Lavcat, thanks very much for posting. From your numbers, I would say you need at least a 72 watt amp to stay out of clipping at your Rossini setting. (was that a typical volume setting?)
With your 35 watt amp, you'll be clipping a few peaks, for sure, but you may not notice if the amp has a benign clipping behavior. That's how we can "get away with" smaller amps on classical than we really need. A few peaks clipped here or there might not be noticeable. But as other have noted, you can tell the difference if comparing it to a clean amp.
With your 35 watt amp, you'll be clipping a few peaks, for sure, but you may not notice if the amp has a benign clipping behavior. That's how we can "get away with" smaller amps on classical than we really need. A few peaks clipped here or there might not be noticeable. But as other have noted, you can tell the difference if comparing it to a clean amp.
For my typical after dark listening level I measured 0.518 V with the -12 dB 120 Hz test tone. This level is lower than I would like, but I live in an apartment. A normal during-the-afternoon listening level just now measured 2.888 V with the -12 dB test tone. Of course sometimes I'm tempted to turn things up. My poll result of 8.47 V was, as instructed, played "as loud as you ever do."
The speakers are B&W 801 and the amp is a Crown D-75A.
Even playing at the 8.47 V level, the distortion indicators (IOC) of the Crown never came on, ostensibly showing that distortion never went above 0.05%. I certainly did not hear clipping. Clipping bothers me quite a lot when I do hear it.
What to do? I suppose I could get another D-75A at some point in the future and run them in bridged mono. Building a high wattage amp at the moment is a bit beyond me, both financially and technically. I have a pair of large Denon amps, but one is broken, and I don't think they sound as good as the Crown.
The speakers are B&W 801 and the amp is a Crown D-75A.
Even playing at the 8.47 V level, the distortion indicators (IOC) of the Crown never came on, ostensibly showing that distortion never went above 0.05%. I certainly did not hear clipping. Clipping bothers me quite a lot when I do hear it.
What to do? I suppose I could get another D-75A at some point in the future and run them in bridged mono. Building a high wattage amp at the moment is a bit beyond me, both financially and technically. I have a pair of large Denon amps, but one is broken, and I don't think they sound as good as the Crown.
No, because it won't drive a 1 ohm load. 
a.wayne likes low impedance speakers.
Honestly, if your typical listening volume setting gave you a 2.88V reading on the test tone, you won't be clipping the amp at all. You have plenty of headroom left. To look at it another way, at that setting you need 11.5V peak. Your Crown can supply about 25 V peak. No worries.
I don't know why the distortion lights didn't come on, but it make take a sustained clip to get them to come on. On very dynamic material those peaks will be brief.
a.wayne likes low impedance speakers.Honestly, if your typical listening volume setting gave you a 2.88V reading on the test tone, you won't be clipping the amp at all. You have plenty of headroom left. To look at it another way, at that setting you need 11.5V peak. Your Crown can supply about 25 V peak. No worries.
I don't know why the distortion lights didn't come on, but it make take a sustained clip to get them to come on. On very dynamic material those peaks will be brief.
If you look back thru the thread, you'll see that I've stated a few times why no 0dB file. As long as you know that the test tones have an RMS value of -12dB full scale (which is 9dB below the loudest sine possible) then you can do the math.
There are a number of softwares that will let you make tones. Audacity and Goldwave are easy to work with. A sine wave with peaks at digital full scale will have an RMS value of -3dBFS. The test tones I provided have a peak value 9dB below full scale and an RMS value 12dB below full scale.
There are a number of softwares that will let you make tones. Audacity and Goldwave are easy to work with. A sine wave with peaks at digital full scale will have an RMS value of -3dBFS. The test tones I provided have a peak value 9dB below full scale and an RMS value 12dB below full scale.
interesting test idea. 1 question: how are you converting from Volts average to RMS...unless im mistaken, few people are actually measuring RMS, peak conversion is rather problematic, and innaccurate. Aside from collecting relative voltages, i feel its a giant leap to try and use relative data for any real analysis. I use a 'good' DMM, and that does measure RMS, not by chance but by Fluke. Nice idea, gleaning anything useful from the data may be; troublesome.
peak to rms for a sinewave is exactly sqrt(2) factor.
