• State Resolved
  • Locked Locked
  • Replies 14 replies
  • Subscribers 33 subscribers
  • Views 2311 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TAS5630 2.1 application, channel separation problem

Junpu Xu
Prodigy 75 points
Other Parts Discussed in Thread: TAS5630, TAS5630B

Hello experts,

I am developing an amplifier in 2.1 system application by using TAS5630 IC. Everything works fine except the strange trouble of channel separations.

My configurations of TAS5630 amp:

-- Channel A & B are working in BTL mode, as subwoofer amplifier

-- Channel C & D are working in SE mode, as satellites

Problem description:

-- Sin wave signal applies to subwoofer channel,  no audio signal to satellites, but satellites amp have very annoying audiable noise. Noise keeps there even no speaker load connected to subwoofer (means keep subwoofer amp open)

-- Noise frequency from satellites changes when sin-wave input signal's amplitude changes, very strange

Eforts tried:

1) I thought noise might come from the pre-amp stage, possibly bad ground loop or what ever. I did a test to cut out the pre-stage of Channel C & D, connect input (before DC block capacitor) directly to AGND. It does not help, problem still there

2) Can output stage involve noise in? I did a test by even removing the output filter of Channel A&B completely, means channel A&B are working (I do see the outA & outB signals) but no current out. It does not help, problem still there

3)Too much noise at AGND pin? I connect even this pin directly to audio gnd with an additional wires, no help.

4) Can 12V auxilary power supply cause the problem? I solder out the on-board 12V auxilary power supply block, and connect by an additional power source (followed by a 7812 regulator). No help

Additional information

A) I reconfigure the system to stereo application, means Channel A&B in BTL mode, and Channel C&D in BTL mode. Surprised to see that there is no problem any more, means 2 channels work independently well.

B) I did another further test in stereo application, I connect 2pcs 2-Ohm speakers in series as load of Channel C & D, and connect the junction of 2 speakers to 1/2PVCC of channel C & D (you can call this as a similar 2pcs SE amp). Idea behind was to see whether the problem is still there, but just compensated each other between Channel C & D. Result showed that both speakers noisy. (cut the connection between mid of 2 speakers & 1/2PVCC of Channel C&D, problem disappear)

Does any of you meet this problem?

Junpu

 

  • 0
    TI__Guru**** 189401 points

    Hi, Junpu,

    Hmmm, this is indeed very strange.

    Do you get the noise when you have a load connected to the sub-woofer terminals? Leaving those floating may cause some strange things to happen.

    Do you have any oscilloscope plots you can post?

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 75 points

    Hi Don,

    Thanks for your reply! This is very strange, indeed.

    I got noise with speaker load (a 4 Ohm speaker) connected to the sub-woofer terminals, and also got noise with dummy load (resistor) connected to sub-woofer terminals. Actually I noticed the noise first when we did the leasoning test with speakers connected.

    I checked the Channel C & D 's output signal with oscilloscope, to be honest, I could not find strange things there. ~400KHz ripple voltlage (roughly peak-peak value 200mV) overwrite everything else. I used my oscilloscope's FFT function, got to see there is an abnormal peak in spectrum, at frequency range of several hundred Hz. I did not save the plots. If you think that's a very important hint, I can reproduce it and post out ...

    Thanks for all your help!

    Junpu

  • 0 in reply to Junpu Xu
    TI__Mastermind 18510 points

    Junpu, I will try to help with this.

    I wonder how you have grounded the AGND pin of TAS5630.  Did you connect it directly to the ground area under the IC, as in our EVM PCB layout?  In our EVM it is connected that way and also to ground plane on the input side (pin 1 side) of TAS5630.

    I think we need to see your schematic and maybe your PCB layout.  Can you make PDFs of those and send them to me?

    Best regards,

    Steve.

     

  • 0 in reply to SteveCrump
    Prodigy 75 points

    Hi Steve,

    Thank you very much!

    My AGND was directly connected to ground copper area under IC. Actually, I was playing around the AGND connection with several different ways, directly connect to GND copper area under IC, connect AGND seperately to audio ground in front of TAS5630 (Pre-Amp part), both did not work.

