Loxjie P20 technical measurements

Discussion in 'Headphone Amplifier Measurements' started by atomicbob, Mar 20, 2019.

  1. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Loxjie P20 technical measurements

    If you are unfamiliar with audio measurements please use a search engine with the query:
    "audio measurements" or "audio measurement handbook"
    Look for publications by Richard C. Cabot and also by Bob Metzler, both from Audio Precision. There are other useful publications as well. These will provide basic knowledge.
    Interpretation of the following measurements is beyond the scope of this post.

    The data presented were collected as follows:

    1. PrismSound dScope III, picoscope 5243B, Micsig DP10013 balanced probe, Keysight 34465A DMM
    2. 300 and 30 ohm loads used for measurements
    3. Balanced cables Belden 1800F with Neutrik XLR connectors
    4. Vaunix Lab Brick USB hub for measurement equipment
    5. Audioquest Forest and Schiit Pyst USB cables used for measurement equipment
    6. Amplifier gain adjusted to 0 dB gain @ 0.775 Vrms 0 dBu
    7. P20 gain set to -16 to achieve 0 dB gain (varies a little between 30R and 300R load)
    8. 14 and 16 AWG sheilded power cables used

    Reference data for two headphones to keep in mind:
    HD 650 impedance 300R, sensitivity 98 dB/mW
    HE-500 impedance 38R, sensitivity 89 dB/mW

    SPL levels for above headphones for reference:
    +9 dBu 300R 16.00 mW - 110 dBSPL @ 98dB/mW
    0 dBu 300R 2.00 mW - 101 dBSPL @ 98dB/mW
    -10 dBu 300R 0.20 mW - 91 dBSPL @ 98dB/mW
    -20 dBu 300R 0.02 mW - 81 dBSPL @ 98dB/mW

    +9 dBu 30R 160.00 mW - 110 dBSPL @ 89dB/mW
    0 dBu 30R 20.00 mW - 102 dBSPL @ 89dB/mW
    -10 dBu 30R 2.00 mW - 92 dBSPL @ 89dB/mW
    -20 dBu 30R 0.20 mW - 82 dBSPL @ 89dB/mW

    All testing performed at 0 dBu unless otherwise noted.
    This level is consistent with listening to headphones (referenced above) at 90 dBSPL average with peaks to 100 dBSPL, if the music has 10 dB Peak to Avg ratio. That is LOUD for long listening sessions.

    Highlights
    True balanced operation input to output
    JRC NJW1195 4 channel digitally controlled analog volume control - no channel imbalance issues
    Nearly perfect gain linearity spanning over 100 dB range!!
    +/- 1.5 dB gain linearity over 140 dB range!!
    SNR greater than 120 dB!!!!!
    Crosstalk greater than 115 dB!!!
    Bandwidth: high cutoff greater than 652 KHz at 30R and 1.29 MHz at 300R

    Other points of note
    NE5532A opamp output stage
    Capacitive coupled stages
    Output impedance 47R - may not cause joy with planar and other low impedance headphones

    Performance enhancement opportunities
    Works with NoiseNuke(TM) improving response in low end grunt.
    Low end grunt further improved with 12V 2A ZeroZone LPS, a worthy performance upgrade.
    GE 5670W driver tubes provide a smoother presentation for tracks that become noticeably spitty with 6N3. Please note that once the shorter 5670W is inserted, extraction requires careful rocking technique and extreme patience. Does not lend to tube rolling.

    PLEASE DO NOT use WE396 in this amp. While worthy of 5670W at $5 to $10 / tube, the extra performance capability of WE396 is NOT going to turn this amp into a Cavalli Liquid Platinum or ECP Audio T4. It will be a total waste and prevent use of the rare WE396 in amps that would actually benefit.

    Bottom line
    P20 is an engaging, inexpensive amp with some special performance features. High impedance headphone users may wish to take note, especially for desktop second systems.


