AVS Forum banner

2561 - 2580 of 7704 Posts

·
Registered
Joined
·
151 Posts
Does that mean that noise is the noise floor of the amp?
Yes it does. You don´t need to have 50/60 Hz noise component in amps output more than couple of millivolts, and 104 dB sensitive full range speaker like 4722 reports it to you loud and clear...

But hey, you did have completely noise free system already, didn´t you? :D
 

·
Registered
Joined
·
7,132 Posts
Yes it does. You don´t need to have 50/60 Hz noise component in amps output more than couple of millivolts, and 104 dB sensitive full range speaker like 4722 reports it to you loud and clear...

But hey, you did have completely noise free system already, didn´t you? :D
Boy you are never gonna get over that are you? I never said I had noise free. The only thing I said was I heard it to 5ft and didn't feel it would pose an issue in future. But after doing some testing with my ears the other night I feel it could actually pose an issue with a room I will be making big effort for a quite noise floor. I have no problem admitting I'm wrong. I am not here to get my ego stroked. I'm here to learn and better my knowledge base therefore I build a better room and system...

I was thinking having an XPA-5 with 95db SNR would give me the options of having a quieter amp if I absolutely needed to switch that to run LCR...
 

·
Registered
Joined
·
6,980 Posts
That will be tough for a general consumer to find out.
NAD M27, Crown DCi8|300, DSonic M3 ... which has the lowest noise floor?
Correct. My point was that SNR as typically specified will not give you a confident answer to the noise floor. It will take some digging with the companies beyond the sales department to someone with technical knowledge. Emotiva now has a 1W SNR listed which is rather telling for some of the amplifiers as at least you know it's consistent within their models. Once you sail past 94-98dB and on well past 100dB, these things become very important. A large HF horn run active can have a sensitivity in the [email protected]/1m range in our most sensitive hearing range. That can raise the noise floor by 20dB vs a more typical home audio speaker.
 

·
Registered
Joined
·
7,132 Posts
Correct. My point was that SNR as typically specified will not give you a confident answer to the noise floor. It will take some digging with the companies beyond the sales department to someone with technical knowledge. Emotiva now has a 1W SNR listed which is rather telling for some of the amplifiers as at least you know it's consistent within their models. Once you sail past 94-98dB and on well past 100dB, these things become very important. A large HF horn run active can have a sensitivity in the [email protected]/1m range in our most sensitive hearing range. That can raise the noise floor by 20dB vs a more typical home audio speaker.
So in your opinion is there nothing I can do for the DSi amps running active to quiet down?
 

·
Registered
Joined
·
3,543 Posts
Be careful with Emotiva. If you google Emotiva XPA noise you'll see lots of hiss issues. My XPA 3 was noisier than my xls1500. It was only an issue of course with speakers like the 3677, pro10, JBL 8340a etc. Some report no issues so it could be QC issues. Perhaps the reference line may be more of a sure thing.
 

·
Registered
Joined
·
7,132 Posts
Be careful with Emotiva. If you google Emotiva XPA noise you'll see lots of hiss issues. My XPA 3 was noisier than my xls1500. It was only an issue of course with speakers like the 3677, pro10, JBL 8340a etc. Some report no issues so it could be QC issues. Perhaps the reference line may be more of a sure thing.
I thought that was the first gen ones with higher fixed gain.
 

·
Registered
Joined
·
3,928 Posts
Having wrestled with amplifier noise floor a good deal for some very quiet home theaters, note the signal to noise is not really the spec you're looking for. It can suggest a lower noise floor, but the same SNR on 2 amplifiers with greater power means the higher power amplifier is likely to be noisier (but not certain!). The value you are really looking for is relative to 2.8V/1W or whatever nominal base line. In other words, with no input signal, how many mV of noise is at the output?
+1

There actually is no spec for noise floor and SNR doesn't provide that info although a higher SNR generally correlates to a lower noise floor. Rane explains how the SNR is calculated for those wanting to read further.

I gave the formula for converting to SNR to 1 W in this post so that amps can be easily compared:

To go along with this, manufacturers specify SNR at full rated power. To properly compare amps one should convert the SNR rating to 1 watt.

Sqrt(power*resistance) = voltage
SNR - [20*log(voltage/2.82)] = SNR at 1 watt

I type it into the calculator like in these examples below. You end up with a negative number, but just drop the minus.
Example:
Amp A: SNR 105dB @ 200 watts [(200*8)sqrt/2.82]log*20 - 105 = 82dB @ 1 watt

Amp B: SNR 110dB @ 1000 watts [(1000*8)sqrt/2.82]log*20 - 110 = 71dB @ 1 watt

As you can see, Amp A is 11 dB quieter than Amp B despite Amp B's higher SNR rating by the mfg.