    I will send you my circuitry + PCB files later.

    Best Regards!

    Junpu

  • 0 in reply to Junpu Xu
    Prodigy 60 points

    it is simple in my experience, add relay between output  and loader, relay delay on circuit necessary indeed. we have finished products like this.

    Mr.lai

  • 0 in reply to Junpu Xu
    TI__Mastermind 18510 points

    Junpu, I have found another case with SE noise that sounds similar to yours.  In that case it seems the user has not grounded the heatsink or given it a good electrical connection to the Power Pad of TAS5630.

    This is a problem because the substrate of TAS5630 is connected to the Power Pad on the top of the device.  The heatsink must be well grounded and the Power Pad must be electrically connected to the heatsink so it can shunt any currents that enter the substrate to ground.  Without these connections any currents that enter the substrate can flow randomly and cause problems like noise.

    We recommend using a high-conductivity thermal compound instead of a thermal pad and especially instead of anything that insulates.  We prefer a compound called Ceramique from a company called Arctic Silver, because it has the correct properties.  I have inserted a data sheet for this compound.

    1033.Arctic-Silver-Ceramique-110615.pdf

    If you do not have a good ground to the heatsink and a good electrical connection between the Power Pad and the heatsink, can you try this and let me know if it improves the noise?

    Best regards,

    Steve.

  • 0 in reply to SteveCrump
    Prodigy 75 points

    Hi Steve,

    Thank you very much for your help!

    Actually, I do have an electrical connection between heatsink & power GND. I started without GND connection to Heatsink (and a insulation layer between heatsink to IC pad), it caused a lot of trouble (oscillation on top area of sin wave) when output power reaches around 70% of full power. A good connection between power GND & heatsink (heatsink directly touching IC pad) solved the oscillation problem.

    BTW, I got to know that TAS5630 is not recommended anymore for new projects, and TAS5630B is prefered. I got several TAS5630B samples, and tested with my circuitry. Noise problem is still there, without any noticeable difference... :(

    Can you please give me (junpuxu@gmail.com) your mail address, which I can use to send you my circuitry + PCB ?

    Thanks + Best Regards!

    Junpu

  • 0 in reply to XINLI LAI LAI
    Prodigy 75 points

    Mr.Lai said:

    it is simple in my experience, add relay between output  and loader, relay delay on circuit necessary indeed. we have finished products like this.

    Mr.lai

    Hi Mr.Lai,

    Thanks for your information!

    Does TAS5630 in your application works as SE mode, or 2.1 mode? I have no noise trouble in BTL mode, but only SE mode.

    Thanks!

    Junpu

  • 0 in reply to Junpu Xu
    Prodigy 60 points

    our products can works at 2 mode, 2.1 or 3way.pls check it here:

    http://www.aliexpress.com/fm-store/601158/210502886-522265351/DSP-CLASS-D-AMP-SMPS-2-1-OR-THREE-WAY-HI-FI-AMP-WITH-ANALOG-SPDIF.html

    1.Power out:

    2.1mode:1x150W/4ohm(subwoofer)+2x75W/8ohm or 4ohm (stereo)

    3way: 1x150W/8ohm(low)+1x100W/4ohm(mid )+1x50W(high)

    2.Input
    Analog:0dBV/100kΩ

    USB:1.1-2.0version

    SPDIF:coxial or optical

    3.Frequency Response:20-20KHz
    +/-0.5dB@1w

    4.Loader:2Ω/4Ω/8Ω

    5.AC Power:100-265Vac 50 or 60Hz

  • 0 in reply to Junpu Xu
    TI__Mastermind 18510 points

    Junpu, since your heatsink is well grounded and connected to the TAS5630 PowerPAD, that cannot be a source of the problem.  So maybe you are hearing crosstalk between channels rather than "noise", which sounds broadband.  I would like to see the waveforms that you are hearing, if that is possible.

    It is probably necessary to filter them, and I have used an RC+LC filter that is reasonably easy to build.  It is described in the report below, and you are welcome to use it.

