiver back in 1981-ish. You have to try pretty hard to even make a power amplifier that's as bad as this - more often than not they're around 20 dB better. Ground loop in the measurement setup, perhaps?
T+A:
113 dB
That's more along the lines of what I'd expect, ref. to full power.
Stereoplay:
XLR (2,83 V at 8 Ω) 93 dB
XLR (10 V at 8 Ω) 104 dB
That, too, is rather more like it should be.
All those completely different values are confusing. Why isn't there a standard?
"The nice thing about standards is that you have so many to choose from; furthermore, if you do not like any of them, you can just wait for next year's model."- generally attributed to Andrew S. Tanenbaum (also see xkcd #927)

Seriously though, if you read old German reviews you'll find 50 mW / 4 ohm SNR (weighted and unweighted). 1 W / 8 ohm is more of an Anglosphere thing.

In case of a typical power amp, its noise performance tends to be entirely described by full power SNR and gain, IOW you can easily do the math on its expected output noise floor. That is not so much the case when preceding volume controls are involved (since the source impedance presented by a pot changes over its range, ranging between ~0 and Rpot/4), particularly multistage ones (typical modern-day PGAs, or 4-gang volume pots). Hence why testing with the volume turned down is a thing when testing integrated amps.

EDIT:
That output voltage will likely be well below what the preamp is capable of. For instance, I have plenty of preamplifiers where they specify say 1.5V output for all their specs, but those preamplifiers are more than capable of 12V output.
And then your seemingly-impressive 126 dB(A) of SNR becomes a much more modest 108 dB(A).

The unbalanced input of the T+A A200 is also unusually sensitive, with a gain of 37.4 = 31.4 dB. Then our example preamp would result in an output noise floor of ~224 µV(A), which is decidedly not inaudible if you're going to stick your ears close to the tweeters.

The headphone amplifier does have a symmetric output interface (fixed output). I used this and connected the A 200 (digital in to headphone amplifier, out via XLR to A 200). This way, I couldn't hear any hiss / noise in the tweeter / midrange. Seems the interface parameters preamplifier / power amplifier are not "optimal".
The XLR input has the immediate advantage of being 6 dB less sensitive. Plus, I would expect noise levels on a modern headphone amp's balanced pre-out to be rather lower as well, maybe 2-3 µV. The combination of both would result in a dramatically lower output noise floor.

As you noticed, a TDA7303 isn't particularly exciting, it's more the kind of thing I'd expect in a micro stereo. It would honestly be fine if used with modest gain, but definitely not under these circumstances. T+A's specs are still possible if maximum output is higher than the 1 Vrms reference in the TDA7303 datasheet (the chip is quite capable of >2 Vrms if you insist, though distortion is not exactly going to improve).
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
63 dB re: 50 mW / 4 ohms (or 316-ish µV) is knot grate. That would have been a rather average receiver back in 1981-ish. You have to try pretty hard to even make a power amplifier that's as bad as this - more often than not they're around 20 dB better. Ground loop in the measurement setup, perhaps?

That's more along the lines of what I'd expect, ref. to full power.

That, too, is rather more like it should be.

"The nice thing about standards is that you have so many to choose from; furthermore, if you do not like any of them, you can just wait for next year's model."- generally attributed to Andrew S. Tanenbaum (also see xkcd #927)

Seriously though, if you read old German reviews you'll find 50 mW / 4 ohm SNR (weighted and unweighted). 1 W / 8 ohm is more of an Anglosphere thing.

In case of a typical power amp, its noise performance tends to be entirely described by full power SNR and gain, IOW you can easily do the math on its expected output noise floor. That is not so much the case when preceding volume controls are involved (since the source impedance presented by a pot changes over its range, ranging between ~0 and Rpot/4), particularly multistage ones (typical modern-day PGAs, or 4-gang volume pots). Hence why testing with the volume turned down is a thing when testing integrated amps.
The voltage for 50 mw @ 4ohms is interestingly 1/20th the voltage at 1 watt for 8 ohms.
 

restorer-john

Grand Contributor
Joined
Mar 1, 2018
Messages
12,579
Likes
38,274
Location
Gold Coast, Queensland, Australia
I hate RPN and HP calculators.

