• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

I feel bass difference between headphone amps. How to check phase mismatch?

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
Before measurements, I sensed bass difference between Topping L30 and SMSL HO200.

L30 seems "cleaner" than HO200 in bass "feels".

After voltage matched, I still feel the same.

Voltage matched done:
*Matched amp's left channel within 0.001V.
*Made sure left and right channel balance of each amp are within 0.001V.
*Made sure each amp is not effected by ground loop and 60hz tone voltage output are stable.

Chain is laptop USB to Topping E30. E30 RCA to amp. Amp to AKG K371.

Song used to compare is this:

I listen to the bass guitar tones starting from 46 seconds of the song. I keep looping back to that point to do listening comparisons.

For my downstairs 2.2 stereo setup, I need to adjust my subs' phase to my speakers to get cleaner bass notes "feels".

Could SMSL HO200 amp has slight phase mismatch between right and left channels in bass freqs, thus it is not as clean sounding as L30?

How do I check for left and right bass notes phase mismatch?

Thanks!
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,523
Likes
37,054
Maybe a crude method only showing significant phase mismatches. Put the headphones so each earcup touches the other. Playing bass from one channel should result in some bass you can hear, while adding a 2nd channel should cancel out almost completely the first channel as face to face the headphone drivers are out of phase.

Another method, use a voltmeter to check AC voltage between the plus pins of each channel. It should be zero or very close to it. If you see voltages of any size then they don't match.

If you have a way of recording the voltage going to the headphones, then record each channel and compare them to each other using Deltawave.
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
Blind test?
I would love to be able to do that, IF my family members are willing.

Since my family members already suffered through two DAC related blind tests with me, the likelihood of that happening anytime soon is slim.
: (

You can see many of the blind tests I have done being tracked here:

I do wonder how to properly do blind headphones amp tests. Each time my family member flips a coin to decide which amp to connect, I would leave the room and then come back blind folded and listen?

I think my next step for now is again measurements. I will check with @Blumlein 88 suggestions maybe sometimes next week. Too lazy this weekend.

: )
 

half_dog

Senior Member
Joined
Jun 28, 2018
Messages
452
Likes
361
Location
Brazil
Because L30 is AC coupled this causes a veeery little roll off (less than 0.5dB) on bass close to 18Hz and usually when bass has some recess this gives a cleaner punch. I could perceive this with a NX4 DSD. Its high pass filter took 0.5dB off at 20Hz and it was audible in some songs and some headphones (K518LE and Starfield). But honestly I can't tell apart a Atom (dc coupled) vs L30. Perhaps the issue is in OH200...
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
Because L30 is AC coupled this causes a veeery little roll off (less than 0.5dB) on bass close to 18Hz and usually when bass has some recess this gives a cleaner punch. I could perceive this with a NX4 DSD. Its high pass filter took 0.5dB off at 20Hz and it was audible in some songs and some headphones (K518LE and Starfield). But honestly I can't tell apart a Atom (dc coupled) vs L30. Perhaps the issue is in OH200...
I looked at Amir's measurements for L30 and HO200 and it seems both amp's freq response are flat down to 10hz. May I know how you find that 0.5db roll off for L30?


Thanks!
 
Last edited:

half_dog

Senior Member
Joined
Jun 28, 2018
Messages
452
Likes
361
Location
Brazil
Probably I'm confusing myself because of NX4 DSD - this I can confirm the 0.5dB roll off. I "found" it measuring L30 with an audio interface and comparing to Atom. There was a difference between these two but it wasn't that big. Atom FR kept equal to the internal loopback but L30 didn't. It rolled off "earlier". I coul read it wrong, 0.5dB instead of 0.05dB
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
Probably I'm confusing myself because of NX4 DSD - this I can confirm the 0.5dB roll off. I "found" it measuring L30 with an audio interface and comparing to Atom. There was a difference between these two but it wasn't that big. Atom FR kept equal to the internal loopback but L30 didn't. It rolled off "earlier". I coul read it wrong, 0.5dB instead of 0.05dB
For DAC, I just remember that I was told a while back that it is very hard to channel balance DAC with dual chips. My Gustard x16 has ~0.2db channel imbalance and I was told by Gustard that 0.2db is within their accepted level of deviation.

So I am wondering if a "fully balanced" (as claimed in https://www.smsl-audio.com/portal/product/detail/id/766.html) amp could also have certain deviations.

I wonder what "fully balanced" means in this case for smsl ho200. Two fully separate paths for left and right channels, from input all the way to output?

Would this "fully balanced" design cause slight timing/phase issues in bass tones?
 

half_dog

Senior Member
Joined
Jun 28, 2018
Messages
452
Likes
361
Location
Brazil
For DAC, I just remember that I was told a while back that it is very hard to channel balance DAC with dual chips. My Gustard x16 has ~0.2db channel imbalance and I was told by Gustard that 0.2db is within their accepted level of deviation.
It seems acceptable. E30 and E50 have a similar tolerance as well (less than 0.3dB), although I got much less than this when measured them. What I understand about balanced circuit in this case is when a signal is splitted in two polarities (positive and negative, the amp may receive the signal already split) and each polarity has its own circuit (gain, buffer etc) until get into headphone. To separate left and right it is not necessary to have balanced signals.
About balanced circuits they are more prone to have noise and phase issues, they require higher precision to work properly because at the end both polarities signals are summed after one of then being inverted. The signals need to be the most similar as possible to avoid those issues... talking about it and like that, it seems OH200 may be the "culprit".
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
It seems acceptable. E30 and E50 have a similar tolerance as well (less than 0.3dB), although I got much less than this when measured them. What I understand about balanced circuit in this case is when a signal is splitted in two polarities (positive and negative, the amp may receive the signal already split) and each polarity has its own circuit (gain, buffer etc) until get into headphone. To separate left and right it is not necessary to have balanced signals.
About balanced circuits they are more prone to have noise and phase issues, they require higher precision to work properly because at the end both polarities signals are summed after one of then being inverted. The signals need to be the most similar as possible to avoid those issues... talking about it and like that, it seems OH200 may be the "culprit".
Thanks for the explanation!

