• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 1 answer
  • Subscribers 31 subscribers
  • Views 656 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.

TAS5630B runs warm?

Hernan Freschi
Prodigy 180 points
Other Parts Discussed in Thread: TAS5630B, TPA3116D2, LM317

I built a test PCB for TAS5630B - the layout is not ideal and I'm aware of that. But I decided to try it anyway with small supply voltage and found that the device actually runs quite warm, under the same conditions (supply voltage = 25V and load = 6 ohms), compared to a TPA3116D2 board I built the other day with outstanding results (after 1 hour of music at maximum volume it was barely warm. The heatsink must have  been at 36-38C max).

The TAS5630B, on the other hand, was getting much hotter even at low volume setting. The heatsink was about 45-50C.


Is this normal for this device to be so warm?
Or should I look for faults in my layout?


Some things that are out of spec in my build are:

1. I don't have any 10nF capacitors rated at >25V so I'm not using them today. Waiting for them to arrive.

2. I'm using an output LC filter composed of 22uH/330nF capacitors. I have a surplus (several dozens!) of power inductors of this value. How does this affect the performance vs. the recommended 7uH or 10uH inductors? I see in the datasheet this is a "minimum" value.

3. I'm not using the ouptut snubber network (again, no capacitors rated for 100V in this case)

4. I don't have any 2.2uF MLCC capacitors for GVDD decoupling. I'm using two 1uF in parallel, but they're unknown SMDs. I don't have any specs for them, except they are 1uF 50V.

The rest of the board is within spec, as far as I know.

  • 0
    TI__Guru**** 189401 points

    Hi, Heman,

    I suspect the issue is your output filter. Please see this app note for more information on the output filter.

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 180 points

    Don,

    I'm reading the appnote but I don't understand the following:

    On page 4, the appnote states that it's recommended not to use a filter that's peaking. but the recommended values for inductors do that. For example on the datasheet they use a 7uH/680nF combination. This, according to equation (2) has a Q of (assuming speaker Z of 6) 6*sqrt(680e-9/7e-6)=1.87. Only when Z is <2.3 the filter is not peaking (Q < 0.707). What Z are the datasheet filters calculated for?

    On the other hand, starting from the R, L and C components I'm using (6 ohms, 22uh and 680nF), I get a F cutoff of 41KHz and Q=1.05.

    Now, all these inductors I found in consumer audio boards. All of these have 6 ohm speakers. And i've looked at several brands and noticed all of them use 22uH inductors. So I decided to fix my inductor at 22uH and Z=6 ohms, and vary C, surprise, I found that the most approximate filter for 6 ohms Z seems to be 22uH/330nF, calculated at Q=0.734 and F=59KHz. The ideal inductor/capacitor, calculated by your appnote, agrees that for Q=0.707 and Z=6, C should be 312nF and L 22.5uH, so 330/22 is the most approximate (and yes, the capacitor found in audio boards made by LG and Philips, using some TDA part from NXP, is 330nF).

    So in short, I think the values in the datasheet are calculated for Z= 2 ohms. Then, how do these systems behave with higher output impedances?

  • 0 in reply to Hernan Freschi
    Prodigy 225 points

    I recommend you download the free TINA spice simulator from TI.  You can then model the output filter to check the response into different load impedances.  22uH is ideal for an SE application, but is too high a value for bridge use, unless your load impedance has a minimum value or 8 Ohms or so.

  • 0 in reply to Chris Morriss
    TI__Guru**** 189401 points

    Hi, Hernan,

    We have found that some inductors have more loss than others, so in addition to Chris's recommendation above; you may want to swap out different inductors to see how they behave.

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 225 points

    Very true!  A lot of people try to get away with the smallest physical size of inductor, which will tend to have much higher core losses.  (The core of the inductor is being continually magnetised one way then the other as the high frequency switching current passing through the inductor increases and decreases.  This obviously can cause eddy current heating if the core is not an insulator, but even worse, the hysteresis losses in ferrite cores can result in quite a lot of self heating.  At low powers this has a much greater effect than the I^2 x R losses in the wire.

    It is worthwhile measuring the supply current at idle (on an AD-mode design in particular) as the idle current can be a good indication of the core losses in the output inductor.  I saw the idle current drop by 30% in my TAS5630B design by changing inductor types.  This was with the chip in AD mode.  The effect may not be so noticeable (at least at idle) in BD mode.

    By the way:  Anyone thinking of using the TAS5630B in the DKD package in AD mode should read my cautionary tale in the thread I posted here a few days ago!

  • 0 in reply to Chris Morriss
    Prodigy 180 points

    For the record: These are the inductors I'm using with TPA3116: http://i.imgur.com/mTRNkQM.jpg

    And these are the ones I'm using with TAS5630B: http://i.imgur.com/l5F2eL3.jpg

    The little black ones are DR127 type from Cooper-Bussman (now Eaton), rated for 4A. The larger ones are some inductors I recovered from a "mini component system", and all the information I have about them is that they are "22uH 9A 500V". There's no more info.

    The supply current on idle, at 24V, is 90mA (this includes the LM317 regulator I'm using for 12V)