Hi. I hope someone can help me out.
I bought the MA12070 board from Audiophonics (MA12070 board) and I'm gonna use it in BTL mode in order to drive 2 top speakers. My input is balanced out from a 3E DSP board.
There's only 1 ground input on the amp, does that mean that I have to wire the 2 ground connections from the DSP board together, and otherwise connect + and - to the respective inputs?
I bought the MA12070 board from Audiophonics (MA12070 board) and I'm gonna use it in BTL mode in order to drive 2 top speakers. My input is balanced out from a 3E DSP board.
There's only 1 ground input on the amp, does that mean that I have to wire the 2 ground connections from the DSP board together, and otherwise connect + and - to the respective inputs?
Yes, just wire the grounds together at the amp and connect the + outputs to IN0A/IN1A and - outputs to IN0B/IN1B.
@altus fidelitas - your amp looks brilliant, thanks for the pics!
Would you be open to sharing the schematics/layouts? I'm planning on building a multi-channel amp using the MA12070P and Linkplay A31 source, and would love to use your layout as a starting point if you're open to it?
Hi all, wondering if anyone can help diagnose a noisy Shenzhen Pillor MA12070P board (PL-AD-160, marked V1.4)?
Thanks to some great tips earlier in this thread (nice even the designer has dropped in) I finally got this board working, hooray! At first I thought problems were all software related - I am connecting to an Orange Pi One SBC running Armbian which lacks the drivers available on the Raspberry Pi and trying to port them got a bit tricky. But fixing the PCB population errors to remove caps on the I2S input pins, connect master clock etc I now have a working I2S connection and can manually control volume and power mode, etc over I2C without the complete Merus driver.
However the board produces a disappointingly loud high frequency hiss from my (desktop) speakers, audible on top of music at low volumes. It is much worse than the other class D amps I have. Again I first suspected software, the volume control adjustment on input or mistakes over format conversions, but the I2S looks good and hiss remains constant if I mute the channels over I2C or use I2C volume control. (Using the MUTE hardware pin or stopping the I2S stream turns off the speakers). I've tried a range of PSUs/voltages and noise varies slightly but is still the same volume even with batteries. I'm using 2.0 BTL which shouldn't need blocking output caps according to the datasheet. It is as if the gain is set too high, but I2C doesn't help with that - I can increase the gain with register 0x25 but that just makes hiss worse.
Has anyone succeeded to get a clean/clear output from this board variant, are more alterations needed?
Thanks in advance.
Thanks to some great tips earlier in this thread (nice even the designer has dropped in) I finally got this board working, hooray! At first I thought problems were all software related - I am connecting to an Orange Pi One SBC running Armbian which lacks the drivers available on the Raspberry Pi and trying to port them got a bit tricky. But fixing the PCB population errors to remove caps on the I2S input pins, connect master clock etc I now have a working I2S connection and can manually control volume and power mode, etc over I2C without the complete Merus driver.
However the board produces a disappointingly loud high frequency hiss from my (desktop) speakers, audible on top of music at low volumes. It is much worse than the other class D amps I have. Again I first suspected software, the volume control adjustment on input or mistakes over format conversions, but the I2S looks good and hiss remains constant if I mute the channels over I2C or use I2C volume control. (Using the MUTE hardware pin or stopping the I2S stream turns off the speakers). I've tried a range of PSUs/voltages and noise varies slightly but is still the same volume even with batteries. I'm using 2.0 BTL which shouldn't need blocking output caps according to the datasheet. It is as if the gain is set too high, but I2C doesn't help with that - I can increase the gain with register 0x25 but that just makes hiss worse.
Has anyone succeeded to get a clean/clear output from this board variant, are more alterations needed?
Thanks in advance.
Hopefully I'm not confusing your board with another one, but I was under the impression that master clock should not be connected. See post #524.
FWIW: Got a Aiyima A8 with a single MA12070 and it outperforms my DIY amplifier. It has drawbacks like many opamps and ceramic caps but it still sounds better. Bummer. When you switch it on the first seconds the air core coils make some noise 🙂 Also no caps after the coils which is awkward. It has a complete useless "bass boost" function and a non standard appearance which I like. When an A8 is bought make sure to clean the PCB as it was relatively dirty with flux etc. For the price I paid it has an excellent price/quality ratio.
So back to the drawing board. it will be a MA12070 module based amplifier driven by transformers so a passive input section. Linear Power Supply too.
So back to the drawing board. it will be a MA12070 module based amplifier driven by transformers so a passive input section. Linear Power Supply too.
Last edited:
Guys - the MA1270 is a great chip but the new MA5332MS strikes me as even more musical (more full-bodied sounding), more precise soundstage and no fatiguing highs.....and this in an unmodded amp....
MA5332MS is a new chip but looking at the block diagram it seems to be IRS2092 repackaged in stereo form with on-chip MOSFETs.
Hi Jean Paul,
I just bougt this amp.
Do you think the input is balanced?
I want to skip all opamps and feet the chip from AK4493 balanced via Transformer
Klaus
Yes it was so cheap that I had to try it out. Strong suspicion that the tube in the preamp is just a gimmick.
Yes the AM12070 itself of course is balanced input but I put it back together. Same idea to skip a few opamps.
The circuit of “bass boost” can be omitted just like the “active output”.
Last edited:
Thanks for the suggestion. But if I disconnect (M)CLK I get no sound, this was one of the original problems to start. I think post #388 is accurate. Also if you check Infineon's circuit diagram for their RPi Zero Hat it connects both MCLK and SCK to BCLK from the Pi.
I'm now wondering if the 5V AVDD/VDD from the TPS62175 buck converter on this board is the cause of the noisy output. It looks like the voltage divider used for feedback has three times the recommended resistance in the TI datasheet, which remarks that higher currents are better for noise immunity. Or maybe there is some bad modulation with the regulators used on the SBC.
Did you get ESP32 squeezelite working with this chip in the end, by the way? That was my original idea for driving this board!
Yes I did, but not using that board. When they released the 32 bit version of Squeezelite-ESP32, I installed it and it worked! Later, I realized that it only worked because the 32 bit version created a BCLK that the MA12070P accepted. Using the 16 bit version, a 44.1K sample rate gave me an invalid BCLK.
By forcing a sample rate of 88.2K or 96K the generated BCLK became acceptable.
They also added MCLK as an output, and that worked, but I got more noise using it so I went back to shorting BCLK and MCLK together and feeding BCLK only.
I haven't fired that amp up in a while, but I would be happy to do so and send you the NVS parameters that worked for me.
Mike
A look inside. Currently feeding the amp and buck converter from a 12 volt battery. The 5 volt output from the buck converter powers the ESP32.
