Infineon MA12070 Class D

The board looks generally a bit odd: they are using the values instead of designators on the silkscreen, no grounding vias near the caps and it's quite hard to understand the track routing and component placement. If this is not a 4 layer board, I assume there are a lot of traces on the back side of the board, and without proper stitching, that creates GND islands that might cause problems.
Are the caps labelled 47u6v3 the supply caps? if so, they seem to be 47uF 10V, according to the ref design, they should have 100uF 35V. This could potentially introduce PSU noise into the system.

1659470639265.png

Infineon-UserManual reference design: page 11
 
Last edited:
Disabled Account
Joined 2002
@weissi: Bourns PTD902 can fit. Apparently Alps RK097 was also made in 50 kOhm log in the past as I found a few in my stock.

Apart from the 22 kOhm resistors that should be 100 … 200 kOhm I implemented the proposed changes but I am not fully satisfied yet. I used OPA2156 as I had no other opamps but today OPA1642 came in :D
 
Last edited:
  • Like
Reactions: 1 user
Thanks for your replies @theAnonymous1 and @jean-paul: Well, I don't even know, why I've asked, as my basic plan is to drive the MA12070 by 4xTDA1541As driven by ADAU14xx in symetrical right out of the DSP.
Well, modding of the BRZhifi goes on, I'm in the process to fit a step-down switcher from 24V to ~8V going into whatever linear regulator for 5V. I've set for an ordinary 7805 right now for the low-freq. channels but will compare it with the high-freq.'s PT6818 x2 if there's any noise to be recognized within 20cm distance. Might try low-noise regs as well.

Also on the list: replace the "fake" Elna RVT 100uF by Chemicon HXEs. I suspect them ELNAs to be fake - just because. But then, I mentioned it before, ELNA is not a tier #1 provider for industrial electrolytics.
 
Last edited:
@weissi

are you planning to do your own amp design?
Infineon sells an affordable PI version, if you choose to use a DSP anyway you are not in need of a DAC chip and can feed the ma120xxp directly via I2S. With good planning of the PSU filtering, you can use a 12, 15 or even 24 V supply rail without having a big step-down converter and just have a small and efficient regulator for the 5V rail.
IMHO you would not get benefits from a high-quality DSP using a board without a proper output filter. (depending on your application) The LC output filter design considerations have already been discussed in this thread, I think someone did a separate board for the MERUS chip, that might come in handy improving the filterless BRZhifi design.
 
Disabled Account
Joined 2002
That board was the first MA12070 board I bought. Ran into trouble as the casing I meant to build it into is too small. It looks way better indeed. Output filtering is datasheet minimum standard with ferrite beads and 2 x 1 nF.

@Solve: no it sounds quite good as it is but I think a DIY amplifier with MA12070 is the best solution. All the boards are cheaply built. Please note Infineon to mention production costs to be $ 2.15 (!!!) for a complete board based on 1000 pieces. Yes that is parts, PCB and assembly. It is a real miracle it all sounds so nice as an amplifier board cost less than 1 good cap :D
 
Last edited:
  • Like
Reactions: 2 users
Ok I ordered one BRZ as addicted to cheap amps.
So now I have one A8 and one BRZ in an active two way system.
PC, Komplete Audio 6, SPDIF out, nanoDIGI, two Khadas Dacs, "Bitrate perfect" so software volume no problem.
Amps at full volume.
So my source is ok.
Speaker, XRK Tl with Purifi and Dayton RST-28F.

If there any TPA32xx with this 3D sound I buy it!
3e don't have it.

This just sound right to me.
Not for all...
Ok it's not nirvana the BRZ at HF have an piezo sound on some songs that also could be mixing.
The bas is musical not excessive as TPA.
The middle is ok more laid back which I like.
But this chip have some nice pontential!