How I improved the B&O sound quality for free

Discussion in 'Infotainment and Electronics (Nav, Audio, Video, Bluetooth)' started by GTP, May 8, 2020.

  1. GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #1 GTP, May 8, 2020
    Last edited: May 10, 2020
    TLDR takeaways:
    1. The factory set the level of the B&O subwoofer way too high by at least 12dB.
    2. The tonal balance can be adjusted close to flat with somewhat extreme settings of the 3 tone controls.
    3. Imaging is better with the balance 1 click shy of centered on the passenger seat centerline. (For when there is no passenger, and assuming center channel speaker is OFF.)

    After much testing and experimenting, I improved the sound of the B&O system just by adjusting the controls.

    Here is the final result:
    MicPassHeadRestFullRangeMid#4Treble#6SubMinBal0_SeatUpWindowsUpTrunkClosed.png

    Even after setting the subwoofer tone control to minimum, and the midrange control up by 4 clicks and the treble up by +5 or +6, the subwoofer level is still too hot by around 6dB.

    Here are the settings (the balance is set for driver with no passenger):
    20200508_152051.jpg

    20200508_152624.jpg

    The OmniMic was placed into passenger headrest:
    20200508_152859.jpg


    I will post supporting data that lead me to this alignment next.
     
    gpierce, jbeez and Järn like this.
  2. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #2 GTP, May 8, 2020
    Last edited: May 10, 2020
    Flat Frequency Response

    I auditioned the B&O system at a dealer before getting my car. I considered ordering and wanted to know if it was worth the extra $900. At first I concluded that it wasn’t. And my opinion was unchanged until now. The bass was way bloated and I couldn’t seem to turn it down enough. The sound quality was rough, and there was not a good stereo image.

    It’s as if some Ford project manager swooped in just before sign-off and exclaimed, “Where’s the bass?! We gotta crank up the bass or no one will like it!” Well, I don’t like artificially bloated bass. Here is the custom target curve for my high-end home system. There is about 2dB gradual emphasis from high to low frequency.

    2020-05-08 17_18_29-Dirac Live Calibration Tool (DiracPreset1(EllenTargetTo1300Hz)).png

    Regardless I found a nearly fully loaded car and it came with the B&O system. Oh well, I thought, at least it will have a CD player, and subwoofer wiring harness in case I want to replace it. I already knew about snipping the center speaker wire on my 2016 Mustang.

    Some of you may have read my thread “Join me in fixing the center speaker issue”. The overall conclusion of that thread (and as confirmed by many others) is that the sound is simply better with NO center speaker playing. And so we have snipped the green wire (positive lead) going to the speaker, near the amp connector.
     
  3. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #3 GTP, May 8, 2020
    Last edited: May 14, 2020
    I read some other threads on what others have done with their systems. Replacing speakers, amps, programming, and so on.

    My opinion (even as a lifelong audiophile and professional loudspeaker designer) was that
    1. The door 3-way speakers are pretty good
    2. The center speaker was cheap and providing horrible sound at a nice loud level, and destroyed the stereo imaging
    3. The package shelf speakers were terrible in the 9-speaker (non-B&O) system. They are decent in the B&O system. They play quieter than the front, and so I no longer have to fade most of the sound to the front.
    4. I did not have a subwoofer on my last (9-speaker) car, and it lacked low and middle bass. But the sub plays way too loud on the B&O system.
    Even though I have high standards for my home system, I don’t carry those over to car audio. There are too many intangibles such as background noise, reflective cabin acoustics, driver delays, etc. Even worse is the challenge of working on a car system such as getting into tight spaces, every connector is custom, and endless research just to know what I’m working with.
     
    marklboris and CS19 like this.
  4. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #4 GTP, May 8, 2020
    Last edited: May 10, 2020
    And now for the technical supporting data.

    I first removed the subwoofer box, and then the woofer itself. Of course that required the obligatory breaking of plastic parts.
    20200506_151048.jpg

    I then verified connector pins and polarity for the two voice coils. This is a very important step to avoid electrical mishaps.
    20200506_130137.jpg

    20200506_130309.jpg

    I wanted to know the voice coil DCR because I intended to test with my lab amp, and so I was concerned about the load impedance being too low. Here is the DCR for one coil, indicating it is nominally 4 ohms.
    20200506_130944.jpg

    Next I tested the free-air and in-box impedance of the factory woofer. I used a 4” PVC tube I had to lift the woofer up off the clip leads. It was a real struggle to get the leads to stay on the pins. (Refer to my complaint about specialized connectors.)
    20200506_131743.jpg
    20200506_131805.jpg

    Here is the impedance graph:
    ZfaibFactoryWoofer.png

    The sharp rise in impedance above 100Hz indicates that the woofer has no copper in the motor. Something that is featured on higher-end woofers. Not a big deal for an active powered woofer that only serves bass duty.

