Your meter might base the reading on the assumption of a sine-wave, whilst the noise is likely to be much more random. However as a comparison it is probably fair enough IMO, since many manufacturer's often don't specify their method of deriving RMS noise figures anyway.
With such caveats, the SPL that your voltage would result in is fairly easy to calculate, according to the speaker sensitivity. Though notice many drivers/speakers specify SPL at 1watt, rather than a particular voltage, so their sensitivity to a given voltage will also depend on their impedance.
That said, none of this involves perception. A-weighting can help to some degree if offered, but there are still many ways in which such simple measure might appear similar between amps and yet not sound similar. Such as peaks vs RMS and the frequency at which the noise presents itself. Probably the waveform too; I'm quite surprised at how the noise from my class-D amps is so very different to AB amps in how it sounds, even accounting for the level.
With such caveats, the SPL that your voltage would result in is fairly easy to calculate, according to the speaker sensitivity. Though notice many drivers/speakers specify SPL at 1watt, rather than a particular voltage, so their sensitivity to a given voltage will also depend on their impedance.
That said, none of this involves perception. A-weighting can help to some degree if offered, but there are still many ways in which such simple measure might appear similar between amps and yet not sound similar. Such as peaks vs RMS and the frequency at which the noise presents itself. Probably the waveform too; I'm quite surprised at how the noise from my class-D amps is so very different to AB amps in how it sounds, even accounting for the level.
Feel free to check my work. I previously posted to this thread a spreadsheet of dB vs voltage as I understand it.
I'm trying to understand if the measurement I got is in a reasonable range for a class A/B amplifier with a switching power supply (i.e. Emotiva XPA Gen3). As I commented previously, reviewers love to discuss dynamic range of amplifiers, but I haven't heard them mention this noise floor with no signal present.
On my amp I measured 15mV, so 15,000µv, and ICEpower is claiming 20-80µV. If my measurement and their ratings are both correct, then I should have built my own amp long ago. But the difference is so large that I'm tempted to think something is wrong with either my measurement or their ratings.
And if a good class D amp really has a noise floor that is ~45dB lower than an A/B amp, why isn't that making the headlines? I would have built my own amp in the beginning, if I knew it would be so much quieter!
Noise and horns. I just worked on a tube based preamp for an unacceptable hum level when driven into Khorns (yah, 105Db). In my system the hum as it were, wasn't an issue, but then we are now using speakers with maybe 89Db sensitivity. What a difference. While I really love the sound of some horns, this noise issue is almost always an issue. Too bad that you can't simply buy an amp with a return guarantee just to listen and see if you are satisfied. Perceived/actual noise from one amp to another may be the difference that you can be happy with.
There are excellent class D stereo amps available, but they come with an impressive price. I've seen some close to $5k.
But those won't work for me, because I have custom 4 way active speakers. I need an 8 channel amp, and 4 of those channels need to be especially low noise. And of course I want to do it all for the same price or less than the Emotiva amp I already have.
Assuming that ICEpower's specs are accurate, this is what seems like the best solution for my specific 8 channel amp requirements. Here I am assuming that all drivers should achieve the same SPL, and I am factoring in their sensitivity to determine how many watts is needed.
In the SPL Vs Power section, the highlighted number is the rated amplifier output for each amplifier, except for the horn. The amplifier I list for the horns is actually 25 watts into 8 ohms x2, but it only requires 8 watts to match the output of the other drivers. So I have some room to pad it down. This should guarantee zero or near zero audible hiss.
The amplifier modules themselves will cost just under $1400. I would anticipate spending a few hundred more, but I'm starting to feel like I understand what I'm getting into, and that it'll work well as a complete system.
In the SPL Vs Power section, the highlighted number is the rated amplifier output for each amplifier, except for the horn. The amplifier I list for the horns is actually 25 watts into 8 ohms x2, but it only requires 8 watts to match the output of the other drivers. So I have some room to pad it down. This should guarantee zero or near zero audible hiss.
