## Raspberry Pi 4-Based NAS Using USB-Connected Disks

Our little DNS-323 has been rock solid for the last decade but it’s getting long in the tooth and it’s just pokey enough to be annoying when I’m trying to do things on it. After a bit of consideration, I decided to see if I could build one with a Raspberry Pi.

It didn’t take a lot of research to find out that the Pi 1 through Pi 3+ aren’t particularly suited for NAS work. They may have the CPU horsepower, but with the on-board Ethernet and USB sharing the same USB2 port, their performance is reportedly not all that great.

The Pi 4, on the other hand, has entirely different hardware; the USB ports (of which there are two USB3 and two USB2) have their own controller, while the on-board Ethernet has its own controller. This makes for an ENORMOUS improvement in the performance of both USB:

and Ethernet:

With numbers like that, I thought it would be worthwhile to try building a little RAID1/mirrored home NAS around a Pi 4. Here’s what I used:

• Raspberry Pi 4, 2GB model, qty 1
• Official Raspberry Pi 4 power supply, qty 1
• 32GB Sandisk UHS-1 MicroSD card, qty 1
• 4TB Western Digital Blue 3.5″ hard drive, qty 2
• Vantec NexStar TX 3.5″ external USB3 enclosure, qty 2
• 5ft Category 5e patch cord, qty 1

I went with mirroring two disks (RAID1), so that is what I’m going to go through here. If you want to set up a single disk, or set up something like a four-disk mirror, or RAID5/6/1+0, you can use the same software I did but you’ll have to do a bit of research into the settings.

Oh, and if you are going to create a RAIDset with more than one disk, make sure they’re all the same size, otherwise the mirror will only be as large as the smallest of the disks that are part of the RAIDset!

I had originally planned to use openmediavault to mirror the disks and create the network shares, but unfortunately it doesn’t support USB-connected disks. That’s what I get for not reading enough before I buy stuff, I suppose. With omv out of the picture, I decided to try mirroring the disks and set up the network shares myself. Here’s how I did it:

Part 1: Set Up Your Raspberry Pi

I’m not going to get into this because there are already a ton of sites out there that will show you how to do this (and describe it better than I can). I built my NAS with the Raspbian Buster Lite image, dated 2020-02-13. Do not use wireless (don’t bother with a wpa-supplicant.conf file), but make sure you enable ssh, go through the raspi-config menu and don’t forget to apt update and upgrade!

Part 2: Assemble And Format The Disks

This part’s pretty easy. Install the hard drives in the enclosures, connect them to your PC, and using the software of your choice, remove any existing partitions on the disks and create a single NTFS partition (or ext4 if your PC is a Linux machine) that uses the entire capacity of the disk. Once that’s done and you get no errors, safely remove the enclosures from your PC.

Part 3: Put Together And Connect Your Hardware

This part’s pretty easy too. Connect the USB disks to the Pi and turn them on. If things are working properly, the physical enclosures and the disks will be present. To check if the Pi sees the enclosures, type lsusb:

To see if the actual spinning hard drives have been detected, type dmesg | grep sda for the first enumerated disk, and dmesg | grep sdb for the second:

Do not go any farther if the output of either of those commands doesn’t look correct, or if the disk capacity listed is different than you expect. Go back, check all of the parts and connections, and try again.

Part 4: Set A Static IP On eth0

You should at this point already have a cable connecting your Pi to your router or a switch. If you’re using wireless… well, I suppose you can do that if you really want or need to, but you’re going to be making your Pi do its work with one foot in a bucket. Use that RJ-45 jack and get yourself some nice clean Gigabit Ethernet goodness.

To set a static IP, use your favourite editor to edit the /etc/dhcpcd.conf file. Make sure to use sudo so you’re editing it as the superuser. Go right to the bottom of the file, and add the following lines:

# NAS Static IP for eth0
interface eth0
static routers=Z.Z.Z.Z
static domain_name_servers=A.A.A.A B.B.B.B

Where:

• X.X.X.X is the static IP address on your network that you want your NAS to be reachable at
• YY is the CIDR representation of your subnet mask (most home or small businesses will be /24)
• B.B.B.B is the IP address for your secondary DNS server (if you have one)

Make sure that you type everything exactly. Even if you only have one router and one DNS server, you still need to type static routers and static domain_name_servers with the “s”.

Once you’ve finished setting the dhcpcd.conf file, reboot your Pi. Once it comes back up, see if you can ping other devices both inside and outside your network. Then try to do another apt update and upgrade to see if your Pi can talk to the Raspbian repositories. If not, go back over the file and make sure the changes you made were saved. Also check to make sure everything is spelled correctly (remember it’s case sensitive).

Part 5: Mirror The Disks

Congratulations! If you’re here, that means you have successfully set up your Raspberry Pi, can see it on the network, and have two hard disks connected via USB. You are now ready to do something not a lot of other people do – use a Raspberry Pi to make a RAIDset out of a pair of USB-connected disks. Fortunately, the software (like the hardware) has made leaps and bounds since the last time I tried it and it’s pretty easy to set up.

First, let’s make sure the right stuff is installed:

sudo apt update

Since you already made sure the disks were working in Step 2, you can go ahead and create a RAID1 mirror. In my case, I didn’t care about the partition size so I used the entirety of both disks with the following command:

sudo mdadm --create --verbose /dev/md0 --level=mirror --raid-devices=2 /dev/sda /dev/sdb

In this case, mdadm is being told:

-- create : Make a new RAIDset
-- verbose : Show what’s going on while the command is running
/dev/md0 : The name of the RAID device you’re creating
--level=mirror : Create a mirror (RAID1)
-- raid-devices=2 : How many disks will be used
/dev/sda /dev/sdb : The names of the disks that will be used

Again, this is how I set up my own little two-disk mirror. If you have a different number of disks or want to set up a different kind of RAIDset, the syntax is pretty much the same but the options are different. You may also want to use particular partitions instead of entire disks like I did. Check out the mdadm man page.

