I’m running into a persistent, randomized audio issue on a fully assembled 89Sabers Anakin Ep3 running a Proffieboard v3.9, and I could really use some hardware or software advice.
Randomly on boot, all audio stretches out into a deep, low-pitched, slurred “slow-motion” playback. Once it happens on a boot, it stays stuck in slow-mo until the next hard reset, where it might boot up fine or trigger the glitch again. It feels exactly like a clock/timing desync or a capacitance issue.
Here is the data and what I have already ruled out:
The Trigger: This problem started right after I reflashed the board. To troubleshoot, I reflashed the original factory config that worked perfectly when I first received the saber, but the randomized slow-mo issue is still happening now.
Cross-Board Verification: My other custom configuration layout works great on my Proffie v2.2, so the underlying code structure and files are verified stable.
Power: Testing strictly on a fully charged 18650 battery at 4.2V (completely disconnected from USB).
OS Version: Running the latest ProffieOS v7.15, so the early v3.9 I2S clock alignment patches should be active.
SD Card Performance:sdtest returns an average speed of 1730.51 kb/s and 19.62 simultaneous audio streams.
Cross-Card Test: Swapped in a known-good SD card from another perfectly working saber. The slow-motion glitch still occurs randomly.
Internal Flash Test: If I pull the SD card completely out to trigger the internal flash storage, the spoken “Font directory not found” error voice still plays in slow motion.
Because the internal error voice is slurring even without an SD card on OS 7.15 and a full battery, it seems like the central CPU clock and the audio peripheral clock are failing to handshake or the crystal oscillator itself is failing to stabilize.
Has anyone seen this persistent slow-mo glitch on a v3.9 even on modern OS versions? Is there a specific define or clock setup I should look at in the source code, or is this indicative of physical hardware/crystal degradation from chassis flex near the micro-USB port?
This is most likely a clock circuit problem. The clock circuit is somewhat sensetive, and having wires, metal, plastic or fingers can cause it to not work properly, and when that happens the CPU runs at a lower speed.
For me, this mostly happens when I’m testing stuff and accidentally put my fingers on the clock circuit.
I think a little capton tape on the CPU and clock circuit would be enough to prevent the problem, but I’m not really sure if it’s caused by electrical or EMI interference, so I can’t be sure.
Yeah, I had researched the other threads where you talked about that before, but my fingers are far from it and it seems to free of being touched by anything else. Is it permanently damaged from me touching it a few times while testing? Also, the board is glued to the chassis, so I can’t really see the other side if anything is messing with it. Is my next order of business to put tape over the circuit, or clean the circuit, or rip it from the glue? Or is it possible that the speaker is just faulty. I will try to upload a video tomorrow when i get the chance, but thought I’d ask this now.
Thank you always for your help, I bought this second saber because of the amazing things I’ve been able to do to my first one thanks to you and Fett <3
While it’s certainly possible to damage a circuit by touching it, that has not been my experience with the clock circuit. I suppose it could be some dirt or something causing a more permanent problem. It’s also possible that the components that make up the clock circuit are not good, or the soldering that connects them.
If you have access to the “top” of the board (where the clock circuit, USB connector and buttons are), then I would try cleaning it with some IPA (isopropyl alcohol) first and see if that helps.
Second, I would record a video of the problem. We should be able to tell the difference between a bad speaker and a bad clock circuit.
If the clock circuit turns out to be bad, do you have warranty on your board?
What version of the arduino-proffieboard plugin do you have?
I am running version 4.8 of the proffieboard plugin.
I cleaned it with the IPA and a Q-tip, no difference unfortunately after I dried it and let it dry for a while.
I’m just confused on why it worked so well originally but then after the first flash it was different.
Here is a video comparing my working proffie 2.2 and the slow 3.9. They’re running essentially the same config and have the same files and everything:
I’m not sure if I have warranty or not, but I sent CC Sabers an email asking.
That is weird, and I’ve never heard a board do that before. To clarify, that doesn’t sound the same as touching it with a finger. Touching results in something more constantly slow and low, as if the sample rate is halved or “playback speed” is 1/2.
Yours is like a modulated pulse, as if you were touching it / untouching it hundreds of times per minute. If I had to take a wild guess, the L1 chip is bad.
But, maybe a speaker wire is contacting something? That’s a PWM signal that might do something like that? But those pads are across on the other side of the board from the chip. Hmmm. Don’t know, sorry.
CC Sabers said this:
"With the blade effects running smoothly and the audio alone changing pitch and speed, I think I have a better idea of what may be happening: the Proffieboard may be under pressure, particularly around its audio circuitry. Pressure on that area can affect the sound, and a similar effect may occur if the board in your working saber is pressed.
One thing you could try is removing the battery, then gently warming the hot glue along the sides of the Proffieboard with a hair dryer or a heat gun on its lowest setting. Once the glue softens, carefully lift the board just a little so it is no longer being pressed down, then let everything cool and test the audio again."
I’m going to try that tomorrow, unless that’s a bad idea?
In my opinion, pulling things apart to see how they work then reassembling is a great learning experience. In fact, I usually do it regardless if something is broken or not.
I agree with the rest of the people, that is a weird and unusual thing, and it’s probably not a problem with the clock circuit. I don’t think I’ve heard that specific effect before. I would describe it as a warble.
Are the two boards in the video using the exact same configuration? (Because if so, we can rule out anything software-based causing this problem.)
There was one other person who had “warbling” going on with their board, but we never found out what caused it. (He just installed a new board.)
It could be a power problem. The CPU on the proffieboard has a kind of brown-out protection which triggers some weird slowdowns. To check, you would need to measure if the 3.3v pad actually has 3.3v on it. This is just guessing though.
Curious if the same effect is heard if you swap the SD cards but change nothing else. Since it’s the same config, the SD card should work in either one. I only say this because I’ve had a bad card give me audio issues and once I swapped it out everything went back to normal without changing the wiring. Just an alternative thought to add to the mix. I’m generally in favor of rewiring anyway if it comes to that.
Yeah, I’ve swapped the SD cards and get the same results out of the respective sabers. (3.9 still slow). So I think that rules out a bad SD. Very annoying.
I haven’t had time to heat up the board and unglue it according to CC Sabers, but will do that Thursday, and it would be cool to learn how the actual wiring of the proffieboard works, as I have no idea and would be a fun project. Very scared to mess it up though. I’m sure there are tutorials here though if I look
My own experience has been that slo mo issues can sometimes be caused either by power issues, as in maybe high resistance between one of of the battery connections to the board, or flux residue between critical components on the Proffie. Before doing anything expensive or complicated, I would definitely try sorting those things first. And do the cleaning several times to really be sure there is no trace of dirt, grease or flux anywhere.