The amplifier modules themselves will cost just under $1400. I would anticipate spending a few hundred more, but I'm starting to feel like I understand what I'm getting into, and that it'll work well as a complete system.
Maybe you noticed the SPL numbers?
I am not listening at 116dB, although achieving THX low frequency reference levels with a stereo pair of floor standers is something that makes me smile.
Distance is ~10' right now, and it is an open floor plan, so kind of medium size.
15mV s quite a lot. When you were measuring, did you have anything connected to the amps input? If so, try to remove it and if you are really serious, you should short the input before you measure the output.
15mV could also be 60Hz hum... which is not you usual "noise"... check out for that... Could be a e.g. ground loop...
Goo luck!
//
15mV could also be 60Hz hum... which is not you usual "noise"... check out for that... Could be a e.g. ground loop...
Goo luck!
//
I have gone through the usual culprits without any success in the past. I'm using XLR connectors, so interconnects aren't creating noise. My amp does not have a transformer, so there is no 60hz hum.
But on your suggestion I decided to power off devices up stream of the amp, and sure enough, my budget signal processor is producing a bunch of noise.
That being said, I still need an aggressive L-Pad on my speakers to kill the amplifier's own noise. I think it is -20dB. With the signal processor off, the amount of hissing caused by the amp is still audible, but acceptible. Turning on the procesor makes it worse
I keep looking for a better quality processor, that doesn't cost $6k. I have a Peavey 48vsx...pure chinesium. It even has a button with chinese characters to change the UI to chinese! It works, but I didn't realize it was making the noise worse. I'm surprised I didn't try isolating it earlier. I had a Behringer graphic EQ that was really noisy once, so I should have known...
That being said, I'm not seeing a real chane in the mV on my meter. I suspect it isn't good at this type of measurement. I'd probably need a oscilloscope.
I mentioned previously that my amp is a "class H" amp, which means it has a switching power supply...so no transformer to cause 60hz hum.
I have the 2X150, 2X300, and the dual mono. They a great little amp but I suspect the have some noise on sensitive speakers, when I get home I will hook one up and put my ear next to my XRK open baffles . My friend bought the Kef LS50 they are biamped and they use a A/B amp for the tweeter and D for the woofer, that would be the way I would go. I built the Wolverine amp it is a very quiet class A/B amp but may be more power than you would need. Ostripper Pete may have some lower wattage designs.
I can definitely suggest trying out a second hand Extron XPA1002 from Ebay. Set their attenuation properly and there is virtually no hiss - and the character of the leftover hiss when listening near the tweeter feels more "nice" than of any other amp I have. I have 4 units and will buy more. After I installed these, the most prominent source of hiss is now the DSP processor, which will be replaced some day too.
Just make sure you get the low impedance version, not the 70 V or 100 V output ones.
Just make sure you get the low impedance version, not the 70 V or 100 V output ones.
The built in attenuator is interesting. I haven't heard of these. Do you think the amps themselves are quieter, or does the attenuation make them seem quieter by reducing input noise?
I was unclear there - my DSP is definitely producing more noise than the amp, so I set the attenuation at the input to reduce the noise from the DSP and the source, the amp with disconnected inputs produces just a little bit of noise regardless of the attenuation setting. I stop hearing any hiss from the amp itself in a pretty quiet room with my ear around 60 cm away from the mouth of the horn where D.A.S. M50 is used (rated at 107 dB/W.m), ear pointed directly at the driver. And as I wrote, the hiss character is sort of more pleasant than of other sources. I really like these amps, the only downside might be the non standard connectors they use, requiring custom wiring.