You should now see the lights on the enclosures blinking furiously and/or be able to feel/hear the the hard drives doing something. Depending on the kind and size of disks you have and the type of enclosure, creating the mirror and syncing it up may take up to a day. You can check the progress with the following command:

cat /proc/mdstat

which should give you output something like this, which shows you the status of the mirror sync:

Personalities : [raid1]
md0 : active raid1 sda[3] sdb[2]
3096885440 blocks super 1.2 512K chunks 1 near-copies [2/2] [UU]
[>....................] resync = 0.2% (61504/3096885440) finish=398min speed=14298K/sec
unused devices: <none>

Once it’s done, instead of the above, you should see something similar to the following when you run the same command:

pi@PI-0:~ $cat /proc/mdstat Personalities : [raid1] md0 : active raid1 sda[1] sdb[2] 3906885440 blocks super 1.2 [2/2] [UU] bitmap: 0/30 pages [0KB], 65536KB chunk Note the line with the [UU] at the end. Each U represents an active and healthy RAID disk. If you run cat /proc/mdstat and you see a _ (underscore) instead of a U, there’s a problem with a disk that requires your immediate attention. Now that the RAIDset is built, you need to save its configuration so your Pi knows what to do with it when it boots: sudo -i mdadm --detail --scan >> /etc/mdadm/mdadm.conf exit Now confirm it was saved: cat /etc/mdadm/mdadm.conf and look for the line that says something like this (obviously your UUID will be different): ARRAY /dev/md/0 metadata=1.2 UUID=061a78a9:ceadf64b:b124c1d4:7e35ae85 name=PI-0:0 Now reboot your Pi, and once it comes back up, use cat/proc/mdstat and blkid to see if everything’s okay: Part 6: Create A Filesystem Now that the disks are mirrored, it’s time to put a filesystem on them. I use ext4, and I’m creating a filesystem on the RAIDset, NOT the physical disks, so the command is: sudo mke2fs -t ext4 /dev/md0 This may also take a little while. Once it’s done, edit the /etc/fstab file so that the filesystem on the RAIDset will automount at boot. I use the /mnt directory as the mount point, here’s what my fstab file looks like: Notice that the UUID of the partition is the same as the UUID for /dev/md0 in the output of blkid. Type the following to see if your fstab file is set up right: sudo mount -a df You should see something like this: In this case, /dev/md0 is mounted at /mnt and everything looks good. The use is at 12% because I’ve already been using the NAS – yours will probably say 0% or 1%. Now, shut your Pi down, then turn off the disks. Turn on the disks, wait for them to spin up, and boot up the Pi. Run those two commands again and make sure everything looks good before going any further. Part 7: Set Up A File Share With Samba Samba is a mature, stable, and very useful batch of software that makes it pretty easy to create simple network shares. It may already be installed, but just to be on the safe side: sudo apt install samba sudo apt install samba-common-bin Once it’s installed, you’re going to need to configure it by editing the /etc/samba/smb.conf file. I’m only going to set up one network share for now, and if you’re new to Samba, I suggest sticking to one share until you’re familiar with it. Before we edit the file, though, we need to create the directory that Samba will use to share over the network: sudo mkdir /mnt/NAS_FILE sudo chown pi /mnt/NAS_FILE sudo chgrp users /mnt/NAS_FILE Do NOT use /mnt as the directory for your file share – always use a directory that resides on the device you’re mounting. If for some reason /dev/md0 doesn’t mount properly, you may end up writing data to and filling up the SD card instead of using the disks! The default smb.conf file contains a number of examples, including one for a network share and one for a print share. Copy the default file: sudo cp /etc/samba/smb.conf /etc/samba/smb.conf_OLD Now, with sudo, use your favourite text editor to open /etc/samba/smb.conf and go to the bottom section of the file, labelled “Share Definitions”. Delete everything in that section and replace it with the following: [NAS] comment = NAS Fileshare path = /mnt/NAS_FILE browseable = yes read only = no writable = yes create mask = 0775 directory mask = 0775 valid users = pi So the “Share definitions” section in my smb.conf file looks like this: Notice how the “valid users” section has the name “pi” in it – you can change that to anyone you’d like (or have more than one user on that line), but for each user on that line, you’ll need to create an account on the Pi for them. I just stuck to the pi user because I wanted to keep things simple to start. To test your smb.conf file, run the following: testparm The “Loaded services file OK” is a good sign that your smb.conf file has no obvious errors in it. If you don’t get that message, go back through the file and make sure everything is spelled properly, etc. Now, for each user account you want to grant access, you need to run the smbpasswd utility to set them up in Samba. It will ask for a password, and that password really, really should really be different than the password that’s used by the user to log into the Pi itself! To add someone to the Samba system: sudo smbpasswd -a USERNAME To disable someone’s Samba account: sudo smbpasswd -d USERNAME To re-enable someone’s Samba account: sudo smbpasswd -e USERNAME In my case, I ran: sudo smbpasswd -a pi and gave it a password that was different from the password that I use for logging in with the pi user. At this point, Samba is installed, you’ve created a directory on the RAIDset that Samba will use for the file share, you’ve edited the smb.conf file, ran smbpasswd for every user that’s listed in the smb.conf