EEVblog Electronics Community Forum
General => General Technical Chat => Topic started by: paulca on April 01, 2021, 03:45:51 pm
-
Seriously, WTAF?
https://www.youtube.com/watch?v=MMrVvWCBFRQ&ab_channel=PaulCampbell (https://www.youtube.com/watch?v=MMrVvWCBFRQ&ab_channel=PaulCampbell)
That's my video. It's not just that one, they all seem to do it. It's not electrical signal interference on the camera either, cause I can listen to these things processing.
Should be obvious... there is no speaker. Just solid state components and passives.
-
I'm going with the Terry Prachett explanation. The video requirements for the ESP32Cam require the demons in the ESP32 to work extra hard, so they can't be quiet about it.
I wonder if they have names?
The one in the video is stuck in a reboot/crash loop, but when they are chooshing along doing actual work they make more interesting music.
-
I haven't noticed that with the one I have. Ghost in the machine!
-
Seriously, WTAF?
https://www.youtube.com/watch?v=MMrVvWCBFRQ&ab_channel=PaulCampbell (https://www.youtube.com/watch?v=MMrVvWCBFRQ&ab_channel=PaulCampbell)
That's my video. It's not just that one, they all seem to do it. It's not electrical signal interference on the camera either, cause I can listen to these things processing.
Should be obvious... there is no speaker. Just solid state components and passives.
Autofocus motor, maybe? It sounds to me just like the chatter of focus hunting.
-
Autofocus motor, maybe? It sounds to me just like the chatter of focus hunting.
It has no motor.
-
microphonic capacitors !
-
Also it's not clear if camera has moving parts, those can produce noise too when focusing, besides MLCC.
-
It has a fixed focus. No moving parts.
-
microphonic capacitors !
I know of at least one product which takes advantage of that effect to provide diagnostic beep codes, without an actual speaker. By running specific instructions in a loop modulating the CPU power supply current, you can make regular ceramic capacitors emit surprisingly loud sounds.
-
Autofocus motor, maybe? It sounds to me just like the chatter of focus hunting.
It has no motor.
Well poop, then I have no idea! If it doesn’t have focusing motors, then there’s also no way it’d have optical stabilization or an actual iris, which are the only other moving parts in non-zoom lenses.
-
Should have said, the only reason I videoed it was because I can hear it with my own ear, so it's not the camera I'm using making the noise either. It definitely processor noise, when it running and not crashing you can hear to doing things when you request a picture.
I expect the empty SDcard slot is amplifying whatever is making the noise.
-
What code is it running?
-
Should be obvious... there is no speaker. Just solid state components and passives.
I think the sound(s) are caused because it was recorded and/or published on April 1st.
Mysteriously, the sound(s) may have gone today, April 2nd ?
-
That noise is caused by focus hunting. You are recording the mechanical noise of the auto-focus actuator/lens, that keeps adjusting the focus continuously.
How a small cam moving lens looks inside. At minute 4:20 the focus moving lens actuator (coil) can be seen.
https://www.youtube.com/watch?v=KH0MZctnJlo (https://www.youtube.com/watch?v=KH0MZctnJlo)
-
What code is it running?
It's some RTSP cam software off github. Which had dodgy network code that didn't work for me, just put it into a reset loop.
I think the sound(s) are caused because it was recorded and/or published on April 1st.
Mysteriously, the sound(s) may have gone today, April 2nd ?
Nope. I swear, it's not an April fool.
-
That noise is caused by focus hunting.
You are recording the mechanical noise of the auto-focus gears/lens, that keeps adjusting the focus continuously.
It's a tiny like OVC2560 camera, it's fixed focus, has no moving parts.
-
Then, the noise might be produced by the recording camera you are filming that youtube video with.
That's focus hunting noise. It's a very specific noise, hard to find something else with a similar rattle noise.
While on macro mode, most filming devices enter in continuous auto-focus mode.
