WaveTooEasy issue

Support
#1

Hi.
Before I start with the issue, I have a question regarding the WaveTooEasy latency.
How did you measure the latency?
What is the expected maximum latency? I suspect it can vary a little bit, especially when playing many sounds at once.

Now for the issue I have with the board.
I’ve soldered a USB lead to the board on the EXT5V and GND pins (power supply pins).
I plug the USB lead to my laptop, it’s all good.
My goal was to modify the firmware, so I used the SD card and put a wte.bin file at its root.
Did that many times, no issues. At some point I reverted to the original firmware, but when I plugged the USB into the laptop, nothing happened. The LEDs stayed off.
Tried with and without the SD card inserted, same thing.

It is possible that the two wires between 5V and GND touched each other, but I’m not sure.
I have tried to use the AK-FT232RL to get a LED to turn on the board, but nothing.
So the board doesn’t seem to have power and nothing is working any more.
Any ideas as to what’s happening? And how to fix this?

I have a STLink-V2, I tried to hook up its 3.3V and GND to the JTAG port, but same thing, no LED is on.

Thanks for you help.

#2

Hi @spin,

It’s been a while since we characterized the board. If I recall correctly, latency was measured with an oscilloscope probing between an input trigger and sound output. Of course latency depends on many factors as quantity of sounds playing as you say (but it didn’t vary much), quality of SD card, sampling rate, etc.

Regarding the dead-board problem. It can be a shortcircuit. The board has 2 DC regulators on board. If the 5V regulator is gone, you can try and use the board using 3.3V power supply through 3.3V pin. That bypasses the 5V regulator (audio will be lower).

If you suspect it’s firmware/software, you should be able to probe the board with a JTAG (you don’t have to download anything, just check if the MCU is being seen by your JTAG software).

#3

Hi @Ivan,

Thanks for the response.
When you say latency didn’t vary much, do you mean something like less than +/- 1ms?

I have tried to power the board using the ST-Link V2 3.3V, but without success, still no LED blinking whether the SD card is there or not, and with and without wte.bin on the SD card.
I will try using the JTAG port and let you know.

#4

If I recall correctly, it was under 10ms.

Regarding the JTAG 3.3V pin, please note this pin is usually for the JTAG to detect the target voltage. It can’t be used to delivery power to a board. It’s an “input”, so to say.

You need to provide 3.3V from a stable power supply.

#5

Hum, 10ms is a lot more than 4.5ms, I need something that stays below 6ms.
Are the sounds cached in some way?

I haven’t tried the JTAG to check if the microcontroller is responsive or not. I’m temporarily leaving that aside, because I have custom boards that work at the moment. My guess is the 5V regulator died, it’s probably my fault anyway…

#6

10ms if I recall correctly is when playing several audio files at the same time using IO mode.
A single sound file should be around 4.5ms.
There is no cache except for a few samples. It’s a microcontroller project, with limited RAM so SD card access is key for good latency.

#7

Yes, that’s not too bad for such a tiny microcontroller to be honest.
I measured the latency on my custom board, it’s around 5ms when there’s only one sound to play. I use a PCM5102 which adds about .5ms of latency, so that’s on par with your 4.5ms measurement.
So, well done, because this is the lowest latency I’ve found for such a low power microcontroller.
The only thing that’s missing for me is to switch the ‘busy’ variable back to false when a file finishes playing (in the Player class). It shouldn’t be too hard.