There is no inaccuracy involved.
Average to rms is a simple set up. Read a known sinewave and adjust the average reading meter to output exactly the true rms value.
If you have an rms reading meter, then feeding a sinewave in, should give a reasonably accurate output.
There is no inaccuracy involved.
Average to rms is a simple set up. Read a known sinewave and adjust the average reading meter to output exactly the true rms value.
If you have an rms reading meter, then feeding a sinewave in, should give a reasonably accurate output.
Last edited:
andrew, i agree. Cheap meters read average and not RMS, i work closely with a 2libration eng, and average is not RMS, unless zero harms are present. Im well enough versed in fourier to understand the point of the test. The assumption is made that there is no THD produced anywhere in the signal chain. Close enough to truth at low levels, but less so at high levels, depending on the users equipment. Speaking from my own experience, i know that with even a 1kohm load i can make my 30watt amp clip on voltage at levels which, until i tried, were a shock to me.
Last edited:
Yes, we are making some assumptions for the sake of ease in testing. It isn't meant to be a rigorous, 1 PPM test . Just a close idea of what voltages you use.
The reason I chose 120 Hz as a test tone is that all the meters I used gave very accurate results on a 120 Hz sine wave. All the DVMs that had the appropriate AC volts scale, that is. If the meter is picking up some THD then it might be slightly overestimating the voltage, right?
It would be nice to know about an amp clipping into a high impedance load, as some here have measured with the speakers disconnected. I've never seen a solid state amp do this, and I would not run a tube amp with no load. Can you tell us more about your case?
. Just a close idea of what voltages you use.The reason I chose 120 Hz as a test tone is that all the meters I used gave very accurate results on a 120 Hz sine wave. All the DVMs that had the appropriate AC volts scale, that is. If the meter is picking up some THD then it might be slightly overestimating the voltage, right?
It would be nice to know about an amp clipping into a high impedance load, as some here have measured with the speakers disconnected. I've never seen a solid state amp do this, and I would not run a tube amp with no load. Can you tell us more about your case?
ok so maybe i was a little OTT with the accuracy bit The amp i use is 25 to 30W. Ive had it in pieces before, and i know the PSU to be 18V winding, so around 24VDC supply rail. Im not an amp designer but the clip occurred at low op volts, due to the very high impedance load and trying to push volts. After this i use 10R or 100R for ts measures. Very extreme case IMHO. FWIW i use a DMM with 200mV AC V and A ranges, and even this 'measures' RMS, ie calc'd from average, 40hz to 400hz -3db. A deal better than many DMMs ive used. Despite this I still found it necessary to calibrate the DMM with the span of frequencies, for doing a simple Z plot and TS calcs. Maybe im picking? Give me a moving coil meter and id have more faith in it
The amp i use is 25 to 30W. Ive had it in pieces before, and i know the PSU to be 18V winding, so around 24VDC supply rail. Im not an amp designer but the clip occurred at low op volts, due to the very high impedance load and trying to push volts. After this i use 10R or 100R for ts measures. Very extreme case IMHO. FWIW i use a DMM with 200mV AC V and A ranges, and even this 'measures' RMS, ie calc'd from average, 40hz to 400hz -3db. A deal better than many DMMs ive used. Despite this I still found it necessary to calibrate the DMM with the span of frequencies, for doing a simple Z plot and TS calcs. Maybe im picking? Give me a moving coil meter and id have more faith in it
No, just never liked the sound of Croewn amplfiers and i have had a few over the years...The 801 on stands is a fine sounding speaker ...
