Notifications
Clear all

ESP32 CAM MB sister board

5 Posts
3 Users
0 Likes
1,363 Views
BunnyKiller
(@bunnykiller)
Eminent Member
Joined: 3 years ago
Posts: 28
Topic starter  

The ESP32 Camera has a board that goes with it to ease the process of downloading programs etc, and it is known as "ESP32-CAM-MB".
It comes with a mini or is it the micro dont know... so many configurations to keep up with (pretty much the "standard sized small one") also has 2 push button switches to assist with the DLing.Well anyway, the board ( or the 2 I got from Amazon) seems have a design flaw or a part placement issue. The diode next to the main chip promptly made the pixies to become flatuant...   the nasty smelling gas was released and the board ceased to live. Dont know if the diode is a reverse polarity protection or a ripple reducing function. When the USB plug made its way into the socket, and about several seconds passed it toasted itself. I checked the USBs voltage and it was sitting at 5.02VDC with .2 VDC ripple ( I luv my sillyscope). So thinking it was maybe a one of bad diode, I put the second one into service and the pixies farted again. OK...  2 in a row, the only thing I can assume now is that the board has some bad circuitry going on. The USB port checks out fine, the Cameras are fine, and I cant think of anything else that would cause it to pop as it did...

So, If anyone else experiences this issue or knows what happened, Id luv to hear from yall, do a follow up here...

 

Thanx,,,   BunnyKiller


   
Quote
(@davee)
Noble Member
Joined: 2 years ago
Posts: 925
 

Hi @bunnykiller,

  Random thought ... might your anti-social "diode"s actually have been (surface mount) polarised (aluminium electrolytic or (possibly but unlikely) tantalum) capacitors. Such capacitors can obviously be soldered in two ways .. the right way and the wrong way .. and have been characterised in the past as 'functioning' long enough to pass a 10 second production line test when soldered either way ... but failing in seconds to minutes when soldered in the wrong way in 'normal' usage.

Looking at photos on the web, I assume the 'main chip' you refer to is type CH340C. Data sheet for the CH340x family at https://www.mpja.com/download/35227cpdata.pdf shows 10 microFarad decoupling capacitor, in parallel with 0.1 microFarad ceramic capacitor, across the power input pins, which are 1 and 16 for the C type.

You can probably check to see if your defunct components are connected this way .. determining which way they soldered might be more tricky if they have self-destructed. Warranty could be an option if you don't want to risk fixing yourself.

Good luck.


   
ReplyQuote
BunnyKiller
(@bunnykiller)
Eminent Member
Joined: 3 years ago
Posts: 28
Topic starter  

@davee what gave me the idea that it was a diode was the color, black, with a line printed on it as most diodes have and a code of SL. When the "item" failed it developed a little extrusion of bubbly plastic near the end opposite of the printed white line. Most of the time when SMT caps fail they crack and crumble ( from what Ive experienced so far) and the few failed resistors Ive tried to replace turn black and powdery and usually come apart when removed from the board. This mystery part held together when removed. Im considering "repairing" this board but not knowing exactly what type of diode it is could very well destroy the mosfet and main chip...  I was thinking of stuffing a very small standard diode to the weld pads and power it up and see what happens.. its not doing any good as it is now either, so who knows, maybe I will learn something from this pixie fart...


   
ReplyQuote
(@davee)
Noble Member
Joined: 2 years ago
Posts: 925
 

Hi @bunnykiller ,

  Sorry, I have no idea what the code SL means. Components from the 'big name' companies often have high resolution printing, so that they often (not always) print logo, capacitance, voltage, tolerance, etc all on tiny package ... but (speculating) these may be from a smaller outfit, or even a 'relabelling' outfit that buys out of spec parts in bulk and relabels with a simple machine. Initially, I had thought aluminium was more likely than tantalum, because it is usually cheaper, but supply chain games might reverse that cost preference.

I suspect the dead capacitors you found crumbled in the past were ceramics, which are not polarised. And neither are resistors, which usually get overheated and turn into burnt toast.

I agree, diodes can look like that part, but what job would a diode be needed for? Sometimes, a diode is put as protection for polarity reversal in the incoming power line, but this board is powered by USB which is unlikely to suffer this problem. Checking how this component is wired into the circuit is the only way.

  I have checked that my first thoughts were plausible. Polarised capacitors with that package style are available. I found a few aluminium electrolytics (although most are cylindrical tin can), but more commonly tantalum.

An example of a component with similar appearance to photo of your MB I saw on the web, except for the printing, from a Vishay data sheet https://www.mouser.co.uk/datasheet/2/427/tmcu-1763538.pdf

image

Wikipedia has some notes of interest .. https://en.wikipedia.org/wiki/Tantalum_capacitor

All tantalum capacitors are polarized components, with distinctly marked positive or negative terminals. When subjected to reversed polarity (even briefly), the capacitor depolarizes and the dielectric oxide layer breaks down, which can cause it to fail even when later operated with correct polarity. If the failure is a short circuit (the most common occurrence), and current is not limited to a safe value, catastrophic thermal runaway may occur. This failure can even result in the capacitor forcefully ejecting its burning core.

A particular cause of confusion is that on surface mount tantalum capacitors the positive terminal is marked with a bar. Whereas on aluminium surface mount capacitors it is the negative terminal that is so marked.

Whilst I am only looking at photos and considering the data sheet I referenced in my 1st reply, I can see why there might be 10 microFarad 6V capacitor but not a diode. I still think a capacitor is top of the suspect list.

I suggest you try to check the circuit of your board, comparing to the CH340x data sheet circuits. If "SL" is directly connected to the 5V and ground supplies the CH340C, then a decoupling capacitor seems much more likely than a diode.

If you want to try a capacitor, aluminium electrolytic should be fine. Make sure you get the polarity correct .. remove the remains of "SL", power up the board and measure 5V to check, before soldering in the new part. From the data sheet, I would suggest about 10 microFarad or more (say up to 50uF), 6V or more voltage rating.

Of course, this is all speculation ... I take no responsibility for any errors, etc.!!

Good luck.


   
ReplyQuote
jBo
 jBo
(@jbo)
Estimable Member
Joined: 2 years ago
Posts: 98
 
Posted by: @bunnykiller

The ESP32 Camera has a board that goes with it to ease the process of downloading programs etc, and it is known as "ESP32-CAM-MB".

@bunnykiller - I also have purchased a couple of ESP32-CAM modules with that separate small board. Mine have not farted, exploded or otherwise misbehaved. I've found them quite convenient, compared to last year when I had just a ESP32-CAM I had to hook up via an FTDI, and remember to ground GPIO0 each time. Oh, and as Bill mentions, it should be programmed at 3.3V but will probably run better at 5V, due to peak power required by WiFi. So testing with this little sister-board is quite convenient, and just needs USB.

The couple I got this year came with the following description and link:

ESP32-CAM WiFi Module ESP32 serial to WiFi ESP32 CAM Development Board 5V Bluetooth with OV2640 Camera Module DIY

I do find it a little bit of a gamble ordering from Aliexpress, but I like it well enough. I just don't always understand what it is I'm ordering.

Good luck!

In theory, theory and practice are the same.
In practice, they're different.


   
ReplyQuote