    8304.Class-D-meast-RC+LC-filter-120411.pdf

    At 1kHz with PHD package and a good PCB layout crosstalk to SE channels is around -60dB, depending on load.  This also depends on PVDD ripple and frequency.  It is higher at 100Hz and a little lower at 10kHz.  These levels are low but still audible if they are within the passband of the loudspeaker used for listening.

    I would like to know if this is what you are hearing.  Can you measure and post the results?

    If that is not possible, please tell me how you are loading your outputs and how you are driving your inputs, and tell me what the relationship is between frequency of the "noise" and frequency of any active output.

    Best regards,

    Steve.

  • 0 in reply to SteveCrump
    TI__Mastermind 18510 points

    Junpu, I would like to check status of this case.  Have you been able to resolve the question, or determine if what you hear is normal TAS5630B crosstalk?

    Best regards,

    Steve.

  • 0 in reply to SteveCrump
    Prodigy 75 points

    Hi Steve,

    Sorry for late reply on this.

    I tried all possible solutions including your proposals, to be very honest, I did not find too much difference. Funny thing is our customer does think there is no problem ...

    I start wondering whether I am a little bit too picky on this.

    Please wait several more days, since I want get a chance to further verify this ...

    I will post my result whenever everthing is clear and settled. If it's really my over worring this issue, I will concel this problem report.

    Thanks!

    Junpu

  • 0 in reply to Junpu Xu
    TI__Mastermind 18510 points

    Junpu, thank you for this reply.  I think what you say indicates that what you hear is normal crosstalk in TAS5630 in SE mode.  It can increase with high PVDD ripple, and in a case like that reducing ripple will reduce the crosstalk.  Otherwise there is not much that will change it.

    We have a number of customers using TAS5630 in 2.1 and 4xSE applications, and they find that crosstalk from the BTL channel is not a significant problem.  The BTL channel usually provides subwoofer output, at low frequencies, and crosstalk from the subwoofer output generally is attenuated by the mid-high frequency satellite speakers.  Maybe this is why your customer is not concerned.

    I will naturally be very interested to hear what you conclude when you complete your investigation.

    Best regards,

    Steve.

  • 0 in reply to SteveCrump
    Prodigy 10 points

    Hi Steve,
     
    I like to pick up the problem Junpu mentioned already. We are using the TAS5630 with 2.1 mode as platform for different projects. Meanwhile we built several thousand units. Reliability is absolute great.
    We of course noticed normal crosstalk between the channels.  Even the virtual GND  for SE-amps created  by 2 x capacitors opens the door for crosstalk through modulated VCC. But this is acceptable and no problem.
    Problem is: Each signal which crawls from BTL to SE channel is causing additional strange small spikes of few mV at the active output of both SE-channels.  The virtual GND keeps clear. Unfortunately this  spikes have a very wide bandwidth and they are audible. With normal music you won´t  hear anything because the spikes are simply masked by the full music signal.  But listening to a pure bass guitar the annoying spikes are audible immediately because of their much higher frequency range.  Effect sounds a little bit like a whirring piece of metal after a hammer stroke.
    Mentioned problem can be created like this :  
    -              Mute audio inputs of both SE-channels. Either by small change of DSP-software or   -to make it more save-  by hardware interrupt directly in front of TAS5630
    -              Connect 4 Ohm dummy load to BTL-channel
    -              Connect  normal speakers to both SE-channels
    -              Drive BTL-channel with 60 Hz sine signal
    -              Vary input level of 60Hz sine signal and listen to the signal from both SE-channels. 
    We have searched for the reasons for a lot of hours i.e. : Changing GND and VCC tracks manually, adding capacitors, replacing class D-coils by other types. The effect cannot be influenced.  Even the use of a laboratory DC-supply does not influence anything.  Different feed points for VCC also don´t  influence the effect.  So it seems to be no layout problem, no VCC-problem and no problem caused by magnetic field of filter-coils.
    Meanwhile we are running out of ideas. It would be absolute great if you can support us with a fresh hint.
     
    Thanks a lot for your efforts and best regards!
    Mario