-
This reminds me of a noise problem I had recently. I was getting a lot of noise on the power rail of a circuit I had built. I eventually tracked it down to EMI from the controller for a LED tape I had stuck to the underside of a shelf for lighting. Every millisecond or so, the microcontroller would wake up and transmit a signal to any nearby switches. If I put my ear next to the controller, I could hear a few kHz whine.
If you have an oscilloscope, create a crude antenna with a few cm long piece of wire connecting a probe tip and ground. You might see the noise. That's how I tracked down the source of my problem.
-
If you have an oscilloscope, create a crude antenna with a few cm long piece of wire connecting a probe tip and ground. You might see the noise. That's how I tracked down the source of my problem.
Sound conduction works incredibly well when came to identify the place where from the noise originates, and what component exactly squeaks: Use a plastic or wooden rod/stick/pencil with one end touching your ear's external lobe/cartilage, and with the other end probing for noise by mechanically touching various parts.
-
Its not the camera. I removed it and it still makes processor noise.
https://www.youtube.com/watch?v=zDdb7NSLok4&ab_channel=PaulCampbell (https://www.youtube.com/watch?v=zDdb7NSLok4&ab_channel=PaulCampbell)
-
The second video has a different type of noise, IMO the second one sounds like a singing power supply.
You can use a skewer stick stick to your ear while with the other head of the stick touch various parts. First the coils, then the capacitors (especially the MLCC type likes to sing or act like any other part or sensor - MLCCs like to act like a microphone, a thermistor, a humidometer, speaker, etc.), ferrite beads, and so on.
-
The second video has a different type of noise, IMO the second one sounds like a singing power supply.
Different firmware. The LED flash indicates successful connection to the Wifi.
Maybe I should write some code with an ascending delay loop and see if I can get it to sweep.
-
The second video has a different type of noise, IMO the second one sounds like a singing power supply.
Different firmware. The LED flash indicates successful connection to the Wifi.
Maybe I should write some code with an ascending delay loop and see if I can get it to sweep.
Instead of sweep, go full SDR, like the dude that made a radio from the COM port of any PC. ;D
https://hackaday.com/2018/12/06/your-usb-serial-adapter-just-became-a-sdr/
https://hackaday.io/project/162477-serial-port-sdr
-
This is usually caused by harmonics in the buck converter. The noise is usually caused by the inductor, but can also be caused by ceramic caps exhibiting the piezoelectric effect.
-
This is usually caused by harmonics in the buck converter. The noise is usually caused by the inductor, but can also be caused by ceramic caps exhibiting the piezoelectric effect.
There aren't any inductors.
I have noticed if I drag my nail across the metal can the MCU is under the timbre of the sound is exactly that of the processor noise. I think this has to be some form of magnetic resonance with the metal can, yet it does it on both internal and external antenna models.
-
Who can I send this ESP32Cam who can test it properly? Anyone game?
-
I'm NOT especially sure, but those 3 bigger style, surface mount capacitors, between the SD card and the camera connector, are possible sound source culprits. 10uF's. (Value via online schematics).
Your video when they can be seen is way too blurry and moving too fast, so I will pass, at trying to identify them. I can't quickly find a parts list online.
(As other(s) in this thread, have effectively asked as well, in general) Does touching them with something appropriate change/stop/increase the sound/noise ?
Although they seem to be linear regulators, perhaps the rapid current fluctuations, as the MCU does things (and not), especially if I/O stuff is switched on and off, may cause rapid current fluctuations, I suppose. Hence sound, in some component types, such as those 3 prominent capacitors.
EDIT: (Some of my online sources of information)
Much more visible picture:
https://microcontrollerslab.com/wp-content/uploads/2021/02/AI-thinker-ESP32-components.jpg
Schematic:
https://i1.wp.com/randomnerdtutorials.com/wp-content/uploads/2020/03/ESP32-CAM-AI-Thinker-schematic-diagram.png?quality=100&strip=all&ssl=1
-
I plugged mine in again...doesn't make a sound.
-
I have recently received an ESP32 cam identical to yours. I'll let you know if I can hear anything.
That noise may be from whatever camera you film your Youtube videos with. Maybe its autofocus is hunting.
