TPA3116D2 + Arduino Nano with AD9833 issues

[thedeadliestcatch]

Hi,

The pinouts for the Arduino Nano should be A0 for FSYNC and then the standard SPI pins (10.11.13)

I'm testing mic-jammer-ad9833, intending to do some improvements, but with the wiring as is I do not see any output from TPA3116D2 (it is wired properly as far as I can tell), consuming only 0.9A at 12V. Maxed the pots for the volume.

Tested two TPA3116D2 boards.

It's a standard Nano clone. The transducers are mounted in a (tested) PCB, 25 per board. All wiring P/N is correct there too.

The TPA3116D2 emits a hissing noise when the Nano finishes booting.

[thedeadliestcatch]

Note: no effect on nearby mic, and the consumption at the PSU is abnormal. Any ideas?

[mcore1976]

1. What are the transducers. 40KHz or 25KHz? There are separated versions of the source code per each type.
2. Do you have an oscilloscope and can you see the output signal of the AD9833 AOUT/AGND pins ?
3. Are you using sinusoidal ( AD_SINE) type of the signal in the code or pulse type (AD_SQUARE) ?
4. Some chineese sellers are selling fake TPA3116D2 modules ( XH-M542 ) that do not work above 20kHZ, I have recently found 2 of 3 delivered boards not working with piezzo transducers..

I have checked with Arduino Pro Micro this design which has the same chip as Nano ( ATMEGA 328P ) and was working 100%

[thedeadliestcatch]

Hi,

1. 25Khz, and they are verified 25KHz, polarity is also correct (tracking down datasheets is a chore but the ones with the insulated shield -not connected to either pin- have a plastic bevel in one of the pins).
2. Yes, will diagnose with the scope soon, but the AD9833 is functioning as it should as far as I can tell.
3. Default from sketch, I've considered using more than one AD9833 and drive separate channels with two different waveforms.
4. Do you have any details about fake TPA3116D2s? Did you remove the heatsinks?

By the way, I am doing the testing with 25 transducers. Designed my own PCB for them, also for heat dissipation purposes. Range is not huge in a closed space, a few meters.

Have you played with phasing at all?

[thedeadliestcatch]

Ah, forgot to ask: any tuning/adjustment tips appreciated. I think distortion is an issue at some point. I'm driving the amplifier with 13V~, have you tested any practical differences with 24v? (will need a switching PSU/DC-DC for the MCU in that case).

Any idea re DC line noise? This is a breadboard proto right now, so, maybe thats a problem too.

[mcore1976]

Some of TPA3116D2 ( XH-M542 ) module has been torn apart by one of the buyers which is visible in www.aliexpress.com review section : https://ae01.alicdn.com/kf/A400a250df4ea4b1fbdd12765ab9dc5edY.jpg
However I have bought couple of boards from the same seller and they were just fine. Maybe it happens from time to time or they were simply faulty DOA.

The best is to use actually not 25KHz transducers but 20kHZ transducers - they are very expensive but provide couple of meter range of jamming for apple devices ( link https://www.aliexpress.com/item/1005001670524154.html ) . To use it you need to modify the source code in random function from (24000,26000) to (19000,21000) as well as values in while loop.

About power source : if you use oscilloscope you will find that with 12V power source you are getting almost 60 Volts on ultrasonic transducers. If you increase this value they may overheat and break, which mostly happens with power source around 20V. So be carefull and observe maximum allowed Vpp on transducers (consult with transducers vendor)

I have experimented with setting random Phase in AD9833 signal, it helps a bit but the signal is more audible.

[mcore1976]

[mcore1976]

Above is how you can identify proper TPA31186D2 mono modules that work properly with the piezzo transducers