My scientific calcs are old and I just cannot get used to the "natural" entry of the 'new' ones everyone uses. Sure it's way better and more logical but I'm hardwired with the 1980s style entry (backwards). RPN is worse again, but it is what you get used to I guess.

I'm also a sucker for scientific calculators. I pick them up in op-shops for close to nothing. Not sure how many I have...:facepalm:

I still have my calc from secondary school, perfect in its box.
 
OP
C

chris256

Member
Joined
Jan 5, 2023
Messages
59
Likes
14
a TDA7303 isn't particularly exciting, it's more the kind of thing I'd expect in a micro stereo
That's true! The preamp module as provided by T+A is additional and obviously more a workaround to get a preamp into a device which is originally designed for use with an integrated amplifier. Unfortunately I didn't know this before (the quality of the preamp module) ... . On the other hand, it's a smart "solution" to prevent unnecessary waste, but it would have been nice if I had some more detailed information about the device beforehand.

Meanwhile I extended my tests and I have to update some findings. It's arbitrarily complicated to talk about phenoms like this without having usable measuring instruments but subjective descriptions like loud or not so loud instead. By chance I discovered that my left ear is not quite so fit anymore. Therefore I did all the tests again with the right ear... . The results are here:

ConditionDistance ear <-> tweeterRemarks
1. preamp volume > 0~ 35 cmXLR connection via preamp module.
- Output: 5 V[eff] / 22 Ohms
- Input sensitivity: 1.6 V[eff] / 20 kOhms
Tone control and loudness is always off.
2. preamp volume = 0~ 15 cmXLR connection
I could slightly hear some music in the midrange speaker if some music is given. It shows: no mute is involved.
You can hear a silent plop (changed DC offset?) when switching from 1 to 0 in the tweeter (or back again). It is not a relay but a circuit switched by a transistor.
3. preamp muted~ 10 cmMuting is realized via relay.
XLR connection
4. preamp off~ 10 cmpreamp switched off.
5. preamp / power amp disconnected~ 10 cm
6. fixed output~ 15 cmThe Music Player was connected via the fixed output (RCA!) to the power amp.
- Output: 2.5 V[eff] / 22 Ohms
- Input sensitivity: 800 mV[eff] / 5.8 kOhms
No mute testable as mute function doesn't exist for the fixed output.
7. preamp volume 0~ 15 cmRCA / couldn't find an audible difference between RCA vs. XLR

My conclusion now:
Regarding test 1, 2 and 6, the TDA 7303 seems to be the main reason for the high noise level. If the volume is set to 0, the TDA seems to be out of the signal path as the noise is now mostly the same as I heard when using the fixed output.
The power amp itself isn't that noiseless, too. Without the TDA, it looks like most of the noise would be produced by the power amp itself (regarding test 3,4 and 5).
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
My scientific calcs are old and I just cannot get used to the "natural" entry of the 'new' ones everyone uses. Sure it's way better and more logical but I'm hardwired with the 1980s style entry (backwards). RPN is worse again, but it is what you get used to I guess.

I'm also a sucker for scientific calculators. I pick them up in op-shops for close to nothing. Not sure how many I have...:facepalm:

I still have my calc from secondary school, perfect in its box.
I had several old calcs. Lots of TI-30s. From the early models with led readouts to those with lcd. All ended up with a stuck key or hard to register key issues. Started using Casio brand and never could wear one out. Now I use a smartphone mostly. I do still pull out a solar powered Casio when I'm doing lots of calculations.
 

pogo

Major Contributor
Joined
Sep 4, 2020
Messages
1,239
Likes
382
Tested it - but it has no effect. But I wouldn't have expected anything to be honest at this point.
Interesting, I have been able to detect tonal differences here.
Which speakers do you use?

If you speak German, you could discuss your noise observation directly with a T+A Fellow in the HIFI FORUM.
 

pogo

Major Contributor
Joined
Sep 4, 2020
Messages
1,239
Likes
382
What does the damping factor switch do?
This was explained quite well here: Link
See also here: Link

'A high damping factor tends to produce a more clearly defined, very precise and analytical sound image, whereas a reduced damping factor produces a more warm and softer sound image.' <- extract from the T+A A200 manual
 
Last edited:
Top Bottom