To further complicate things, most of the time SNR is A-weighted. This weighting makes the measured output at 10 kHz down 2.5 dB from actual and by 20 kHz it is down by 9.3 dB.
 

·
Registered
Joined
·
3,543 Posts
I thought that was the first gen ones with higher fixed gain.
That may be the case. They have a return policy so other than shipping it would be a safe experiment. I'm sure you tried adjusting the gain/attenuator on the Crown. I had to turn down my QSC DCA to get it silent to the seats. It should attenuate the entire signal, noise included.
 

·
Registered
Joined
·
7,132 Posts
That may be the case. They have a return policy so other than shipping it would be a safe experiment. I'm sure you tried adjusting the gain/attenuator on the Crown. I had to turn down my QSC DCA to get it silent to the seats. It should attenuate the entire signal, noise included.
Gain doesn't matter. I only have DSi amps plugged in then speaker wire connected in active config. So one channel to low frequency section and one channel to high frequency section. Gain knobs can be all the way up or all the way down but noise stays the same...

I wonder if I should try a different XLR cable. I did make the one I'm using so perhaps I screwed it up...
 

·
Registered
Joined
·
3,543 Posts
I wonder if the input attenuators (that's really what those knobs are, often called gain knobs) are locked out internally or something. You should get quite a bit of difference form min to max... like several db.

For example I can turn the knobs down and have to be 2" from my 2150s (101db sensitivity) tweeter to hear any hiss. At max I can slightly hear it 12 feet away.
 

·
Registered
Joined
·
7,132 Posts
I wonder if the input attenuators (that's really what those knobs are, often called gain knobs) are locked out internally or something. You should get quite a bit of difference form min to max... like several db.

For example I can turn the knobs down and have to be 2" from my 2150s (101db sensitivity) tweeter to hear any hiss. At max I can slightly hear it 12 feet away.
I dont know. Any suggestions on what or how to test tonight?
 

·
Registered
Joined
·
3,543 Posts
I'm pretty sure something is up here. Maybe being in crossover mode they're locked out which would make sense as you could really mess up the response messing with them. You may be able to adjust internally or set one knob to attenuate both inputs.. just brian storming here.

It sounds like the signal is coming in full signal. Google amp hiss turn down gain (everyone mistakenly calls it gain). Also look up "gain matching" .


For example...

http://www.crownaudio.com/media/pdf/amps/137467.pdf

Just an example where even the manufacturer warns of noise with it all the way up. There's a thread on AVS explaining how to gain match using a volt meter.
 

·
Registered
Joined
·
3,928 Posts
The Emotiva is >101 db @ 1watt and >117db at full power

The XLS 1500 is >103 dB (rated as dBr to full rated 8 ohm power output)

So if the XLS 1500 is 103 db at full power then it is probably much less at 1w. If I am understanding this correctly, the Emotiva should be a large step up from the Crown. Am I reading this right?


I called Outlaw but they are not going to run any sales to compete with Emotiva and they charge for shipping.
Using the formula I provided earlier, the XLS 1500 is 78.2 SNR @ 1 watt.
 
  • Like
Reactions: Molon_Labe

·
Registered
Joined
·
7,132 Posts
I'm pretty sure something is up here. Maybe being in crossover mode they're locked out which would make sense as you could really mess up the response messing with them. You may be able to adjust internally or set one knob to attenuate both inputs.. just brian storming here.

It sounds like the signal is coming in full signal. Google amp hiss turn down gain (everyone mistakenly calls it gain). Also look up "gain matching" .


For example...

http://www.crownaudio.com/media/pdf/amps/137467.pdf

Just an example where even the manufacturer warns of noise with it all the way up. There's a thread on AVS explaining how to gain match using a volt meter.
Well I planned on using the process Jon posted weeks back about setting gain...Makes sense it could be set in bi amp mode. But I don't understand how turning the volume down would mess up FR. I don't see why if they warn of noise they would auto set gain at or near max in bi amp mode. If that is what is happening...

So I will try hooking one up passively when I get home...
 

·
Registered
Joined
·
6,980 Posts
So in your opinion is there nothing I can do for the DSi amps running active to quiet down?
I haven't followed all the posts and details, but assuming you are using the analog inputs I would start by connecting nothing but the power cord and the speaker cable. Disconnect any other wires and start playing with the settings. While less likely, some amps might pick up some noise from the power line if there is not a good ground or on a circuit shared somehow with some lighting dimmers in the room. Turn any lights full on or full off for testing. This should help you determine what your minimum noise floor is before adding other system connections.

I would try bypassing the active DSP settings and turning gains down to see if the noise changes. When I saw many talking about going active with these in the home I wondered if this would be an issue. The passive crossover will reduce the sensitivity of the horn some but without testing there's no info on how much. Most of the pro amps with internal DSP do have a higher noise floor, but the DSi amps are only rated at about 100dB S/N where some others are 4-12dB higher. As others noted, higher doesn't always guarantee less hiss, ultimately someone has to connect to a speaker and put an ear to it while playing with the gain & settings.
 

