When would headphones generate clipped sound because of insufficient power?

[duyu389]

We know that when the source doesn't have enough voltage and/or current to meet the peak demands of the music, it clips off the peaks.

I made an experiment with my hd600.

Plugged into Audient ID14 MKII, played 1khz sinewave at max volume while monitoring the sound generated by hd600 with an oscilloscope app on iPhone named sonic tools, an true-rms digital multimeter was connected to the cable of hd600.

I looked up Julian Krause's database and founded out that the headphone output power of ID14 is lower when using Type-A USB power supply.

When using Type-A USB power supply, the oscilloscope showed obvious harmonic waves and I clearly heard distortion from headphone. Volume was 107 db, voltage was 2.7 Vrms.
When using Type-C USB power supply, the oscilloscope showed no obvious harmonic wave. Volume was 114 db, voltage was 4.3 Vrms.

Then I plugged into lightning to 3.5mm adapter which provide only 0.6vrms voltage into 300 ohm, volume was only 96 db but no clipped.
Then I plugged into Macbook(M1) Pro which provide 3vrms voltage into 300 ohm, volume was 108 db but no clipped.

If the clipping occur when the source doesn't have enough voltage, the lightning to 3.5mm adapter's power is definitely insufficient but the headphone generate wave with no distortion.

So I wonder when would clipping exactly occur with insufficient power? When the amp thought it can provide enough power but with insufficient power supply? I am so confused and expect for a reasonable explanation. Thanks a lot!

 

[duyu389]

why blocked

 

[oleg87]

If the clipping occur when the source doesn't have enough voltage, the lightning to 3.5mm adapter's power is definitely insufficient but the headphone generate wave with no distortion.

If the dac/amp can hit its voltage limits without distorting into a given load, you'll just run out of volume.

 

[solderdude]

Any amplifier will be able to generate a 'clean' output voltage right up to the point where it hits the internal power supply voltage rails (or just below it)
It thus solely depends on the internal power supply rail voltages of the amp section.
In USB-C mode (which can supply more current than USB-A) there is a higher internal rail-voltage possible (more power for the converter) so the output power can be higher as well.

 

[DVDdoug]

If the clipping occur when the source doesn't have enough voltage,clipping

That's true. And if it can't put out the necessary current, the voltage will drop (Ohm's Law).

It's a combination of the power supply and the amplifier. There is some voltage loss in in the amplifier. Some amplifiers tend to soft-clip before hitting the hard limit. And some amplifiers can put-out a short-term peak before the power supply voltage sags and it starts clipping with a sustained signal. (I wouldn't expect the power supply in a headphone amp to sag.)

When using Type-C USB power supply, the oscilloscope showed no obvious harmonic wave. Volume was 114 db, voltage was 4.3 Vrms.

You shouldn't see a "harmonic wave". You should see clipping, which yes, does mean there are harmonics.

USB-C might be capable of more current and the Audient MIGHT have a voltage-boosting power supply.*

With 5VDC you can get 5V peak-to-peak which is a 1.75V sine wave (with no loss through the amplifier). Or a you can get a 2.5V RMS square wave. You can double that with a bridge amplifier, but then you can't have a common ground between left & right so you'd need balanced (4-wire) headphones.




* High power car amplifiers have voltage boosters, otherwise they are limited by the 12V.

 

[duyu389]

If the dac/amp can hit its voltage limits without distorting into a given load, you'll just run out of volume.

Yeah that's true. But I wonder when running out of volume, under which kind of condition would sound get clipped. If 1khz 0dbfs sine wave did not get clipped through lightning to 3.5mm adapter, does that mean any kind of sound won't get clipped?

 

[staticV3]

@duyu389 Apple artificially limits the output of their dongle so that it's impossible for them to clip above 20Ω load impedance.

 

[solderdude]

It just means that the lightning adapter can not produce an output voltage (0dB FS) that reaches the clipping point of the output stage so you cannot clip it (with high impedance headphones, you can with very low impedance headphones.

The Audient headphone out can reach clipping levels.

 

[duyu389]

You shouldn't see a "harmonic wave". You should see clipping, which yes, does mean there are harmonics.

Sorry for my mistake of calling spectrum a "scope". Your explanation of voltage-boosting power supply definitely expand my knowledge of audio gear.

 

[duyu389]

@duyu389 Apple artificially limits the output of their dongle so that it's impossible for them to clip above 20Ω load impedance.

Thanks for explanation. I have never considered that there are so many hidden tiny tech specifics of a little cute adapter.

