Is my amp clipping? Is my DAC too powrful for my amp?
Apr 10, 2012 at 12:15 PM Thread Starter Post #1 of 12

scannon18

100+ Head-Fier
Joined
Sep 3, 2010
Posts
455
Likes
12
I use the audio-gd nfb 3.1 dac, which has a line output level of 2.5 V, and the 3.1 feeds the Fiio E9, which I hear clips around 2.1 V.  However, I am not quite sure what clipping is.  I have noticed that on certain tracks with large dynamic range there is abrupt static at very loud portions of the song. 
 
Is this clipping?  Can this be remedied, or do I need a new amp to keep up with my DAC?  Or do I have this all wrong
blink.gif
?
 
Apr 10, 2012 at 12:22 PM Post #2 of 12
I know that can certainly be possible. My Music Streamer clipped the astro mixamp when I tried that combo out. I'm surprised something robust like the E9 is clipping from the dac, though.


Make sure you have EQ off.
 
Apr 11, 2012 at 7:24 PM Post #3 of 12
Can you describe what the clipping sounds like?
 
For me it is a reproducible static sound that occurs at the same time on the same tracks, regardless of volume.  At first I thought it was because the track was poorly mastered and my new dac was the only gear I had that was resolving enough to expose the static.  But having read here that the E9 clips after 2.1V (and I have no idea what that means) I now suspect that the amp might be clipping. I don't hear the static from the E7.
 
Apr 11, 2012 at 7:32 PM Post #4 of 12
is your dac 2.5v max or 2.5v rms? if it 2.5v rms that means it can be possibly overloading the inputs of the E9 cause i assume 2.1v is the amp's max input voltage it can handle.


have you tried hooking the e9 to something like a tv or computer directly? they output very little voltage so it's good way to test the e9. also hook up the dac directly to things to see if it's not the dac as well just in case.

 
Apr 11, 2012 at 7:51 PM Post #5 of 12
If the data on the E9 and the NFB 3.1 are right, then that's what you'd expect.  Why don't you just test it yourself?  Hook up the E9 output to the line in (or mic in) of a computer, play a 0 dBFS tone through the system, and record it.  You'd first want to turn down the volume on the E9 and set the levels reasonably of course.
 
You could do the recording in something like Audacity, which is free software.  Then you can graphically see if the signal is clipped or not.  You could also try this with music, but a test tone is probably easier to see and confirm.
 
 
By the way, if it really is clipping, all you'd have to do is lower the volume on the computer slightly.
 
Apr 12, 2012 at 4:06 AM Post #6 of 12
I have had that Fiio combo for a while myself, never ran into this. I don't think the voltage in and by itself would be responsible for the clipping... as somebody else pointed out, it may rather be at the source level, either the system volume, or your player's volume, or EQing, or a combination of all those.
 
The only time I personally have had clipping issues (as single clicks or noise as you describe depending on the length) in any of my various setups it has inevitably been EQing (before I learned here to always EQ down (when available) and if necessary raise the overall volume) The iDevice EQs for example have been terrible all around so I simply leave them off, esp. now that I know that the right headphones don't require any of that anyway...
 
Keep in mind that even if you lower your source volume, you can raise it on the E9, and in my experience it has been pretty good even up to a high volume, without the kind of distortion you describe.
 
Apr 13, 2012 at 5:06 PM Post #7 of 12
Quote:
I have had that Fiio combo for a while myself, never ran into this. I don't think the voltage in and by itself would be responsible for the clipping... as somebody else pointed out, it may rather be at the source level, either the system volume, or your player's volume, or EQing, or a combination of all those.

[emphasis added]
 
Well it can be, with respect to a given device.  It's a very well-understood phenomenon easily explained by examining the circuit.  If the voltage to a certain system is too high or low, the value will be clipped when passing through.  It's how electronics work.  In some systems, circuits are used to intentionally clip inputs to certain levels, if clipping is needed for some reason.
 
However, the rest of your sentence doesn't really make sense to me.  What we're interested in is the output voltage level of the source, and whether the amp can handle it.  For amps, it's more common that you could run into clipping from the gain being too high or the output not being strong enough, but it's also possible for an amp to clip an input voltage that's too high for it to handle—no matter what the volume is set to.
 
You're right that improperly-set EQ can easily lead to digital clipping though, and that's common.
 