file, and tested your configuration with testparm. It’s time to restart Samba so it loads the new configuration: sudo service smbd restart If you don’t get any messages or errors, things may actually be working! Part 8: Access Your New NAS Finally, the payoff – your own home-built NAS! How you will access it depends on the operating system on the computer you want to access it with. If you open your file or network browser, it may automatically show up. Otherwise, you will have to browse to it. Open your file or network browser and browse to the static IP address you set way back in Part 4. In Windows, it should look something like this: In Ubuntu, you may have to enter smb:// before the address: When you try to open the NAS share, you should be prompted for a username and password. The username will be: localhost\USERNAME, which in my case was localhost\pi …and the password will be whatever you set it to with smbpasswd. Some of my friends and family may disagree with this statement, but I like it when things are organized. To help with this, I designed and printed a lightweight frame that holds two disks and a Pi, and has several holes for 5mm M3 screws to fasten things like cable management or velcro or whatever to it. Here’s how mine turned out: I put the model on Thingiverse, it’s at https://www.thingiverse.com/thing:4235287 As for performance… this setup is much more responsive and can transfer files to and from the disks much faster than the old NAS. Disk fragmentation can slow things down, and the old NAS is a decade old, but it was only about half full so that shouldn’t have been too big a problem. Transfer rates to and from the Pi are faster than my home wireless is, and browsing the directories on the file share is no different than browsing the directories on my PC. Nice and quick. 150Mbps isn’t too shabby at all! The best I could get out of the old NAS was around 70Mbps. Bottom line – a Pi 4 with two external USB3-connected hard drives makes a serviceable and reasonably fast NAS for home or small business use, although there are security considerations that need to be addressed prior to using it out in the real world. So that’s the deal. I did a bunch of torture testing when I first set things up, and things recovered gracefully. I will do another post soon to discuss how to fix a mirror if there’s a disk failure or if you need to recover the array entirely due to a Pi failure. I hope you found this useful! ## Temperature And Humidity Display Using Arduino, DHT22, And MAX7219 Display I finished another project today. This time it’s a simple temperature and humidity display, and so far it’s working pretty well. It’s built around an Arduino Nano and uses a DHT22 sensor. The display is an extremely cheap MAX7219-based four-module LED matrix (130x32mm), and its brightness is controlled by a capacitive touch sensor (11x15mm). In this case, everything runs off a 5V supply so there are no level shifters needed and a single USB cable can power the whole thing. Any other 5V Arduino with hardware SPI will work fine here, too. The software libraries I used also support software SPI but I haven’t tried that out. Here’s the layout: • Everything is connected to the +5V and GND pins on the USB connector. • MAX7219 CLK to Arduino 13. • MAX7219 DATA to Arduino 11. • MAX7219 CS to Arduino 10. • DHT22 I/O to Arduino 2. • Touch sensor I/O to Arduino 4. I had everything on a breadboard but forgot to take a picture before wiring everything up to fit in the case… here it is wired up and just before being prepared to put into the case. The program uses the MD_MAX72XX and MD_PAROLA libraries for the display, and the SimpleDHT library for the DHT22. It took me a while to wrap my head around the MD_PAROLA stuff, but the examples included with the libraries were very helpful. Here’s the program: /* Temp and RH DHT22 MAX7219 for Dot 04 * Uses Nano to check DHT22 and display on 8x8 dot matrix (x4) MAX7219. * Meant to be used indoors. * Has two brightness settings, 4 and 15 (on scale of 0-15) * Runs off 5V USB. * MAX7219 controlled by MD_Parola and MD_MAX72xx libraries * DHT22 using SimpleDHT * PINS: * DHT22 data: D2 * MAX7219 clock: D13, data: D11, CS: D10 * Intensity: D4 * Uses a MAX7219 32x8 LED module from Banggood. Hardware type is MD_MAX72XX::ICSTATION_HW, 4 devices * Puts temp and RH on display at same time */ /* * MAKE SURE YOU RUN THE MD_MAX72XX_HW_Mapper to confirm the hardware setting for your particular display! * The results I got for the display I have were: * * HW_DIG_ROWS 1 * HW_REV_COLS 1 * HW_REV_ROWS 1 * Your hardware matches the setting for IC Station modules. Please set ICSTATION_HW. * */ #include <MD_Parola.h> #include <MD_MAX72xx.h> #include <SimpleDHT.h> #include <SPI.h> // Set up DHT22 vars for data TX/RX #define h_w 8 #define h_h 8 static unsigned char h_w_bits[] = { 0x3c, 0x42, 0xa5, 0x81, 0xa5, 0x99, 0x42, 0x3c }; #define s_w 8 #define s_h 8 static unsigned char s_w_bits[] = { 0x3c, 0x42, 0xa5, 0x81, 0x99, 0xa5, 0x42, 0x3c }; // Create instance for the DHT22 using pin 2 for data xfer SimpleDHT22 dht22(2); // Define the number of devices we have in the chain and the hardware interface // NOTE: These pin numbers will probably not work with your hardware and may // need to be adapted #define HARDWARE_TYPE MD_MAX72XX::ICSTATION_HW // Found using the MD HW mapping program #define MAX_DEVICES 4 // Four 8x8 modules on this particular board #define CLK_PIN 13 #define DATA_PIN 11 #define CS_PIN 10 #define BRIGHT_PIN 4 // Hardware SPI connection MD_Parola P = MD_Parola(HARDWARE_TYPE, CS_PIN, MAX_DEVICES); byte CountUp = 0; void setup() { delay(500); // Need this because display doesn't seem to start up right