·
Registered
Joined
·
7,132 Posts
I haven't followed all the posts and details, but assuming you are using the analog inputs I would start by connecting nothing but the power cord and the speaker cable. Disconnect any other wires and start playing with the settings. While less likely, some amps might pick up some noise from the power line if there is not a good ground or on a circuit shared somehow with some lighting dimmers in the room. Turn any lights full on or full off for testing. This should help you determine what your minimum noise floor is before adding other system connections.

I would try bypassing the active DSP settings and turning gains down to see if the noise changes. When I saw many talking about going active with these in the home I wondered if this would be an issue. The passive crossover will reduce the sensitivity of the horn some but without testing there's no info on how much. Most of the pro amps with internal DSP do have a higher noise floor, but the DSi amps are only rated at about 100dB S/N where some others are 4-12dB higher. As others noted, higher doesn't always guarantee less hiss, ultimately someone has to connect to a speaker and put an ear to it while playing with the gain & settings.
Yea I will check with lights off. Different circuit too as I have moved the fat speakers :) The noise comes even when AMP is not connected to balanced outs on the Onkyo 5508...Leads me to believe its simply noise floor of CD running active like you said. It does seem when in bi amp mode it does fix the gain...

Jon did say this...And perhaps possibly the only solution...
"I think that I mentioned in the 4722 thread that if the amps are a little too noisy for the compression driver, an L-pad or autoformer can be used to reduce the sensitivity (and the noise) of the c.d. by 10db or so with no performance disadvantages.
It's actually why the M2 comes with a "crossover" even though it too is an active speaker--just to quiet down the compression driver/horn.
You will also lose some db with the acoustically transparent screen. Tough to say how much, as it depends on what you go with.

Here is a link to an L-pad:
http://www.sengpielaudio.com/calculator-Lpad.htm

It is just a couple of resistors.
They turn the excess power into heat. This is the way it is done to bring down the sensitivity of the compression driver to the woofers in a passive network.
An auto former increases the impedance of the compression driver, so the voltage sensitivity is reduced (and the noise), but with no heat. I'm not sure what device would be ideal as I haven't researched it that closely."
 

·
Premium Member
Joined
·
2,435 Posts
I am thinking of ditching using all externals and going with a Yamaha 3050. I can sell my unusued Dirac/88A and almost be half way there. I hate Audyssey but a lot of people seem to prefer YPAO since it cures the "wet blanket" mess of Audyssey. I would still need amps for Atmos, but the Crowns work fine for the SCS 8 with the gain turned down. I am fast approaching simple, set it and forget it. These long-chained setups are becoming too problematic quickly.
 

·
Registered
Joined
·
21,336 Posts
That will be tough for a general consumer to find out.
NAD M27, Crown DCi8|300, DSonic M3 ... which has the lowest noise floor?

the dsonic is easy. the specs indicate 55 uvrms (millionths of a volt rms "mu") residual noise. that is low. :) with 400 watts into 8 ohms, the voltage there is 56.6 volts. 55uvrms / 56.6vrms = 9.72 10^-7, which is -120.4 db. the dynamic range spec is 120db, so everything looks good.


nad specs 118db s/n at "rated power" which is 250 watts. 250 watts into 8 ohms is 44.7 volts. the spec is -118db. -118db is 1.26*10^-6, multiplied by 44.7 volts is 5.63 *10^-5 volts or 56.3 uvrms residual noise. that is low too. :)


the crown is going to lose this game. :) signal to noise is only -108db and rated power is 300 watts at 8 ohms. that is 49 volts. -108db is 4*10^-6 multiplied by 49 vrms = 1.95 *10^-4 volts or 195 uvrms residual noise. 195 uvrms is actually pretty good, but obviously not as good as the other two amps.


all three were specified using "a-weighting", so are comparable. a-weighting can provide an advantage in s/n vs "unweighted", so be careful when comparing.


note: to convert negative db voltage spl into volts use: 10^(-100/20) or whatever your s/n is if not -100. then just multiply by the voltage at the "rated power" of the amp. to find that simply take the power and multiply by the impedance then take the square root. so for 300 watts into 8 ohms. it is 300*8 = 2400, then square root of 2400 = 49.0 vrms.


note 2: the residual noise is the noise floor of the amplifier with inputs shorted and no signal present. one can always lower the noise floor with various forms of attenuation on the speaker, but that will come at the expense of headroom. however, in most cases, once you have about 100db of dynamic range, you are good to go in a home environment (i.e. a 20db noise floor and 120db peak output will satisfy all but a few folks.)
 
2561 - 2580 of 7704 Posts
Top