This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TDA3118D2 - distortion issue ...

Other Parts Discussed in Thread: TPA3130D2EVM, TPA3118D2

Hi,

I am using TDA3118D2 in a 2.1 stereo system. I suffer from audible distortion - especially noticeable when playing piano Music - kind of high frequency noise >5kHz.

I supply the amp with 20VDC which should give enough head room and the load is a peerless 3inch fullrange driver with high quality.

I run the highest gain and no power limit.

I have a very simple output filter - a ferrite bead of 150ohm followed by a 1nF/50V/COG cap to ground on each output terminal.

Please help me to find the cause of my distortion. Is it the output filter that causes the noise?

Thanks in advance

Markus

  • Some more info:

    The ferrite bead I use as output filter is an axial ferrite from Murata P/N BL02RN2R1M2B.

    /Markus

  • Hi Markus,

    Are you able to run a THD vs. output power and THD vs. frequency plot?
    This way you could figure out if output power limit or frequency was to blame for the distortion.

    My guess is it is the ferrite bead. The standard through hole EMI suppression beads are very bad for Class-D output. Their impedance is usually relatively high in the audible band and will climb very quickly with frequency. This will cause a lot of losses at higher frequency and can sap some of the output power and cause distortion as you mention.

    If you have a scope, you can try looking at the PWM output waveform at idle. If it is rounded or triangular, the beads are providing too much attenuation.

    If using a ferrite bead filter, the key is to pick a ferrite bead the will have minimal attenuation well past the PWM switching frequency. The bead should only provide impedance in MHZ region to help suppress high frequency EMI issues. On our TPA3140 EVM we use Murata NFZ2MSM301SN10L beads. If you take a look at the datasheet, notice that the impedance rises very quickly with frequency in the MHZ region. In the audible band the impedance will be very low and essential the DCR of the bead.

    Best Regards,
    Matt
  • Hi Matt,

    Thanks for your answer! I know that ferrite beads can have bad impact and generate audible noise especially in the upper mid - has experienced that before with another amp.

    The reason I picked this axial ferrite bead was because in the DS for TPA3118D2 there is a link to the TPA3130D2EVM where this type of beads, according to TI, have been tested with good result:

    SLOU341A–March 2012–Revised December 2013

    Anyway, I run my amp with supply=20V and a 4ohm load. I would prefer to use a ferrite bead with at least 7A current rating to avoid too much EMI-block degradation at high power output. How much peak current do you run in the TPA3140 EVM? The ferrite bead you use is rated for 3A.

    I think I will first try BL02RN1R2M2B which has slightly lower impedance below 1MHz and significantly lower at 10MHz.

    After that I will try to the P/N you proposed!

    What about running compleltely without output EMI-filter (no bead, no cap) on the PCB and instead add some ferrite rings on the speaker cable close to the board? I have rather short cables (<500mm)!

    Thanks and best regards

    Markus

  • Hi Markus,

    The ferrite bead used on the TAP3140EVM will have too low of a current spec to run with the TPA3116 flat out. I just wanted to illustrate an example of a good bead for class-D.

    As for ferrite rings around the speaker cable, this may work. However, I fear that you will have the same issue as the through hole beads unless you can find one with good properties as discussed.

    Best Regards,
    Matt