away. P.begin(2); // Using 2 zones, one for temp, one for humidity pinMode(BRIGHT_PIN, INPUT); P.setZone(0,0,1); P.setZone(1,2,3); P.displayZoneText(0, "Hi!", PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.displayZoneText(1, "Hi!", PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.setZoneEffect(0, 1, PA_FLIP_UD); // Need this because I glued the display in upside down >:-( P.setZoneEffect(1, 1, PA_FLIP_UD); // Need this because I glued the display in upside down >:-( P.setZoneEffect(0, 1, PA_FLIP_LR); // Need this because I glued the display in upside down >:-( P.setZoneEffect(1, 1, PA_FLIP_LR); // Need this because I glued the display in upside down >:-( P.displayAnimate(); delay(2000); } void loop() { jumpback: // If touch sensor is active, cycle through the 16 levels of brightness until sensor is inactive. int brightness_change = digitalRead(4); while (brightness_change == 1){ if (CountUp == 16){ CountUp = 0; } P.setIntensity(0, CountUp); P.setIntensity(1, CountUp); delay(250); CountUp = CountUp + 1; brightness_change = digitalRead(4); } float temperature = 0; float humidity = 0; int err = SimpleDHTErrSuccess; if ((err = dht22.read2(&temperature, &humidity, NULL)) != SimpleDHTErrSuccess) { // If we're here, there was a problem reading the DHT22. Show an error then try again. P.displayZoneText(0, "Dht", PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.displayZoneText(1, "Err", PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.displayAnimate(); delay(5000); goto jumpback; // I know, I know. Don't say it... } // Convert the float to a string to display char temp_result[6]; dtostrf(temperature,2,0,temp_result); // Convert the float to a string to display char hum_result[6]; dtostrf(humidity,2,0,hum_result); P.displayZoneText(1, hum_result, PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.displayZoneText(0, temp_result, PA_CENTER, 75, 0, PA_PRINT, PA_NO_EFFECT); P.displayAnimate(); delay(3500); // DHT22 max sample rate is about 2 seconds. } If you are using a MAX7219-based display, save yourself some time and frustration by connecting it and running the MD_MAX72XX_HW_Mapper program that comes with the MD_MAX72XX library before you do anything else. It will tell you how your display is set up, regardless of how it actually looks. After doing some testing, I found that the capacitive touch sensor I was using could reliably detect my finger out to about 5mm away. That was great because then I could hide it inside the case and there’d be no switch, no pad… just a “magic” spot on the back that changes the brightness if you put your finger there. I designed a case for this particular project, including the specific display and touch sensor I had on hand. It’s vented, has a hole for a USB cable, and is closed up with four 6mm M3 screws: With everything wired up and tested, I hot-glued everything… and I mean everything. Every connector, every module (except the Nano’s mini-USB port – never know if I’ll want or need to reprogram it) … it’s all quite secure inside the case. I glued put a piece of plastic on the back of the display just in case any other parts work their way loose and came in contact with it. I’m still a little wary of doing things this way, but it sure beats drawing up and etching boards for this kind of stuff! Once the glue had cooled and I confirmed everything was stuck good and tight, I closed up the case and plugged the cable into a 5V USB power supply. The LEDs flashed, and then… everything was upside down. I’d glued the display in upside down. So… another 45 minutes or so of pondering and looking and I found how to flip the display in software so it looked right again. If you run into this problem, check out setZoneEffect() in the MD_PAROLA documentation. Here it is, from the back: And from the front, display pointing the right way: The STL files for the case are available at https://www.thingiverse.com/thing:4202464 ## Find A File That Contains A Particular Text String (Windows) Windows File Explorer is many things. One thing it is not, however, is good at searching. It used to be… but somewhere along the way something got broken. Fortunately, there’s a command-line tool that works much better than the GUI, called findstr. You can look up the options by typing findstr /? If you’re looking for a particular text string and you only want to know which files contain that string, use the following command: findstr /S /I /M /C:TEXT_STRING_HERE * The /S means search this directory and all subdirectories. The /I means don’t be case-sensitive. The /M means print only the filename if there’s a match. The /C: specifies the text string you’re looking for. The * means check everything. ## Fix For Arduino IDE Compiling Very Slowly In Windows My Raspberry Pi has been compiling Arduino programs in about a tenth the time that my main Windows machine has. The Windows machine is not a slouch – i7-7700HQ with 16GB RAM and a SSD, and it was getting annoying waiting for several minutes while some of the larger ESP programs compiled. Turns out the antivirus was running full-tilt, scanning every file as it was loaded into the compiler, pinning the CPU. The fix for me was to set an exception in my AV software for the C:\Users\[yourname]\AppData\Local\Arduino15 folder. It made a huge difference in speed and my laptop no longer sounds like a jet engine when it’s compiling a 20-line program for a microcontroller with 2.5kB of RAM… ## ESP32-CAM Low Power Trail Camera I’ve been spending a lot of time lately working with the ESP32-CAM module. It doesn’t produce the best pictures I’ve seen, but for the cost (I’ve found them for$9, including the OV2640 camera!) and the number of features and horsepower, they’re tough to beat.