Have to admit, i did try ...No, because it won't drive a 1 ohm load.
also on more benign loads 8/4 ... I dont believe the 801's are 8 ohm speakers Pano and if he is using 2.88V where are the 12db peaks coming from , also the 801's need sustained current for the bass to sound right, that Crown is far from it ...
If he is using 2.88v and the 801's are not 8ohm, they are zmin of 5ohm at the top end , 6 in the bass.....
Last edited:
So they are 6 ohm speakers, not a big deal. Yes, that means more current needed than a nominal 8 ohm load, but his voltage needs stand.
As to the suitability of the Crown I can't speak. According to its specs it should be able to deliver about 3.7 amps (RMS) into a 4 ohm load. Using the volume setting that gave the 2.88V reading of the tone, the 801s should need about 2 amps, peak. How gracefully the Crown does that, I don't know.
As to the suitability of the Crown I can't speak. According to its specs it should be able to deliver about 3.7 amps (RMS) into a 4 ohm load. Using the volume setting that gave the 2.88V reading of the tone, the 801s should need about 2 amps, peak. How gracefully the Crown does that, I don't know.
The only real bit about this test is how to relate to real non steady state reproduction. It may sound strange but i cant listen to test tones as loudly as music. Pano, if i am correct, you are saying the tones are at -3dBfs PEAK? does it then follow that media played from the same device will have this same amplitude? Perhaps many recordings are at -3dBfs also, and i may be picking, humour me. Should the recording be at 0dBfs, or in the case of older stuff at lower levels. Im struggling to put my point across...surely you need to adjust the amp powers by +/-3dB to deal with the difference in level?
No. All of this has been discussed before, over and over. But I'm not above repeating myself.
The tones I provided has an RMS value of 12dB below Full Scale. That means their peaks are at 9dB below full scale. As Andrew pointed out above, that's the RMS to peak value of a sine wave 3dB.
When doing the measurements it does not matter what level music is recorded at. Because we know that the test tones have a peak exactly 9dB below the maximum possible. No music, no matter how loud, can have peaks above this. For some reason this seems to be very difficult for people to grasp. In digital, there is an absolute maximum. We know what that maximum is - it's 0dB. Everything else must be below that.
You DO NOT use the tones to set your level. You use music of your own choosing. You use the tones only to get the voltage reading. They bear no direct relation to how loudly of softly any music has been recorded. A very loud recording might be at -12dB average or even hotter. Many are at -18dB or -22dB. The test tone is not related to that.
The tones I provided has an RMS value of 12dB below Full Scale. That means their peaks are at 9dB below full scale. As Andrew pointed out above, that's the RMS to peak value of a sine wave 3dB.
When doing the measurements it does not matter what level music is recorded at. Because we know that the test tones have a peak exactly 9dB below the maximum possible. No music, no matter how loud, can have peaks above this. For some reason this seems to be very difficult for people to grasp. In digital, there is an absolute maximum. We know what that maximum is - it's 0dB. Everything else must be below that.
You DO NOT use the tones to set your level. You use music of your own choosing. You use the tones only to get the voltage reading. They bear no direct relation to how loudly of softly any music has been recorded. A very loud recording might be at -12dB average or even hotter. Many are at -18dB or -22dB. The test tone is not related to that.
Ok that answered my concerns fully Pano, no offence meant but im about 600 posts late to the party, and its tough going reading it all!
FWIW i just tested (22:00 here, top floor apartment...wooden floors...)
120hz I get 0.791 V on 2 Volt scale
220hz I get 0.792 V on 2 Volt scale also.
On mid 80's dB speakers too.....expected to see some impedance rise on the 120 due to TQWT with no stuffing showing 3rd Harmonic peak.
FWIW i just tested (22:00 here, top floor apartment...wooden floors...)
120hz I get 0.791 V on 2 Volt scale
220hz I get 0.792 V on 2 Volt scale also.
On mid 80's dB speakers too.....expected to see some impedance rise on the 120 due to TQWT with no stuffing showing 3rd Harmonic peak.