If you you might be able to get rid of it by using an external microphone on a cable. You'll be most likely to hear it if the mic is built into the camera
-
A loop works better than a monopole for this. Dave has a video on how to build a small RF probe loop. It really is a useful aid to finding all kinds of EMI.
This reminds me of a noise problem I had recently. I was getting a lot of noise on the power rail of a circuit I had built. I eventually tracked it down to EMI from the controller for a LED tape I had stuck to the underside of a shelf for lighting. Every millisecond or so, the microcontroller would wake up and transmit a signal to any nearby switches. If I put my ear next to the controller, I could hear a few kHz whine.
If you have an oscilloscope, create a crude antenna with a few cm long piece of wire connecting a probe tip and ground. You might see the noise. That's how I tracked down the source of my problem.
-
I plugged mine in again...doesn't make a sound.
I'm not 100% following this thread, so I could be talking nonsense.
But, surely you need to run the same software the OP has been using. Which presumably causes the audible current fluctuations, hence sounds via a capacitor or some other component.
I.e. (bold/font-size-change added/done by me) :
What code is it running?
It's some RTSP cam software off github. Which had dodgy network code that didn't work for me, just put it into a reset loop.
-
A loop works better than a monopole for this. Dave has a video on how to build a small RF probe loop. It really is a useful aid to finding all kinds of EMI.
This reminds me of a noise problem I had recently. I was getting a lot of noise on the power rail of a circuit I had built. I eventually tracked it down to EMI from the controller for a LED tape I had stuck to the underside of a shelf for lighting. Every millisecond or so, the microcontroller would wake up and transmit a signal to any nearby switches. If I put my ear next to the controller, I could hear a few kHz whine.
If you have an oscilloscope, create a crude antenna with a few cm long piece of wire connecting a probe tip and ground. You might see the noise. That's how I tracked down the source of my problem.
Yes - I connected a loop of wire between the probe tip and the probe's ground clip.
-
IT IS THE USB BOARD!
ESP32cam does not have USB port, so users normally adapt some USB to Serial boards. And using this there's no sound whatsoever. BUT recently i acquire two ESP-CAM-MB, who are some "official" USB addapter for ESPCAM, AND THOSE BOARDS MAKE THE NOISE!!!
I found this topic after hearing on one of my ESPCAMs. I tested the cameras using an extended flexcable camera (who allows me to put it in my ear!), and I could confirm that, somehow, the ESPCAM working together with this especifically board, make a low scratch sound every time the WIFI connection is used. I changed the ESPCAM to my other addapter and no sound! Put another ESPCAM on the ESP-CAM-MB and heres the sound!
I'm not sure what causes, because the board just have a couple capacitors, some resistors, one transistor (i think), two buttons and a CH340 chip.
-
At least one person accepts I'm not mad! They really do make noise.
It has to be magnetic coupling something such that it makes sound which relates to the software its executing. The kind of noise you get on dirty power rails with PCs on them. Either the CPU, memory buses or often the higher power stuff like the HDMI port will make high speed modem like noise. That is the same noise these ESPCam's make, but it's audible physically somehow.
I even flashed it with different software and it made a different set of beeps, squeaks and tappity tap sounds.
-
It has to be magnetic coupling something such that it makes sound which relates to the software its executing.
Given that the USB-serial board probably doesn’t contain any magnetics (like transformers or inductors), it’s very unlikely to be a magnetic effect. IMHO, it’s much more likely to be a ceramic capacitor.
-
I noticed that on my camera, specifically the programming shield under the camera, there was an audible noise. I separated the shield from the camera (as shown in the photo!) and clearly identified that the noise was coming from the shield—it was audible to my ear.
I suspect the ceramic capacitor on the shield is causing the noise. When the WiFi transmission starts, power spikes, which cause a variation in voltage across the capacitors and induce the audible noise. I removed the capacitors, and the noise stopped!
To clarify, the camera doesn’t have any moving parts, as it is fixed-focus, and even when the camera is completely removed, the noise continues, as demonstrated by the original author.