    Test Data Organization
    Post
    1 - introduction
    1 - Bal input to 4XF output 300R load part 1 of 3
    2 - Bal input to 4XF output 300R load part 2 of 3
    3 - Bal input to 4XF output 300R load part 3 of 3
    4 - Bal input to 4XF output 30R load part 1 of 3
    5 - Bal input to 4XF output 30R load part 2 of 3
    6 - Bal input to 4XF output 30R load part 3 of 3
    7 - PCB picture, gain control accuracy, channel balance

    My personal preferences include listening typically at an average of 75 dB SPL. Assuming a typical Peak to Average ratio of 10 dB puts peaks at 85 dB SPL. This suggests the measurements at -20 dBu represent my experience with HD6x0 and HD800 headphones. Auditory experiences for someone listening 10 dB SPL average higher may diverge from my own experiences.

    editorial - I believe there is a tendency to focus excessively on numbers such as THD attempting to reduce a complex topic to a single number representing a sort of goodness rating. Graphs can present far more data conveying a better understanding of the measurement over a range of frequency, levels, etc. Please don't read too much into the numbers presented. Read the tutorial presented here for a good example of why.

    Picture of Audio Lab setup for P20 measurements
    DSC_0065_small.jpg

    Bal input 4XF output 300 ohm load part 1 of 3

    Output impedance measured 47R

    P20 THD+N THD nth harmonic distortion (A04) FFT 0dBu 300R load
    20190221 P20 A04 THD+N  THD  nth-HD FFT 300R 4XF.png

    P20 Gain vs Amplitude 300R load
    20190221 P20 1 KHz gain linearity 300R.PNG

    P20 THD THD+N vs Freq 0dBu 300R load
    20190221 P20 THD THD+N vs Freq 300R 4XF.png

    P20 1 KHz spectrum Left and Right 0dBu 300R load
    20190221 P20 1 KHz THD+N THD FFT 0dBu 300R 4XF.png

    P20 50 Hz + 7000 Hz spectrum Left and Right 0dBu 300R load
    20190221 P20 50+7000Hz 0dBu 300R 4XF.png

    P20 600 Hz + 1700 Hz spectrum Left and Right 0dBu 300R load
    20190221 P20 600+1700Hz 0dBu 300R 4XF.png

    P20 IMD spectrum Left and Right -16dBu 300R load
    20190221 P20 19+20KHz -16dBu 300R 4XF.png

    P20 residual noise spectrum Left and Right A=0dB 300R load

    20190221 P20 residual noise A=0dB 300R 4XF.png
     
    Last edited: Mar 20, 2019
  2. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Bal input 4XF output 300 ohm load part 2 of 3

    P20 square wave 20 Hz 2 Vpp 10 mS / div 300R load 100 KHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10mS div 100KHz filter 300R.png

    P20 square wave 20 Hz 2 Vpp 10 mS / div 300R load 5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10mS div 5MHz filter 300R.png

    P20 square wave 20 Hz 2 Vpp 10 uS / div 300R load 2.5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10uS div 2_5MHz filter 300R.png

    P20 square wave 20 Hz 2 Vpp 1 uS / div 300R load 5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 1uS div 5MHz filter 300R - BW calc.png
    Bandwidth estimation: BW (MHz) = 0.35 / RT (mS)
    Where RT = 10 to 90% Rise Time
    0.35 / 271.6 nS = 1.29 MHz

    P20 amplitude - phase - gain (A01) 300R load
    20190221 P20 A01 amplitude - phase - gain 300R.png

    P20 signal to noise (A07) 300R load
    20190221 P20 A07 SNR 300R.png
    Impressive dynamic range!!

    P20 THD+N THD nth harmonic distortion (A04) 300R load
    20190221 P20 A04 THD+N  THD  nth-HD 300R.png

    P20 Crosstalk (A08 A09) 300R load
    20190221 P20 A08 A09 Crosstalk 300R.png
    Pretty impressive crosstalk performance, almost dual mono.
     
    Last edited: Mar 20, 2019
  3. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Bal input 4XF output 300 ohm load part 3 of 3

    P20 Frequency Response and Interchannel Phase pg 1/2 300R load
    G2_L1T2_1_A.png

    P20 Frequency Response and Interchannel Phase pg 2/2 300R load
    G2_L1T2_1_B.png

    P20 THD+N vs Amplitude Left 300R load
    G2_L1T7_1_A.png

    P20 THD+N vs Amplitude Right 300R load
    G2_L1T7_1_B.png

    P20 Crosstalk Left to Right vs Frequency 300R load
    G1_L1T11_1.png

    P20 Crosstalk Right to Left vs Frequency 300R load
    G0_L1T12_1.png
     
    Last edited: Mar 20, 2019
  4. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Bal input 4XF output 30 ohm load part 1 of 3

    P20 THD+N THD nth harmonic distortion (A04) FFT 0dBu 30R load
    20190221 P20 A04 THD+N  THD  nth-HD FFT 30R 4XF.png