    Then I tested the nearfield SPL response of the factory subwoofer, but this was driven by my Crown K1 lab amp. This provided me the response knowing that there is no EQ programmed in anywhere.

    OMG, look at the peaked response!
    NFibFactoryWoofer.png

    No worries, I thought. Surely Ford will have EQ’d this hump out of the response. Nope! It is still there in the factory amp driven sub.
    Lab-v-Ford_driven_SPL.png
    20200506_150949.jpg

    I then spent quite a bit of time designing a passive EQ network to flatten the response. These graphs show a better bandpass response from 30 – 100Hz, and intentional loss of -12dB so as to reduce its output. Notice that the impedance is unchanged at very low frequency, and is overall higher than original throughout most of the pass band. This makes it easier to drive by the amp.

    Unfortunately, four very large value inductors are called for. Two are needed in series simply because of sourcing a large value, and then two pairs are needed because there are two voice coils.

    A total of ten crossover parts was required. EQd_factory_sub.png Z_EQd_factory_sub.png
     
    Bryantbjb12 likes this.
  5. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #5 GTP, May 8, 2020
    Last edited: May 8, 2020
    At this point I decided to do what I should have done from the outset – SPL testing inside the car.

    As it turns out, the cabin has a suckout at 60Hz. Maybe that is the reason for the 66Hz response peak? Even though there is a large peak, it still does not fully overcome the 60Hz dip (this curve is with the trunk and windows open – I forgot to close those). This graph was with the mid and treble controls all the way down.
    60HzDip.png
     
  6. StangTime

    StangTime Well-Known Member

    First Name:
    Todd
    Vehicle(s):
    19' GT PP1, Manual
    Joined:
    Apr 16, 2019
    Location:
    Ontario
    Posts:
    753
    Likes Received:
    477
    Rating:
    100%
    1   0   0
    Interesting. I have the 9 speaker system. I did a basic RTA on it. No calibrated mic available so I used the internal mic on my ipad (which is compensated for in the RTA). The 9 speaker system takes a dump at 10khz and there is no more high frequencies produced. The B&O must use a different tweeter or the APIM is EQ'd differently. Thanks for sharing this info.
     
  7. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #7 GTP, May 8, 2020
    Last edited: May 8, 2020
    I continued with more measurements.

    I first wanted to characterize the gain range of the bass control. IOW, what is the dB span between minimum and maximum settings of the bass tone control?

    A new challenge is that I cannot inject my favorite audio analyzer signal into the car, because there is no provision for analog audio input. Therefore I had to fall back on CD test signals.

    To probe the connector I made some probe pins out of paper clips, and insulated them (refer to previous comment about electrical mishaps).
    20200508_114702.jpg

    I connected my DVM to one of the subwoofer cable channels.
    < photo missing >

    I played a constant 50Hz test tone from the CD and then plotted the 15 voltage steps from changing the tone control from minimum to maximum. Here is the resulting chart with voltages converted to dB and normalized for 0dB at the middle setting:
    BassMinMax-at-50Hz_Chart.png
    So each step is less than a dB, and the graph is nearly a straight line. The tone control has a range of +|- 5dB.
     
  8. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #8 GTP, May 8, 2020
    Last edited: May 8, 2020
    Effect of right rear seat up versus down:RearSeatDownUp.png
    It's about 2dB increase with the seat down, but the response shape remains the same.
     
    Bryantbjb12 likes this.
  9. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    #9 GTP, May 8, 2020
    Last edited: May 8, 2020
    Effect of windows up (blue) versus down (red):
    WindowsDownUpTrunkOpen.png

    The response is a little smoother at 200Hz and in the 1-2kHz octave. Some cabin gain is seen below 20Hz. Again the trunk was accidentally left open, so the 60Hz notch remains.
     
