iBasso D6 Volume Impression...where's the beef?

I'm glad I found this, because this was exactly my experience. I've been discussing the D6 performance in this thread:

http://www.head-fi.org/t/527472/ibasso-d6-fer-de-lance-usb-dac-amp   (the first post of mine is here: http://www.head-fi.org/t/527472/ibasso-d6-fer-de-lance-usb-dac-amp/675#post_7988175 )

After doing some measurements, it seemed to me that it was not getting anywhere near it's stated specifications of 650mW into 32 ohms, when used as a DAC. I was using 120 ohm headpones, and found that the output clipped at about 2.8V RMS, and if you calculate the power into 32 ohms, at this voltage, it's only 245mW.

Greg.

The other thing to note with the D6 is that the maximum gain of 10dB may not be sufficient for some portable devices. I.e, assuming it met it's specification of 650mW into 32 ohms (a voltage of 4.56V RMS), and assuming a portable device is being used with, say, a 0.5V RMS output, if you calculate the resultant output with a 10dB gain, it is 1.58V RMS, which is far less than the maximum specified output of 4.56V RMS, and it's not even reaching the maximum output that I actually measured at the onset of clipping (2.8V RMS)  Note that I realise the specs do not explicitly say 4.56V RMS - that is just my interpretation of the specs.

On the positive side, iBasso were willing to adjust the gain for me, before shipping it to me. (I didn't take them up on that offer though, reasoning that I will usually be using sensitive headphones with my portable device) Those with low output devices may want to explore that option, if you are considering buying a D6.

Greg.

@Justin,

Even when loaded with 120 ohm headphones, though, it clips at 2.8V RMS (at a volume level of about 4 o'clock), which is still far shy of 4.56V RMS, so I don't think that explains it. Completely unloaded (except for a digital multimeter), it reaches 3.6V RMS at max volume, however I think that this signal would be highly clipped, so it's not a true RMS figure, given that my DMM is probably an averaging meter. (it's certainly not a true RMS meter).  I don't have an oscilloscope, however I could have recorded this signal with another audio interface. I regret not doing that before returning the unit. 

I also do not understand yet how it could possibly reach 12.9V peak to peak (that's what 4.56V RMS is), given that the supply voltage is circa 12V, unless it generates a negative supply, or something like that.

Note that it is still possible that I made errors in my measurements.

If anyone is wondering, yes, I was using a sine wave frequency within the frequency range of my DMM. I measured another piece of equipment as a cross check, too.

Greg.

Just btw, it has been suggested to me that perhaps, when the unit is being used as a DAC, the output of the internal DAC chip is routed to the headphone amp, rather than the Line Out signal.  The first thing that iBasso said was that the DAC had an output of 1.2V, and if that's a peak voltage, it would tally nicely with my 2.8V RMS maximum clean output, at a 10dB gain. However, they then seemed to change their mind. They said the 1.2V was a mistake, and that the Line Out spec of 1.5V RMS is correct. 

Perhaps someone else here could take a measurement. Feed it with a full scale sine wave, at a low frequency (say, 100Hz, to 200Hz, check the specs of your multimeter) on the USB, and measure the AC RMS voltage of the output with the volume on 100%.  Any cheap DMM should be able to give a useful ballkpark measurement I think. You could then connect headphones (starting with a low volume, of course), and gradually turn up the volume, listening for distortion. (clipping). You'll want relatively INsensitive headphones for this test - take care not to damage your headphones or your hearing.

Greg.

Thinking carefully about how I did my testing, I think it's very possible that only one "side" of the amp was clipping. (i.e, the negative excursion of the waveform, or the positive excursion).  The reason I say this is that I remember being surprised at how high the level was, when I looked at the peak level meter on another audio interface that I connected it to, as part of the testing.  I've now done some testing to crudely calibrate the level meter I was using, and my best guess is that the peak voltage, with the D6 volume at max, may have been about 5.5V. Recall that I heard clipping at 2.8V RMS and above. 2.8V RMS = about 4V peak, so if the waveform was clipping at 4V, it should have remained at 4V peak as I increased the volume above the clipping point, with of course a wider flat-line clip as the volume was increased. I don't think this happened, because of the higher peak level I observed. I think the most likely explanation is that one side of the waveform remained at 4V, with the other half perhaps NOT clipping, and continuing to increase in amplitude as the volume was increased. If I'm right, I think it could mean my unit was defective, although it could just mean there is some asymmetry in the design.

The reason I had connected the D6 was initially to record the waveform, to make absolutely sure it was the D6 that was clipping, and not the headphones. However, I took a short cut - I simply connected the headphones to a monitor headphone output on this audio interface, with the volume down quite low, and I then listened to the D6 through this audio interface, isolated from my headphones.  As I could still hear the clipping, I concluded that this proved that it was the D6. I could see that the D6 was not overloading the input to this audio interface, because it has a peak meter. It must be the D6. As I say, in hindsight, I wish I had also recorded the waveform.

Greg.

According to iBasso, the "max" voltage is indeed 4.56V. However, we are at loggerheads over what this means. They insist that it is NOT peak, and it is NOT RMS. It is "max".  I'm going to assume for now that we have a language barrier, and they really do mean "peak".  They have also re-confirmed that the Line Out is 1.5V RMS, and they admit that it is possible to reach clipping if the volume is turned up enough. 

So, if it's a peak of 4.56V, that means the power spec is a peak of 650mW, and the RMS power (which I think is what we're really interested in) for a sine wave would then be 325mW into 32 ohms.   This still doesn't tally with my measured 2.8V RMS/4V peak measurement, although it's now in the ballpark.

I am trying to clarify with them exactly what they mean by "max".

Greg.

Sorry - having now re-read their email, I do believe they have confirmed that it is a peak voltage of 4.56V  They said it's less than the PEAK TO PEAK voltage, but more than the RMS voltage. This is quite clear to me the PEAK voltage (which is of course half of the PEAK TO PEAK voltage).

I'll now try to explain to them that I wasn't quite able to reach a peak/max voltage of 4.56V

EDIT: Considering that I was simply listening to the signal to determine the point of clipping, and that a peak of 4V is just 1dB less than 4.56V, I think it's entirely possible my unit was reaching 4.56V peak.

Greg.