    P20 Gain vs Amplitude 30R load
    20190221 P20 1 KHz gain linearity 30R.PNG

    P20 THD THD+N vs Freq 0dBu 30R
    20190221 P20 THD THD+N vs Freq 30R 4XF.png

    P20 1 KHz spectrum Left and Right 0dBu 30R load
    20190221 P20 1 KHz THD+N THD FFT 0dBu 30R 4XF.png

    P20 50 Hz + 7000 Hz spectrum Left and Right 0dBu 30R load
    20190221 P20 50+7000Hz 0dBu 30R 4XF.png
    50 and 7000 Hz sines, each at 0 dBu sum to +3 dBu. The P20 is running into the power supply limits and clipping at this level. Reducing 50 and 7000 Hz sines to -3 dBu each now results in 0 dBu total which the P20 can handle for 30R loads as shown in next graph.

    P20 50 Hz + 7000 Hz spectrum Left and Right -3dBu 30R load
    20190221 P20 50+7000Hz -3dBu 30R 4XF.png


    P20 600 Hz + 1700 Hz spectrum Left and Right 0dBu 30R load
    20190221 P20 600+1700Hz 0dBu 30R 4XF.png
    Same issue as 50+7000Hz multitone noted above with clipping at 30R loads.

    P20 600 Hz + 1700 Hz spectrum Left and Right -3dBu 30R load
    20190221 P20 600+1700Hz -3dBu 30R 4XF.png

    P20 IMD spectrum Left and Right -16dBu 30R load
    20190221 P20 19+20KHz -16dBu 30R 4XF.png

    P20 residual noise spectrum Left and Right A=0dB 30R load
    20190221 P20 residual noise A=0dB 30R 4XF.png
    Impressive noise floor.
     
    Last edited: Mar 20, 2019
  5. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Bal input 4XF output 30 ohm load part 2 of 3

    P20 square wave 20 Hz 2 Vpp 10 mS / div 30R load 100 KHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10mS div 100KHz filter 30R.png

    P20 square wave 20 Hz 2 Vpp 10 mS / div 30R load 5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10mS div 5MHz filter 30R.png

    P20 square wave 20 Hz 2 Vpp 10 uS / div 30R load 2.5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 10uS div 2_5MHz filter 30R.png

    P20 square wave 20 Hz 2 Vpp 1 uS / div 30R load 5 MHz BW
    20190221 SigGen P20 20 Hz square 2000mVpp 1uS div 5MHz filter 30R - BW calc.png
    Bandwidth estimation: BW (MHz) = 0.35 / RT (mS)
    Where RT = 10 to 90% Rise Time
    0.35 / 536.8 nS = 652 KHz


    P20 amplitude - phase - gain (A01) 30R load
    20190221 P20 A01 amplitude - phase - gain 30R.png

    P20 signal to noise (A07) 30R load
    20190221 P20 A07 SNR 30R.png
    Impressive dynamic range.

    P20 THD+N THD nth harmonic distortion (A04) 30R load
    20190221 P20 A04 THD+N  THD  nth-HD 30R.png

    P20 Crosstalk (A08 A09) 30R load

    20190221 P20 A08 A09 Crosstalk 30R.png
    Very impressive crosstalk measurements.
     
    Last edited: Mar 20, 2019
  6. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    Bal input 4XF output 30 ohm load part 3 of 3

    P20 Frequency Response and Interchannel Phase pg 1/2 30R load
    G2_L1T2_1_A.png

    P20 Frequency Response and Interchannel Phase pg 2/2 30R load
    G2_L1T2_1_B.png

    P20 THD+N vs Amplitude Left 30R load
    G2_L1T7_1_A.png

    P20 THD+N vs Amplitude Right 30R load
    G2_L1T7_1_B.png

    P20 Crosstalk Left to Right vs Frequency 30R load
    G1_L1T11_1.png

    P20 Crosstalk Right to Left vs Frequency 30R load
    G0_L1T12_1.png
     
    Last edited: Mar 20, 2019
  7. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    PCB picture, gain control accuracy, channel balance

    P20 PCB:
    DSC_0102_small.jpg
    Quality components. On board DC converter provides +115Vdc for plate voltages (measured 111Vdc). Design engineer and PCB layout people are definitely capable with seriously good technical skills.

    P20 volume control accuracy:
    20190307 P20 gain accuracy GE5670W 300R.png
    Left and right channels are on top of each other in this graph. Precise, repeatable volume settings.