One of the things I want to do with them is put a couple outside and get pictures of the different kinds of animals that wander through the yard and leave footprints in the snow. When I mentioned this to a buddy of mine, he immediately wanted to know if they could be used to watch for motion and take pictures in case someone was trying to break into his shed or cabin. Sure, I told him – I didn’t see any reason why not. His response was to immediately ask me how many I could make for him and how quickly I could do it.

Unfortunately, it was just an idea at the time and I hadn’t actually tried to do it. I figured it wouldn’t be too tough – after all, the ESP32-CAM AI-Thinker modules I use have several GPIO pins broken out. I was wrong.

Turns out some of the GPIO pins are used by the camera, and the rest are used by the uSD card slot that’s on the board. One of them (D4) seems to be used by BOTH the camera (camera flash) and the uSD card slot (data line).

I tried a bunch of things and didn’t have much luck, and when I looked around for information, there were lots of links but I couldn’t find any information that quite fit what I was doing.

Finally, I found a link to some ESP documentation, which got me started. Looking into the various ESP libraries for the SD card, using a FAT32 filesystem, the camera, and the on-board EEPROM took a while but after I figured one or two of them out, the others were easier.

After going through my various parts bins, I cobbled together a circuit that seems to reliably work. Here’s the schematic of the whole thing:

+V on the schematic is the power supply you want to use. Power for the ESP32 first goes to an AMS1117-3.3 regulator. According to the AMS1117 datasheet, it will run to where the input is only 1V higher than the output. The output is 3.3V, so it should be able to run down to 4.3V. The absolute maximum input voltage is 15V, so powering it from 4xAA/4xC/4xD alkaline batteries (6V) is fine. Even 9 or 12V should be OK, but check the regulator on your board first to make sure.

Remember that if you want to power it from rechargeable NiCd or NiMH batteries, those are 1.2V, not 1.5V, so you’d probably want to use five of them, instead of four alkalines. Same with those long-life lithium batteries – they’re 1.2V too.

The block labelled “OPTIONAL” is there if you want a switch that will keep the MOSFET (and ESP) turned on while programming. You can also just move the ESP GND pin from the MOSFET drain to ground while programming. Or… if your PIR will remain on while motion is present, you can just wave your hand above the sensor until programming is done. That’s what I do.

R1 and R2 are necessary, particularly if you have a MOSFET with a high input capacitance. They keep the MOSFET gate from momentarily pulling a large amount of current which could damage the PIR or ESP.

Why the 4N37? Because the ESP is not connected to the ground rail when the MOSFET is turned off, so GPIO13 does a very noisy “high-ish” float which leads to unpredictable results. Note that the 4N37 diode side circuit goes from GPIO13 to the anode, the cathode goes to R3, and R3 goes back to the GND pin on the ESP – NOT the ground rail from the power supply!

This circuit works best with low-Vf Schottky diodes that can tolerate a reverse voltage at least 2x the highest possible voltage in your circuit. Tested and works with BAT41 and 11DQ06. Tested and works with old-school 1N34 germanium diodes (but I probably wouldn’t use them for real-world applications). It seems to mostly work but not as reliably with typical 1N914 and 1N4001 diodes. It does NOT work with anything with a Vf larger than about a volt, like LEDs.

This circuit also works best with MOSFETs that are fully on at a low voltage, like 2.5 to 4.5V, and have a very low drain-source resistance when on (tens to a couple hundred milliohms). Tested and works with IRLI640G and DMN1019USN-7.

Normally I’d put bypass capacitors across the 5V and GND pins of the ESP, but here’s the ESP32-CAM power circuit:

Note the abundance of capacitors, which is pretty great. A 0.1uF capacitor probably wouldn’t hurt, but I don’t think it’s necessary unless you’re using a very noisy power supply.

So that’s the schematic. Breadboarded up, this is how it looks:

Here’s how it works:

• When power is applied to the circuit, the gate of the MOSFET is low and doesn’t conduct, so the ESP is disconnected from the ground rail. The circuit pulls about 16uA at this time.
• When the PIR sensor detects motion, its trigger pin goes high for two seconds. This signal is sent through a diode and 47k resistor to the gate of the MOSFET, which turns it on.
• With the MOSFET turned on, the ESP now has power and boots. As soon as it can, it sets GPIO13 high. This turns on the input of a 4N37 optoisolator, which turns on its output transistor. The output transistor is also connected to the gate of the MOSFET through a diode and a 47k resistor. This keeps the MOSFET turned on even after the PIR trigger line goes low.
• The ESP goes through the paces of checking for and mounting the uSD card, starting up the camera, and checking the EEPROM and reading the number from it that indicates the last picture that was taken (PIC_COUNT) prior to the previous shutdown.
• If there is a problem at any point in the startup, the ESP will set GPIO13 low, which will turn it off and wait for the next trigger to boot again. I know it’s 2020, but sometimes a reboot still fixes things.
• If there are no problems, it gets to the main loop, which takes the number of pictures specified by a while loop that increments the variable COUNTUP (in the program here it takes five pictures). Each time through, the picture counter (PIC_COUNT) is increased by 1. The circuit pulls about 130-140mA at this time.
• Once COUNTUP reaches the maximum set in the while loop, the ESP saves the current value of PIC_COUNT to the EEPROM and then sets GPIO13 low. This should turn off the MOSFET and remove power from the ESP.
• If there is still power, the ESP waits for 500ms after it tried to shut itself down. If it’s still awake, then that means the PIR has either re-triggered or is still triggered so it’s a good idea to get more pictures. The ESP sets GPIO13 high again and loops back to take another five pictures.

Here’s what it looks like when it’s running. Note the camera flash LED on the board glowing faintly instead of blazing like a million suns like it usually does. When the LED is on, the ESP is communicating with (and hopefully writing an image to) the uSD card.

You may be wondering why it works, though. After all, there don’t seem to be any usable free GPIOs when using both the camera and the uSD card, so what’s with GPIO13? Well, it comes down to changing this line:

SD_MMC.begin("/sdcard")

to this:

SD_MMC.begin("/sdcard",true)

Selecting “true” tells the ESP to talk to the uSD card in 1-bit mode instead of the usual 4-bit mode. This frees up a couple of GPIO pins, one of which is GPIO13. The disadvantage to using 1-bit mode is that it’s slower, but I’m pretty sure the ESP itself is going to be the bottleneck here. Plus, the OV2640 and lens aren’t super-high quality so setting the JPEG image quality high and making huge files isn’t necessary (or useful).