    Bryantbjb12 likes this.
  10. Mustang_Lou

    Mustang_Lou Well-Known Member

    Vehicle(s):
    2020 Mustang Bullitt
    Joined:
    May 1, 2016
    Location:
    Toronto, Canada
    Posts:
    1,533
    Likes Received:
    568
    0   0   0
    I applaud your efforts @GTP but way more than I'd have done for what's likely a minor sound improvement but we've all got our "thing" and sound quality is obviously yours.

    So summary of the testing and all is on the sound settings menu shot at post #1, right? I don't find the bass overbearing at all but I'll try that setting nevertheless.
     
    zero_chance and GTP like this.
  11. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    Yes, you make the obvious and valid point. I.e. that I had to go through all this (from the POV of loudspeaker expert) just to determine that the B&O system is pretty good, but the subwoofer level was set way too high at the factory.

    The sound was unlistenable to me because I am long accustomed to "flat" tonal balance, including throughout the bass. As I said, I discovered this in a dealer parking lot when I began shopping for a 2019 GT. (Once the bass is flat respective to the rest of the spectrum, the listener can easily follow any aspect of the music, including the bass player. The bass no longer "drowns out" the rest of the music.)

    I was prepared to go full monty by ordering a Parts Express 10" DVC woofer, and either retrofitting it into the factory box, or else buy a Zenclosures box. I have the means to test for and design the optimum box size, sealed vs vented, etc.

    Ultimately I determined (through rigorous scientific method, lol) that I could achieve the sound I wanted just by using the controls. So that is the good news! And others can try this adjustment for free after I put in the hours of effort.
     
    Järn, Pittpa, Scoobs and 1 other person like this.
  12. FreePenguin

    FreePenguin Well-Known Member

    First Name:
    Donald
    Vehicle(s):
    17 mustang
    Joined:
    Sep 29, 2019
    Location:
    Ohio
    Posts:
    982
    Likes Received:
    528
    Rating:
    100%
    1   0   0
    try to disable the factory EQ with forescan, maybe that would be better for you. makes everything flat
    that's all I did and I love it.
     
  13. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    Yes, I did consider and still am considering that option. I would have to slog through the long Forscan thread(s) and study up on that. Any direct pointing me to the relevant info is appreciated. Nevertheless, I wanted to checkout what was going on via my own acoustic methods first.

    Here is the Ford subwoofer amp transfer function (i.e. drive voltage):
    1. 1/3-octave data points from 20 to 200Hz.
    2. Low-pass crossover frequency is about 80Hz, and 12dB/octave slope.
    3. Likely capacitive-coupled amplifier seen by the 6dB/octave slope below 20Hz.
    4. Some EQ going on between in the 31 to 63Hz octave.
    FordAmpDriveVoltage.png

    Showing the SPL response again from Post #1: MicPassHeadRestFullRangeMid#4Treble#6SubMinBal0_SeatUpWindowsUpTrunkClosed.png
     
  14. FreePenguin

    FreePenguin Well-Known Member

    First Name:
    Donald
    Vehicle(s):
    17 mustang
    Joined:
    Sep 29, 2019
    Location:
    Ohio
    Posts:
    982
    Likes Received:
    528
    Rating:
    100%
    1   0   0
    There’s really nothing to read up on, you change literally this one code

    Don’t like it? Change code back. It’s on and off

    ACM 727-01-02 xXxx xx 0=EQ Off

    I clipped the wire to center speaker under left kick plastic panel, I think mine was green. Just used pair of finger nail clippers

    Disabling eq plus cutting the center speaker completed my system

    You literally plug in turn on. Find ACM under module. Then click on that box I listed above

    Just gotta download forescan

    5 minutes to download, 2 minutes to register forescan account
    2 minutes to change the code 4C02BAA4-97AC-42F8-8CAB-9A40601EF075.png
     
  15. OP
    OP
    GTP

    GTP Deutsche Pony

    First Name:
    Philip
    Vehicle(s):
    Poor Man's McLaren
    Joined:
    May 27, 2015
    Location:
    Indy
    Posts:
    2,897
    Likes Received:
    791
    Occupation:
    Engineer
    Rating:
    93.3%
    14   1   0
    Yes, I understand. But does that one line just remove EQ from the subwoofer channels? What about the subwoofer overall gain? That is primarily what I want lowered.

    Also, if EQ defeat takes out EQ from ALL the channels then that could be a problem Again, I'd have to know the voltage transfer function to every driver in the car before the defeat. It is possible that Ford/B&O compensated out cone breakups, and other non-flat anomalies in the drivers' responses. Something that I would not want eliminated!
     
Loading...

Share This Page