    P20 channel balance:

    20190307 P20 chan balance GE5670W 300R.png
    Channel balance is exceptionally tight, within 0.15 dB across volume settings.
     
    Last edited: Mar 20, 2019
  8. atomicbob

    atomicbob dScope Yoda

    Pyrate BWC MZR
    Joined:
    Sep 27, 2015
    Likes Received:
    18,652
    Trophy Points:
    113
    Location:
    On planet
    reserved 8/8.
     
  9. Baten

    Baten Friend

    Pyrate
    Joined:
    Mar 18, 2018
    Likes Received:
    1,130
    Trophy Points:
    93
    Location:
    EU
    Perhaps worth noting, from listening the SE input sounded drastically worse than Bal.
     
  10. restrav3

    restrav3 Likes Audio-GD

    Contributor
    Joined:
    Jun 18, 2018
    Likes Received:
    101
    Trophy Points:
    43
    Location:
    PA
    is that your personal experience or Zeos'? because I can tell you that he tends to get these things right, in the opposite universe only. so if he says that it sounds better balanced, I'd bet on SE being superior. in this case I'd bet it is the old issue of the BAL mode being louder and that's it.
     
  11. bilboda

    bilboda Florida boomer

    Pyrate Banned
    Joined:
    Nov 20, 2016
    Likes Received:
    834
    Trophy Points:
    93
    Location:
    Miama
    I went thru the inputs and outputs with my Gungnir Multibit and found the se input to fare worse, almost a different amp entirely, like a wee lil brother.. Verums loved the se output with balanced input.
     
  12. purr1n

    purr1n Desire for betterer is endless.

    Staff Member Pyrate BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    89,771
    Trophy Points:
    113
    Location:
    Padre Island CC TX
    That could mean Gungnir BAL outputs sound better than SE outputs.
     
  13. Baten

    Baten Friend

    Pyrate
    Joined:
    Mar 18, 2018
    Likes Received:
    1,130
    Trophy Points:
    93
    Location:
    EU
    I did say 'from listening'. I owned one. Returned it but seeing these measurements maybe I should have given it some more listening time.
     
  14. bilboda

    bilboda Florida boomer

    Pyrate Banned
    Joined:
    Nov 20, 2016
    Likes Received:
    834
    Trophy Points:
    93
    Location:
    Miama
    A full comparison. SE input is not trash but something needs to be balanced with the Clears.
    Quoting myself from the other thread.

    I put my noise nuke to work, nice. As noted by someone in the review thread, the supplied smps works but needs more amps. He added a Jameco 5 amp smps.
    With supplied smps, the display turns off briefly after it is turned on. With a better power source, this does not happen.
    Mebbe I should re-evaluate the SE input now. Work beckons.....
     
  15. Baten

    Baten Friend

    Pyrate
    Joined:
    Mar 18, 2018
    Likes Received:
    1,130
    Trophy Points:
    93
    Location:
    EU
    Just received my second unit. This one runs better. No weirdness to the sound, works as it should/ as one would expect. Balanced in/Balanced out it does sound more than decent. But eh~~ I'm still sending it back.

    Had it warm up and listened for quite some time. To me it sounds too much like a tube trying to sound solid state; Probably the topology characterizing it so. I'd rather keep saving and get the liquid platinum new (or used if a deal pops up). Not to be crude but the Loxjie... not quite my amp.

    For a $99 device it's great to test/try out balanced without spending big boy bucks. But imo it's not a good way of trying out tubes. This thing sounds very solid statey to me, but I'm definitely curious to hear other's listening impressions.

    PS one more thing, this thing is very SILENT to me. Buzzing/humming is ultra minute. Props to the designers.
     
  16. bluebyte60

    bluebyte60 New

    Contributor
    Joined:
    Jun 12, 2017
    Likes Received:
    58
    Trophy Points:
    13
    Location:
    Sunnyvale, CA
    Ordered one for a try, listen for a while. Compare to THX AAA 789, I think I like it! I use hd800 , P20 sounds warmer (maybe output impedance does the trick?), more engagement while 789 is more analytical and has better separation. Consider it's a $100 amp, I think it worth every bucks despite the looks (just not my type, yet)
     
    Last edited: May 2, 2019
  17. Baten

    Baten Friend

    Pyrate
    Joined:
    Mar 18, 2018
    Likes Received:
    1,130
    Trophy Points:
    93
    Location:
    EU
    High impedance out works well with high impedance Sennheiser series ;)
     

Share This Page