Here’s the program in its entirety (it looks kind of mangled but if you copy and paste it into a text document or the Arduino IDE it comes out properly):

/* ESP32-CAM Low Power Trail Camera v1
* Mark's Bench (http://marksbench.com)
* Uses ESP32-Cam AI-Thinker with OV2640 camera to take pictures and save to SD card when triggered by PIR
* ESP32 is connected to power rail via MOSFET. MOSFET is initially turned on by PIR trigger and kept on by setting pin D13
* on the ESP32 high as soon as ESP starts up.
*
* ESP then takes 5 pictures and saves them to the SD card, after which it sets D13 low, which turns the MOSFET off,
* cutting the ESP off from the GND rail.
*
* If the PIR trigger remains high or goes high again during when the ESP32 would shut down, then take another five
* pictures and try shutting down again.
*
* Advantage to this scheme is a power savings - power draw is less than 20uA when the MOSFET
* is off and the ESP32 is shut down. Power draw is around 130-140mA when pictures are being taken and saved.
*
* Disadvantage is that when using both the camera and the SD card, there are no easily usable GPIO pins available.
* To get around this, use 1-bit SD card access instead of the usual 4-bit.
* It slows things down but you can still get an image with ok quality about once a second
* at full resolution (1600x1200) on the OV2640.
*
*
* Information on and code examples for using the ESP32-CAM library:
* https://github.com/yoursunny/esp32cam
* https://github.com/espressif/esp32-camera
* https://github.com/espressif/arduino-esp32/tree/master/libraries/ESP32/examples/Camera/CameraWebServer
*
* Information on and code examples for using the ESP32 SD_MMC library:
* https://github.com/espressif/arduino-esp32/tree/master/libraries/SD_MMC
*
* Information on using the ESP32 EEPROM (the Preferences library):
* https://github.com/espressif/arduino-esp32/tree/master/libraries/Preferences
*
* VERY useful ESP32 documentation:
* In particular, the "ESP32-WROOM-32 Datasheet", "ESP32 Datasheet", and "ESP32 Hardware Design Guidelines".
* Also, the schematic at:
* https://github.com/SeeedDocument/forum_doc/raw/master/reg/ESP32_CAM_V1.6.pdf
* And the specification page (which has some errors) at:
* https://github.com/raphaelbs/esp32-cam-ai-thinker/blob/master/assets/ESP32-CAM_Product_Specification.pdf
*
*
* ***TO PROGRAM: Set Board to "AI Thinker ESP32-CAM"***
*/

#include <esp_camera.h>
#include <FS.h>
#include <SPI.h>
#include <SD_MMC.h>
#include <Preferences.h>

// The ESP32 EEPROM library is deprecated. Use the Preferences library instead.
Preferences preferences;

// The following defines are for the ESP32-Cam AI-THINKER module only. I haven't tried any others.
#define CAM_PIN_PWDN    32
#define CAM_PIN_RESET   -1
#define CAM_PIN_XCLK    0
#define CAM_PIN_SIOD    26
#define CAM_PIN_SIOC    27
#define CAM_PIN_D7      35
#define CAM_PIN_D6      34
#define CAM_PIN_D5      39
#define CAM_PIN_D4      36
#define CAM_PIN_D3      21
#define CAM_PIN_D2      19
#define CAM_PIN_D1      18
#define CAM_PIN_D0       5
#define CAM_PIN_VSYNC   25
#define CAM_PIN_HREF    23
#define CAM_PIN_PCLK    22

// Create a variable to hold the picture number. Since the SD card is formatted FAT32, the maximum number of files
// there can be is 65534, so a 16-bit unsigned number will be fine.
uint16_t PIC_COUNT = 0;

void setup(){
pinMode(13, OUTPUT);    // GPIO13 available when using SD_MMC.begin("/sdcard",true) for 1-bit mode (set below)
digitalWrite(13, HIGH); // Hold the gate of the MOSFET high as soon as possible after boot to keep the power on
// after the PIR is done triggering.

//Serial.begin(115200); // Uncomment for troubleshooting

preferences.begin("trailcam", false); // Open nonvolatile storage (EEPROM) on the ESP in RW mode
PIC_COUNT = preferences.getUShort("PIC_COUNT", 0);  // Get the stored picture count from the EEPROM.
// Return 0 if it doesn't exist.
// getUShort() fetches a 16-bit unsigned value

// Now, configure the camera with the pins defined above and recommended settings for xclk, led_c, and format.
camera_config_t config;
config.pin_d0 = CAM_PIN_D0;
config.pin_d1 = CAM_PIN_D1;
config.pin_d2 = CAM_PIN_D2;
config.pin_d3 = CAM_PIN_D3;
config.pin_d4 = CAM_PIN_D4;
config.pin_d5 = CAM_PIN_D5;
config.pin_d6 = CAM_PIN_D6;
config.pin_d7 = CAM_PIN_D7;
config.pin_xclk = CAM_PIN_XCLK;
config.pin_pclk = CAM_PIN_PCLK;
config.pin_vsync = CAM_PIN_VSYNC;
config.pin_href = CAM_PIN_HREF;
config.pin_sscb_sda = CAM_PIN_SIOD;
config.pin_sscb_scl = CAM_PIN_SIOC;
config.pin_pwdn = CAM_PIN_PWDN;
config.pin_reset = CAM_PIN_RESET;
config.xclk_freq_hz = 20000000;
config.ledc_timer = LEDC_TIMER_0;
config.ledc_channel = LEDC_CHANNEL_0;
config.pixel_format = PIXFORMAT_JPEG;

// Make sure there is PSRAM available (the AI-Thinker module has PSRAM). Otherwise, don't go any further.
if(psramFound()){
config.frame_size = FRAMESIZE_SXGA; // If there's PSRAM then there's enough memory to capture up to 1600x1200
// The following resolutions are available:
// 96x96 (96x96)
// QQVGA (160x120)
// QQVGA2 (128x160)
// QCIF (176x144)
// HQVGA (240x176)
// 240x240 (240x240)
// QVGA (320x240)
// CIF (400x296)
// VGA (640x480)
// SVGA (800x600)
// XGA (1024x768)
// SXGA (1280x1024)
// UXGA (1600x1200) **Full-resolution for OV2640

config.jpeg_quality = 10; // Valid: 0-63, with 0 being highest quality and largest file size.
// Anything lower than 8 creates large file sizes that take a long time
// to save to the SD card.
// The camera and lens aren't the best quality, so huge files
// won't get you a better picture beyond a certain point.
config.fb_count = 2;  // With the PSRAM, there's enough memory for two framebuffers, which speeds captures.
}
else{
while(true){
// The AI-Thinker module has PSRAM. I haven't tried any module without PSRAM.
//Serial.println("NO PSRAM FOUND"); // Uncomment for troubleshooting
delay(500);
}
}

