Thanks for the advice Doctormord
In parallel, I was looking at the different components of the power supply stage of the TPA3251EVM, TPA3255EVM and TAS3251EVM.
I notice that the LM5010 exists in HTSSOP package, but has a downward thermal pad. This is I imagine difficult to handle for hand soldering. Would there be some more suitable option for hand soldering, with similar performance?
Best regards,
JMF
LM5007 seems to do the job, from V and I and I perspective.
It seems that the TAS3251 draws typically 90 mA from the 12V. The LM5007 can deliver up to 700 mA.
Thanks !
JMF
Hi AIM65,
I'm finishing the reading of the PCB design tutorial and it is very instructive and well done. It helps grasp the basics, which is really important. A valuable reference for me !
JMF
MP2459:
https://www.monolithicpower.com/pub/media/document/MP2459_r1.01.pdf
It's pin-compatible (need pullup on /EN) with the TI LMR16006 at half the price with a bit lower switching frequency. I used LMR16006(Y) in the past at 700kHz and 2.1MHz without issues. Actually i use the MP2459 with my LTC3784 4-phase boost-converter.
Ok ok.
Here's the results for the TPA3128 on a 1.0mm board when knocking at the side edge of the PCB with some Murata 2u2/25V MLCC in place (input coupling):
We then modified the mounting to vertical mounting like so that the MLCC is standing like a tombstone on one pad and is then wire-connected to the other pad. The resulting improvement is shown here:
As already written above, the mechanical influence can be reduced when mounting the MLCC with their long body side to the short side of the board. It is a good practice to keep this in mind when using MLCC. As a side-effect/benefit, flex-cracking-risc is much lower when doing so. (For the latter, flex/soft-termination-caps might help as well.)
The device under test:
Attachments
Skip switching converters and use a higher feed + passive converters I instead. There is enough hf to battle anyways!
?
//
?
//
Correction: It's the TPA3132D2 on the board, not the TPA3128.
Why? Voltwide and i use the LM5010 without further LDO filtering for the AVCC stage and theres no degradation in THD but YMMV. Having problems with HF is mostly due to improper grounding loops and PCB layout.
Last edited:
X7R?
PS: Just tried to knock my HiFi-TOY PCBA - no effect at all, S/N steady stay -104db(A). that's why I like "Power DAC" topology 😉
it is not what I did, I talking about native Ti chipset TAS558+TAS5624, so Ti did that immunity for piezo caps. I believe it is because Ti used silicon capacitors for the loop filter, aka noise-shaper, together with I2S it gives excellent immunity.
hence, for high-resolution audio ability need to add external sample rate converter IC(and EVB has it!). I quickly losing my exciting regarding TAS3251..
The TI answer is ok I believe "The ultra HD was in reference to the THD+N and SNR performance" as these 2 parameters has much more influence of perceived resolution than the fs.
//
//
I had seem that. However as my music is 99.9% 44k, this is fine for me. In my current config, my music server (mpd on Linux, in an OrangePi) has the ability to perform the sampling rate conversion. This is working very well.
Is there really 192 kHz music available yet ? I will develop a new project when it will be available 🙂