I have ordered 3.5 mm to Dual RCA female adapter and ordered longer RCA to XLR cables so that I can capture the headphones output using my ADC (that takes XLR inputs). I hope I will have time to do some captures and comparisons after Christmas.
 

half_dog

Senior Member
Joined
Jun 28, 2018
Messages
452
Likes
361
Location
Brazil
Channel Balance of E50: <0.03dB (~0.025)
Channel Balance of D10s: cant tell haha
 

Attachments

  • E50 vs D10s channel balance.PNG
    E50 vs D10s channel balance.PNG
    164.8 KB · Views: 82

half_dog

Senior Member
Joined
Jun 28, 2018
Messages
452
Likes
361
Location
Brazil
D10s stock vs D10s with L30 on signal path. The roll off is far far from 0.5dB... Completely irrelevant for audio signal.
 

Attachments

  • D10s stock vs D10s with L30.PNG
    D10s stock vs D10s with L30.PNG
    167.7 KB · Views: 101

bravomail

Addicted to Fun and Learning
Joined
Oct 19, 2018
Messages
817
Likes
461
Before measurements, I sensed bass difference between Topping L30 and SMSL HO200.

L30 seems "cleaner" than HO200 in bass "feels".

After voltage matched, I still feel the same.
Thanks!
U r not alone. L30 sounds lean. :)
I don't think any messing with the phase will help.
Try playing with E30 filters. I found brickwall filter works the best. It brings back some of that lost bass.
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
U r not alone. L30 sounds lean. :)
I don't think any messing with the phase will help.
Try playing with E30 filters. I found brickwall filter works the best. It brings back some of that lost bass.
May I know which filter of e30 you are using? I have been using filter 2.

I like to "feel" the "vibrations" of bass notes cleanly. I also like to feel clean bass "kicks". In regards to "cleanliness" of feels, I prefer L30 over HO200. So, I am not looking to improve L30's bass in this case. I am more curious about why ho200 is not that clean.

Phase issue is just a guess. Will see if capturing the outputs and checking them would reveal anything or not.

Thanks!
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,424
Likes
7,940
Location
Brussels, Belgium
What headphones are you using btw?

It’s best to use planar magnetic for these kinds of test to remove impedance variation from the equation.
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
What headphones are you using btw?

It’s best to use planar magnetic for these kinds of test to remove impedance variation from the equation.
As mentioned in my first post, AKG K371.
; )

Would this headphones cause impedance variation issue for either amp?

Thanks!
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,424
Likes
7,940
Location
Brussels, Belgium
As mentioned in my first post, AKG K371.
; )

Would this headphones cause impedance variation issue for either amp?

Thanks!
No because the headphone amplifier's output impedance is very low. but it's something to keep in mind since you're talking about very minute differences.
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
I did some capturings and here are the delta of phase for each channels. Anything of interest?

First, I played from 40 seconds of video for about 1 minute and captured Topping E30's RCA output using Focusrite Forte.

Then, I played back the captured wav file and captured the outputs of amps using these chains:
Laptop -> E30 -> L30 headphone output -> Focusrite Forte -> Laptop
Laptop -> E30 -> HO200 headphone -> Focusrite Forte -> Laptop

Anything interesting in delta of phase when comparing L30 vs HO200?

Thanks!

Note that the file name of HO200 mentioned L30 capture. Those are typos. Should said E30 capture. Both L30 and HO200 using same e30 capture file.
Channel 1 phase delta:
topping_l30_vs_smsl_ho200_phase_delta_right_channel.PNG


Channel 2 phase delta:
topping_l30_vs_smsl_ho200_phase_delta_left_channel.PNG
 
Last edited:
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
I added Sabaj a10h captures for comparisons.

Like before, I played back the captured wav file and captured the outputs of amp using this chain:
Laptop -> E30 -> a10h headphone output -> Focusrite Forte -> Laptop

Compare L30 to a10h, there are less phase delta in the ~40 hz region as compared to L30 vs HO200.
Channel 1:
topping_l30_vs_sabaj_a10h_phase_delta_right_channel.PNG


Channel 2:
topping_l30_vs_sabaj_a10h_phase_delta_left_channel.PNG




However, comparing A10H to HO200, the ~40hz phase difference showed up again.

Channel 1:
sabaj_a10h_vs_smsl_ho200_phase_delta_right_channel.PNG


Channel 2:
sabaj_a10h_vs_smsl_ho200_phase_delta_left_channel.PNG
 
OP
Pdxwayne

Pdxwayne

Major Contributor
Joined
Sep 15, 2020
Messages
3,219
Likes
1,172
Anyone wants me to measure anything else before I return the HO200?
 
Top Bottom