// Start up the camera with the configuration settings made earlier in the "config." statements.
esp_err_t err = esp_camera_init(&config);
if (err != ESP_OK){
// If we're here, there's a problem communicating with the camera.
// Turn the ESP off and wait for the next trigger.
digitalWrite(13, LOW);
//Serial.println("CAM FAIL"); // Uncomment for troubleshooting
while (true){
// Need this loop to wait in case the PIR is keeping the power on.
}
}

// Start up the SD card, using 1-bit xfers instead of 4-bit (set the "true" option). Frees up GPIO13.
if(!SD_MMC.begin("/sdcard",true)){
// If we're here, there's a problem with the SD card.
// Turn the ESP off and wait for the next trigger.
digitalWrite(13, LOW);
//Serial.println("SD FAIL 1");  // Uncomment for troubleshooting
while (true){
// Need this loop to wait in case the PIR is keeping the power on.
}
}

// Query the card to make sure it's OK
uint8_t SD_CARD = SD_MMC.cardType();
if(SD_CARD == CARD_NONE){
// If we're here, there's a problem with the SD card.
// Turn the ESP off and wait for the next trigger.
digitalWrite(13, LOW);
Serial.println("SD FAIL 2");  // Uncomment for troubleshooting
while(true){
// Need this loop to wait in case the PIR is keeping the power on.
}
}
}

// We are now done the setup and should be ready to take pictures in the main loop() function.

void loop(){
uint8_t COUNTUP = 0;  // Create variable to take multiple pictures.
while (COUNTUP <=4){  // Take 5 pictures before shutting down.

// Take picture and read the frame buffer
camera_fb_t * fb = esp_camera_fb_get();

if (!fb){
// If we're here, there's something wrong with the data in the frame buffer.
// Turn the ESP off and wait for the next trigger.
digitalWrite(13, LOW);
while(true){
// Need this loop to wait in case the PIR is keeping the power on.
}
}

// If we're here, the image was captured. Begin the process to save it to the SD card.
// First, create the file name and path. Currently set to make files like /pic123.jpg
String path = "/pic" + String(PIC_COUNT) + ".jpg";

fs::FS &fs = SD_MMC;

// Now, create a new file using the path and name set above.
File file = fs.open(path.c_str(), FILE_WRITE);
if(!file){
// If we're here, there's a problem creating a new file on the SD card.
// Turn off the ESP and wait for the next trigger.
digitalWrite(13, LOW);
while(true){
// Need this loop to wait in case the PIR is keeping the power on.
}
}
else
{
// If we're here, the file was created. Now write the captured image to the file.
file.write(fb->buf, fb->len);
PIC_COUNT = PIC_COUNT + 1;  // Increment the picture count number each time there's a successful write.
if(PIC_COUNT >=65500){
PIC_COUNT = 0;  // FAT32 has a limit of 65534 files in a folder
}
}
file.close(); // Done writing the file so close it.

// Free the memory used by the framebuffer so it's available for another picture
esp_camera_fb_return(fb);

COUNTUP = COUNTUP + 1;  // We are done an image capture cycle. Increment the count.
}

// If we're here then we've taken the pictures and we are ready to shut down. Write the current file number to
// the EEPROM, then set D13 low.
preferences.putUShort("PIC_COUNT", PIC_COUNT);  // Store the picture count number in the EEPROM
// Normally you'd want to do a preferences.end() to properly close the EEPROM but since the intent is to
// shut the ESP down, it's not needed, and not having to open and close it every capture cycle speeds things
// up and saves some wear on the EEPROM.

//Serial.println("Shutting down."); // Uncomment for troubleshooting
digitalWrite(13, LOW);
delay(500);

// The ESP should be shut down at this point. If the PIR is still triggered or has re-triggered and is keeping
// the MOSFET on, then set D13 high and allow the program to loop again to take another five pictures.
digitalWrite(13, HIGH);
//Serial.println("Looping back.");  // Uncomment for troubleshooting

}

Hopefully there are enough comments to make heads or tails of it. A few notes:

• This is for the AI-Thinker module with the OV2640 camera only. It’s the only one I’ve tried. It will probably work with other models of ESP32-CAM, but you will need to check the #define for each pin, see how much of what kinds of storage are available, and what settings the camera you’re using requires.
• The ESP32 Arduino-compatible “EEPROM” library is deprecated; the new way to do things is with the “Preferences” library.
• Before programming, set the board type to “AI Thinker ESP32-CAM”. Again, this is for the AI-Thinker module only.

If you’re looking for documentation on the ESP32-CAM or the libraries I used to get this working, it’s all in the following links. For the libraries, be sure to check out the examples and .h files for options and how various things work:

I haven’t put it outside yet, but as a test of the ESP and circuit, I changed the program so the ESP would take and save as many 1600×1200 JPEGs at compression level 9 as quickly as possible. The circuit was powered by four grocery-store branded AA alkaline batteries that were unused but of unknown age. It ran for 16 hours 21 minutes and took 23475 pictures before the batteries died. Not bad!

I know there is a lot of room for improvement – both in the circuit and in the program. When building this, I was limited to what I had on hand – a P-channel MOSFET might be a better choice, and I’m sure there’s a better way to do things than with the 4N37. For now, though, I’m pretty happy that it works reliably. Now I need to put it in a box and get some pictures of those critters in the yard.

If you made it this far, congratulations! If you build this or have a better way to do this, I’d be curious how it turned out! Feel free to drop a comment here or send me a message using the contact form!

