Audio turns off after a while, when playing 16-channel file from SD card

Pavledin

Hello!

I'm new to the Bela world and this is my first post, so I'm sorry if I'm missing something rather obvious.

I've purchased a CTAG Beast for the simple task of playing a 16-channel audio file (around 10 minutes long) in an endless loop for an embedded audio installation. I'm using PureData to program the board, here's a screenshot of the patch:

Playback is fine for a while, but after some time (sometimes it's 24 hours but sometimes it's less than an hour), I get this message:

audio frame 5756928, errorCode: 3
McASP unexpected transmit frame sync occurred

The last time it happened, audio playback resumed, but 99% of the time it stops all audio coming out of the board, and I have to recompile in order to make it run again. I've tried changing the block size, but it hasn't solved it. CPU usage is around 30%. The file is a 16-channel 44.1kHz 16-bit WAV file.

I'm waiting for a new SD card today which will be a SanDisk Extreme Pro, but the card that's been in there so far has worked well for me in a camera before.

What could be the cause of this behaviour? Am I doing something wrong on the PD side, or is there a problem with the way I'm configuring the Bela? The installation is due new week so I'm kind of in a tricky situation.

Thanks for your help,
Pavle

giuliomoro

Can you try updating the Bela core code to the dev branch? That's explained here using this zip archive.

That should fix it.

EDIT: even updating from the master branch should work.

Pavledin

Will do, thank you for the quick reply! I'll write back when I've tested that out.

Pavledin

I've updated the Bela software via the instructions above, but now I get these errors when attempting to load my Project:

make: execvp: ./resources/tools/detectlibraries.sh: Permission denied
make: *** [/root/Bela/projects/PD_PlaybackTest_01/build/Makefile.inc] Error 127

giuliomoro

That looks like something weird happened while updating. Try running

chmod +x /root/Bela/resources/tools/detectlibraries.sh

in the console at the bottom of the Bela IDE and then build again.

Pavledin

Thanks Giulio! It's loading the project again as before.

Let me run the setup for a while and test if the original issue is resolved, and I'll get back to you.

Pavledin

Hello again, I'm back after some testing 🙂

I've run it for a couple of days now, and the initial errors haven't been showing up, but I've been getting a few of these, which stop the board entirely and require a reboot:

PRU interrupt timeout, -1 110 Connection timed out
PRU interrupt timeout, -1 110 Connection timed out
audio frame 6618880992, errorCode: 1
PRU event loop timed out
PRU interrupt timeout, -1 110 Connection timed out
PRU interrupt timeout, -1 110 Connection timed out
McASP error, abort
PRU interrupt timeout, -1 110 Connection timed out
Makefile:613: recipe for target 'runide' failed
make: *** [runide] Error 1
Bela stopped

I'm don't know if they're related to the previous ones in any way or if they're something else entirely, and if I should create a new post about this?

giuliomoro

Can you tell me which CTAG cape in your stackup is configured to act as leader and which as follower? https://learn.bela.io/using-bela/technical-explainers/configuring-ctag/

Pavledin

The bottom one in the stack is configured as leader (the one connected to the BeagleBone), and the next one is the follower (the one on top of the stack). So the full stack is (from bottom up): BeagleBone / CTAG leader / CTAG follower.

And both of the CTAG capes are configured for external power, rather than USB.

giuliomoro

was full board reboot needed or was it sufficient to restart the program ?
Can you associate this with anything you did on the board while it was running ? e.g.: plugging/unplugging a cable, or switching on/off an heavy electric load (lights, motors), nearby?

When it happens again, run in the console of the IDE dmesg | tail -n20 and show us what you get there, please

Pavledin

giuliomoro When it happens again, run in the console of the IDE dmesg | tail -n20 and show us what you get there, please

Thanks, will do!

giuliomoro was full board reboot needed or was it sufficient to restart the program ?
Can you associate this with anything you did on the board while it was running ? e.g.: plugging/unplugging a cable, or switching on/off an heavy electric load (lights, motors), nearby?

I'm not entirely sure about the reset case, if it happens again I'll test that part out. The first time it happened, I was indeed fiddling around with the cabling since I've been fitting the board inside a printed case. Also, at that time, the USB controller on my MB might've been clogged, messing up the voltage of the USB lines, since I was doing a large file transfer to/from a hard drive connected to the same set of ports.

But the last time it happened I just found it like that in the morning after leaving it to run overnight (the hard drive was still connected though, but no transfer was going on).

giuliomoro

OK, let's see if we can get any meaningful information from dmesg. This is unrelated to the previous issue, btw.

Pavledin

Thanks Giulio, should I create a separate post for this then?

giuliomoro

as you prefer, I don't mind either way.

Pavledin

Ah, it happened again, here's the log:

[ 9.057487] LUN: removable file: (no medium)
[ 10.288368] usb0: HOST MAC be:1a:24:86:bf:cb
[ 10.288830] usb0: MAC be:1a:24:86:bf:cc
[ 10.290118] usb1: HOST MAC be:1a:24:86:bf:ce
[ 10.290323] usb1: MAC be:1a:24:86:bf:cf
[ 10.561783] configfs-gadget gadget: high-speed config #1: c
[ 11.126814] net eth0: initializing cpsw version 1.12 (0)
[ 11.205990] SMSC LAN8710/LAN8720 4a101000.mdio:00: attached PHY driver [SMSC LAN8710/LAN8720] (mii_bus:phy_addr=4a101000.mdio:00, irq=POLL)
[ 11.227393] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[ 12.420262] random: crng init done
[ 12.420287] random: 7 urandom warning(s) missed due to ratelimiting
[ 12.963268] IPv6: ADDRCONF(NETDEV_UP): usb1: link is not ready
[ 14.518605] rtdm_pruss_irq_open
[ 14.533604] rtdm_pruss_irq_open
[ 968.556858] rtdm_pruss_irq_close
[ 968.556927] rtdm_pruss_irq_close
[ 969.685496] rtdm_pruss_irq_open
[ 969.687225] rtdm_pruss_irq_open
[22069.409628] rtdm_pruss_irq_close
[22069.409667] rtdm_pruss_irq_close

Pavledin

Also, there was no need for a full board restart, it was enough to Build and Run again.

giuliomoro

That's annoying because it doesn't actually show anything useful. Try to update to the fix-ctag branch: https://github.com/BelaPlatform/Bela/archive/refs/heads/fix-ctag.zip

That should address this issue.

Pavledin

giuliomoro Try to update to the fix-ctag branch

Will do, thanks! I'll run it again afterwards and get back to you.

Pavledin

I've been running it for a few days and it seems to be ok (fingers crossed) 🙂 I'm setting the installation up over the weekend so we'll see how it goes in a real-world environment.

Thanks a lot for your help!