 

[solderdude]

Using DC/DC converters is very common in battery and 5V fed audio devices.
The 48V phantom voltage is also generated from the 5V input voltage. As that draws virtually no current you can do this with USB-A aleady.
The same is done for the headphone out.
In USB-C more current can be drawn and thus more power is available for creating more output power.

The hidden tiny tech is simply the max. output voltage the DAC chip can provide (in high impedance headphones). There is no magic.

 

[duyu389]

It just means that the lightning adapter can not produce an output voltage (0dB FS) that reaches the clipping point of the output stage so you cannot clip it (with high impedance headphones, you can with very low impedance headphones.

The Audient headphone out can reach clipping levels.

That sound really clear to blow cloud out of mind. Never thought of so many difference between devices. Thank you for your patience.
I usually listen to some heavily compressed mainstream music under 80 db with my 300 ohms headphones through stuff like apple adapter and everything seems to be fine. Some day I browsed excellent website like diy audio heaven and nwavguy's blog then suddenly found out the "clipping" statement really crushed my faith. Just a false alarm.

 

[solderdude]

Clipping occurs when a limit is reached (usually internal power supply rail voltage which can be different from the power supply voltage connected to the device).
You can also clip a 'digital signal' by the way. You can make the dongle sound heavily clipped yet.... the output stage of the dongle is not clipping at that point.
There are many 'squashed to death' recordings that show clipping of the signal yet the DAC itself does not clip.

 

[AwesomeSauce2015]

I responded to your other post, and just saw this so I figured I would restate a bit of what I said, for posterity's sake.
My big post:
https://audiosciencereview.com/foru...age-first”-get-me-confused.51667/post-1861350

In your first test you are seeing the device - to - device variation. Analog circuitry is not a simple beast, and there are may ways to design a circuit to meet specifications.

You noted that the apple hardware did not show a clipping signal on your scope. This is probably because apple designs their headphone circuits to not clip the signal. They can do this by limiting the user's volume control such that the user cannot increase the volume into the clipping region. I am not surprised at hearing this since apple clearly has some talented audio engineers on staff.

The other device you tested showed a clipping signal on one supply and not on the other. This is probably because the company designed that device to use higher USB-C supply voltages (since USB-C is capable of exceeding 5 volts, somehow). They would have designed the maximum user-accessible volume for the higher USB-C voltage, and when you use the lower USB-A voltage, then it would clip due to hitting its voltage limit.

But the simple answer is that: Clipping can occur anywhere in the signal chain. It can happen in the recording studio, in the digital domain (Spotify, windows, etc), it can occur in your DAC / Amp / speakers, and even (sort of) in your ears and the air. However, if all of these are managed properly, then you don't really have to worry about it.

So in conclusion, just listen to it, and if it doesn't get loud enough, or you hear obvious distortion, then go looking for solutions. But if it works, then don't bother looking for problems where there aren't any.

 

[duyu389]

I responded to your other post, and just saw this so I figured I would restate a bit of what I said, for posterity's sake.
My big post:
https://audiosciencereview.com/forum/index.php?threads/“some-amps-run-out-of-current-first-and-some-run-out-of-voltage-first”-get-me-confused.51667/post-1861350

In your first test you are seeing the device - to - device variation. Analog circuitry is not a simple beast, and there are may ways to design a circuit to meet specifications.

You noted that the apple hardware did not show a clipping signal on your scope. This is probably because apple designs their headphone circuits to not clip the signal. They can do this by limiting the user's volume control such that the user cannot increase the volume into the clipping region. I am not surprised at hearing this since apple clearly has some talented audio engineers on staff.

The other device you tested showed a clipping signal on one supply and not on the other. This is probably because the company designed that device to use higher USB-C supply voltages (since USB-C is capable of exceeding 5 volts, somehow). They would have designed the maximum user-accessible volume for the higher USB-C voltage, and when you use the lower USB-A voltage, then it would clip due to hitting its voltage limit.

But the simple answer is that: Clipping can occur anywhere in the signal chain. It can happen in the recording studio, in the digital domain (Spotify, windows, etc), it can occur in your DAC / Amp / speakers, and even (sort of) in your ears and the air. However, if all of these are managed properly, then you don't really have to worry about it.

So in conclusion, just listen to it, and if it doesn't get loud enough, or you hear obvious distortion, then go looking for solutions. But if it works, then don't bother looking for problems where there aren't any.

Thank you very much for your patience! Now that my opinion on this matter is gradually clear, I hope to be as objective and knowledgeable geek as you are!