Edit: This is the manufacturer data, in this case I would really believe them based on my experience:
Frequency response 20 Hz to 20 kHz, ±1 dB THD + Noise 0.05% @ 20 Hz - 20 kHz, 8 ohms, at 3 dB below clipping S/N 105 dB, 20 Hz - 20 kHz, unweighted
There is a review on ASR as well https://www.audiosciencereview.com/.../extron-xpa-1002-plus-amplifier-review.35557/
Edit: This is the manufacturer data, in this case I would really believe them based on my experience:
Frequency response 20 Hz to 20 kHz, ±1 dB THD + Noise 0.05% @ 20 Hz - 20 kHz, 8 ohms, at 3 dB below clipping S/N 105 dB, 20 Hz - 20 kHz, unweighted
There is a review on ASR as well https://www.audiosciencereview.com/.../extron-xpa-1002-plus-amplifier-review.35557/
I was given a schooling in math over at the ASR forum. It's actually really helpful to understand the relationship between the different characteristics, and especially how to derive specs that are not directly stated to compare products directly.
https://audiosciencereview.com/forum/index.php?threads/minidsp-flex-htx.49512/post-1906197
Through this explanation we are able to determine that my signal processor is quite noisy, and it is made worse by the high gain of my amplifier.
Based on Michael's fromulas (From ASR thread), I was able to determine that my Emotiva Amp has ~80µV residual noise, which is on par with many ICEpower modules, even while the Emotiva has higher gain. So it is pretty good.
So the plan. I need to get the quietest components and reduce the number of components, to reduce noise up stream from the amp. I think this means I'm buying the new MiniDSP HTx processor.
And then I need to get a lower gain amp that is also low noise, something like the 50ASX from ICEpower. But most of the improvement will come from the MiniDSP, so I'll start with that.
https://audiosciencereview.com/forum/index.php?threads/minidsp-flex-htx.49512/post-1906197
Through this explanation we are able to determine that my signal processor is quite noisy, and it is made worse by the high gain of my amplifier.
Based on Michael's fromulas (From ASR thread), I was able to determine that my Emotiva Amp has ~80µV residual noise, which is on par with many ICEpower modules, even while the Emotiva has higher gain. So it is pretty good.
So the plan. I need to get the quietest components and reduce the number of components, to reduce noise up stream from the amp. I think this means I'm buying the new MiniDSP HTx processor.
And then I need to get a lower gain amp that is also low noise, something like the 50ASX from ICEpower. But most of the improvement will come from the MiniDSP, so I'll start with that.
I have the same problem as you did before. I was using the Heil AMT before and the efficiency on ESS is 94 dB IIRC and I use Behringer CX3400 as crossover to fed Icepower 1200AS2.
The 1200AS2 itself has 45uV un-weighted noise but I got quite noticeable hiss from the tweeter that's due to the 26dB gain that amplify CX3400's noise. Someone suggested me to get an signal attenuator before feeding into the 1200AS2 and it simply solved the problem.
Now I switched to the DCX464-16 and it has 112dB/W sensitivity. The source I'm using now is MOTU M4. With this combination I don't have hiss problem even at desktop monitor condition.
But it's just for testing. I don't need such high output for my compression driver.
Now I power my DCX with headphone amplifier and get enough volume.
Perhaps you can try the attenuator if the sound quality is already good for the original source.
Attenuator from Shure
The one I used before is very cheap just around $10 each.
The 1200AS2 itself has 45uV un-weighted noise but I got quite noticeable hiss from the tweeter that's due to the 26dB gain that amplify CX3400's noise. Someone suggested me to get an signal attenuator before feeding into the 1200AS2 and it simply solved the problem.
Now I switched to the DCX464-16 and it has 112dB/W sensitivity. The source I'm using now is MOTU M4. With this combination I don't have hiss problem even at desktop monitor condition.
But it's just for testing. I don't need such high output for my compression driver.
Now I power my DCX with headphone amplifier and get enough volume.
Perhaps you can try the attenuator if the sound quality is already good for the original source.
Attenuator from Shure
The one I used before is very cheap just around $10 each.