[eio]

Nice. Good see some plots.

The crosstalk on the V20 is so flat it could be software, but it looks like a common resistance on the ground return. Most likely a layout issue. Crosstalk at -70dB and below is probably not a delibarate DSP feature.

To me I would guess the amp is the improvement. DACs don't have that much intermod

thanks for help analyzing! you are right this crosstalk doesn't make sense as a DSP feature.
 
it was tested with no load (lineout), i'll test with low/high resistance dummy load later.

 

[ljnew]

   
B&O branded (imho there is no tuning) korean version F800.
 
 
yes, V20 really is cleaner and smoother than V10.
 
i use SE846, it has tremendous sub bass amount & extension, paired with ultra low impedance = higher requirement on amp current and internal impedance than its sensitivity (114db) suggests.

Oh ok I really like the se846, but I wouldn't consider it to be an analytical iem like audio-technica ath-im01-3 or e70.  Also, I'm using the us version from tmobile so they may sound quite different.

 

[eio]

it seems the only way to get around SRC on V20, for now, is to play 24bit files with its built-in music app.
16bit files will be resampled to 16/48 even in built-in music app.
and everything will be resampled to 16/48 in poweramp (poweramp's hi-res output path does not work on V20 for now)
 
when it's going the resampled (mixer) path, volume control is also being done in digital domain, which makes ES9218's AVC feature useless.
 
since most materials are 16bit, the best option i've found is to use poweramp alpha, and switch to SoX resampler to use the best SRC, given that i have no practical way to get around it.

 

[eio]

  What your impression of the V20 with the Hi-Fi setting turned off?

pretty good for a bundled codec, i think it can be called hifi a few years ago...
 
but the hifi path of V20 is simply much better 

 

 

[eio]

RMAA continued.
 
20 ohms dummy load tested at 45 to 75 volume levels (75 is the maximum volume on V20)
 
V20's distortion start to rise above volume level 50 on 20 ohms load, looks like being soft-clipped, maybe due to current limit.


 
crosstalk:

 
sweep frequency IMD:

 

[SomeGuyDude]

Anyone else having an issue where the Quad DAC switch doesn't do a damn thing? I'm sitting here in Spotify flipping it on and off and there's zero change in the sound.
 
EDIT: And more than that, for some effing reason, the volume buttons only change the call volume, not the music I'm listening to. Plot twist... I'm not in a call. 

 

[eio]

 

  Anyone else having an issue where the Quad DAC switch doesn't do a damn thing? I'm sitting here in Spotify flipping it on and off and there's zero change in the sound.
EDIT: And more than that, for some effing reason, the volume buttons only change the call volume, not the music I'm listening to. Plot twist... I'm not in a call. 

 
yes, it seems like a UI issue that only reproduces occasionally.

 

[maxh22]

Anyone else having an issue where the Quad DAC switch doesn't do a damn thing? I'm sitting here in Spotify flipping it on and off and there's zero change in the sound.

EDIT: And more than that, for some effing reason, the volume buttons only change the call volume, not the music I'm listening to. Plot twist... I'm not in a call. 

That's weird. Did you make sure the headphone jack is plugged in all the way? Are you playing lossy or lossless files. I think i read somewhere that if the bitrate is of low quality it will not engage the quad dac.

 

[ZJee]

Anyone else having an issue where the Quad DAC switch doesn't do a damn thing? I'm sitting here in Spotify flipping it on and off and there's zero change in the sound.

EDIT: And more than that, for some effing reason, the volume buttons only change the call volume, not the music I'm listening to. Plot twist... I'm not in a call. 

Make sure that youve got the maximum quality you can get. Change the quality to "extreme" in the settings and that should hopefully improve something. I wouldnt be surprised if it didnt though. My v20 is fine but my v10 was riddled with lots of odd software bugs that made it downright frustrating to use. If it really doesnt work you may have to annoy LG until they fix it or give you a new phone.

 

[eio]

i'm trying to estimate a safe volume range for se846, from the safe volume level of 50 for 20 ohms load.
 
se846 is specified at 9 ohms, i'll use 5 ohms as my target since headphone is not a pure resistive load as the dummy load.
suppose the output voltage is constant in the volume range i care about, since I = V / R, with R reduced by 4x, i'll have to reduce V by 4x to keep I in safe range.
10 volume level difference in V20 roughly equals to 5.67db, that is 1.92x voltage difference. so i'll have to keep volume level below 29 = 50 - 10 * (4 / 1.92).
 
omg! that is in my daily listening volume range (20~30), i'm already pushing V20 near its limit with se846 

, and se846 has a sensitivity of 114db which is high. 

 
if i don't care about a little increase in distortion, i can push the volume to 44, which equals vol 65 into 20 ohms load, with -100db THD the sound quality is still reasonablely high.

 

[eio]

RMAA continued.
 
10 ohms dummy load tested at 35 to 75 volume levels (75 is the maximum volume on V20)
 
wow! it is massively overloaded at volume 75, spectacular! 

 

 

[eio]

RMAA continued.
 
100 ohms (detected as high impedance deivce) dummy load tested at 35 to 75 volume levels.
 
performs slightly worse than low gain mode.

 

[eio]

10 ohms (in high impedance mode by hot swap) dummy load tested at 15 to 75 volume levels.
 
it doesn't look like a good idea to override the auto detection...



 
summary of output voltage levels at different gain modes, measured by -3dB 1k tone with 10 ohms load:

 
 
summary of distortion performance in low gain mode:

(i tried to include high gain/aux mode in the same graph, but it seems high gain mode performs significantly worse than low gain mode for low impedance load even at the same output voltage level)

 

[eio]

try to measure the output impedance in low gain mode:
 
10 ohms vs 100 ohms load (all in low gain mode by hot swap), volume is set to 65 (to avoid the clipping at volume 75 for 10 ohms load), using 1k signal.
measured output level difference is 1.65db, that is 1.21x voltage difference.
 
that give us an estimation of 2.39 ohms output impedance.

 

[eio]

but it looks like a common resistance on the ground return. Most likely a layout issue.

agree, with crosstalk of -42.1db / -60.9db / -81.1db for 10 ohms / 100 ohms / unloaded, that would give us an estimated common resistance of 800~900 ohms, right?
 
seems too conductive to me...is this a severe layout issue?