Apr 13, 2012 at 5:38 PM Post #8 of 12
scannon18, simply play a full scale sine wave and compare it to a -6 dBFS sine wave. If the full scale sine wave sounds distorted (not clean) your amp is clipping the too hot input signal.
 
Apr 14, 2012 at 2:09 AM Post #9 of 12
Quote:
Originally Posted by mikeaj /img/forum/go_quote.gif
 
Well it can be, with respect to a given device.  It's a very well-understood phenomenon easily explained by examining the circuit.  If the voltage to a certain system is too high or low, the value will be clipped when passing through.  It's how electronics work.  In some systems, circuits are used to intentionally clip inputs to certain levels, if clipping is needed for some reason.
 
However, the rest of your sentence doesn't really make sense to me.  What we're interested in is the output voltage level of the source, and whether the amp can handle it.  For amps, it's more common that you could run into clipping from the gain being too high or the output not being strong enough, but it's also possible for an amp to clip an input voltage that's too high for it to handle—no matter what the volume is set to.
 
You're right that improperly-set EQ can easily lead to digital clipping though, and that's common.

I'm not an electronics expert, so that explanation will work for me. As for a clarification of the rest: while the first part was about (apparently misunderstood) electronics, the second was simply a list of practical reasons I have experienced distortion and/or clipping myself:
 
Usually, if the volume control was set relatively low on the amp, yet the actual volume in the headphones was relatively high and clipped/distorted, it was because somewhere along the input chain the signal got over-amplified, whether through the overall source system volume, or in combination with a high separate player software volume and/or EQing (even flat), or a mismatched gain setting, or, I guess, that voltage difference.
 
I still think that by keeping the source volume at an appropriate level, he should be able to continue using his combo without clipping even if there is a voltage difference. After all it only gets clipped when going past a certain limit, so if one keeps it below that by controlling the source volume, it should be OK.
 
(Which, after re-reading your other post, you pretty much wrote as well: "By the way, if it really is clipping, all you'd have to do is lower the volume on the computer slightly")
 
Apr 19, 2012 at 11:48 PM Post #10 of 12
I really appreciate the above advice.  Some of the suggestions were a bit too technical for me, but ultimately I was able to determine that the amp was clipping by playing a "static-y" portion of the song and replaying it with the computer volume at 70%.  At 100% there was static, at 70% no static.  Case solved.
 
Any resources explaining what causes this or what this is or how this all works would be appreciated.  As mentioned above, I don't even know what 2.5V RMS means
 
Apr 20, 2012 at 11:01 AM Post #11 of 12
Typically the DAC should be a line out.
This is defined at -10dBV or 0.316Vrms for consumer audio. IMO this should not cause the amp to clip.
If the output resistance of the DAC itself is high then there's a problem.
 
Apr 20, 2012 at 11:15 AM Post #12 of 12
Quote:
I really appreciate the above advice.  Some of the suggestions were a bit too technical for me, but ultimately I was able to determine that the amp was clipping by playing a "static-y" portion of the song and replaying it with the computer volume at 70%.  At 100% there was static, at 70% no static.  Case solved.
 
Any resources explaining what causes this or what this is or how this all works would be appreciated.  As mentioned above, I don't even know what 2.5V RMS means

 
Good that you got it sorted out okay.
 
2.5V is 2.5 volts.  RMS is root mean square, an important distinction if you're looking at some kind of AC signal like we are here.  Often times, the output voltage levels are given in terms of RMS values, mostly just for convention.  What actually causes clipping in this circumstance is exceeding the maximum allowed peak value.  The RMS value is never greater than the peak value, but it can be smaller.  It depends on the shape of the signal.  For a sine wave—and often for electronics testing, we're assuming sine wave inputs—the RMS voltage value is 1/sqrt(2) of the peak value.  A source listed at 2.5V RMS output is doing 2.5 * sqrt(2) = 3.536V peak, or 7.071V peak-to-peak (positive peak to negative peak).
 
http://en.wikipedia.org/wiki/Root_mean_square
http://en.wikipedia.org/wiki/Audio_power
 
In this situation, the amp's input circuitry cannot handle voltages over a certain level, so it will clip any values that get too high (or low on the negative side).
 

Users who are viewing this thread

Back
Top