## Ah, The Good Old Days…

Anybody remember these? Anyone else use wire as bookmarks? Back in the day (when low-power Schottky was actually low power) one of these were worth their weight in gold. For the longest time I was stuck sharing a very worn one (no cover, lots of missing pages, etc) with a group of other people until I saved up enough to get my own.

With the internet, looking up a datasheet is easier and faster than with this old book, and I’ve been tempted to throw it out a couple of times… but last night I was stuck on a problem and found myself leafing through it for ideas. I’m glad I kept it!

## Getting Clone / 3rd Party Arduino 32U4 Boards Working On The Raspberry Pi / Linux

Ever since Ms Geek introduced me to the Arduino last April, I’ve been having a blast tinkering with them and building things. I’ve bought quite a few Arduino boards now, most of which are clones or third-party boards. A couple of them have been duds out of the package, but when I can buy four cheap boards for the cost of one “real” board, it’s tough to rationalize paying the extra money. Plus, a bit of time with a magnifier and a soldering iron fixed all but one of the boards that were DOA.

Up until two days ago, I’ve used the Arduino IDE exclusively under Windows and things have been working well. Two days ago, I installed the IDE (1.18.10) on a Raspberry Pi 4 running Raspbian Linux (Buster), and had no problems connecting to and uploading programs to MEGA328p boards like the Nano and Pro Mini.

A nice thing about Arduinos is that from a programming perspective, they’re pretty tough to permanently kill. As always, though, I make no guarantees that any of the following (or anything on this site, for that matter) will work for you. It’s worked for me so far, but it may not work for you. Who knows – it may not even work for me tomorrow. Anyway…

When I tried to do anything with a board that had the MEGA32U4 chip, though, the IDE couldn’t find it. There were no serial ports listed that pointed to the board:

The same thing happened with every single Pro Micro clone I tested (four of them from two different orders months apart) and a Keyestudio Leonardo clone. After hours of research and tinkering, I wasn’t any closer to fixing the problem, although I did disable modemmanager, which is something I discovered you should do if you want to do Arduino stuff under Linux anyway.

Just before I gave up, I extricated an authentic Leonardo from the project it was wired to and plugged it in. Immediately, a new serial port (ttyACM0) appeared in the IDE and showed a device connected to it:

Finally, I had a starting point. Here’s an assortment of my 32U4-based Arduino boards (the Pro Micro is basically a little Leonardo):

An easy thing to check was the 32U4 chip. They all looked pretty much the same, other than the authentic Leonardo being quite a bit older than the other boards:

Since the boards all worked under Windows, I was pretty sure that despite some different components (regulators, crystals, etc) on the clone boards, all were electrically sound. So, I took a look at what the Pi said when I plugged the boards in. It turned out that Linux was seeing the clones when I plugged them in but they disconnected almost immediately:

Then I plugged the genuine Leonardo in. It showed up in dmesg and stayed connected:

Aside from not disconnecting, the genuine Leonardo showed up a little different in dmesg. Take a look at the “New USB device found” and “New USB device strings” lines.

On the genuine Arduino: idVendor=2341, idProduct=8036, bcdDevice=1.00, Mfr=1, Product=2, SerialNumber=3

On the clones: idVendor=2341, idProduct=0036, bcdDevice=0.01, Mfr=2, Product=1, SerialNumber=0

Everything except the idVendor is different. According to the USB ID Repository, the vendor ID of 2341 belongs to Arduino. The Product ID of 8036 is associated with the Leonardo, while the Product ID of 0036 is associated with… a blank spot:

I still remember going through USB hell when the original Raspberry Pi came out, so to make sure the Pi or Raspbian weren’t causing the problem, I connected the boards to an Ubuntu VM. I got the exact same results, which took the Pi off the list of suspects. I did some more research and discovered that the vendor, product, and other information presented by the MEGA32U4 isn’t hard-coded in the chip or selected by OTP fuses – it’s firmware. As in, the bootloader. Which can be replaced.

I found some instructions for replacing the bootloader on an Arduino here and used a Nano (which works fine in Linux) as the programmer. I hooked one of the 32U4-based clones to the Nano, set the IDE to use the Nano as a programmer (Tools -> Programmer -> ArduinoISP), then burned the new bootloader (Tools -> Board -> Arduino Leonardo, then Tools -> Burn Bootloader).

I hooked the clone Leonardo up to the Raspberry Pi and took a look at the output of dmesg:

After burning the new bootloader, all of the clones showed the same values as the genuine Leonardo for idVendor, idProduct, bcdDevice, Mfr, Product, and SerialNumber… and all of the clones now show up as /dev/ttyACM0 or /dev/ttyACM1, and they are all visible and programmable in the Arduino IDE on a Pi 4 running Raspbian.

So, to sum up… if you can’t program a 32U4-based board with the Arduino IDE in Linux but it works fine in Windows, you can try the following:

• Check dmesg and confirm that the board is connecting and disconnecting. If so, then…
• Disable modemmanager (good idea to do this anyway), then…
• Check dmesg and see if the board has idVendor=2341 and idProduct=0036. If so, then…
• Buy or build an Arduino ISP. Once you’ve got it, then…
• Hook up the board in question to the ISP and burn the correct bootloader to the board. Once that’s done, you can…
• Check dmesg again and see if the board now has idProduct=8036. If it does, then…
• Give it a try (don’t forget to change the IDE settings (i.e. Tools -> Programmer -> AVRISP mkII) back to program your board directly instead of using the IDE!)

Good luck!

## Removing IR Filter From ESP32-CAM

I’ve been playing around with the ESP32-CAM modules a lot lately, and after spending some time searching through pages of code, I think I’ve finally got a handle on configuring the camera and getting it to start up cleanly.

With its boot time being about a second and the price being as low as it is, I think there are tons of applications for it. Everything from an always-on security camera to an occasionally used trail/wildlife camera, eye/brain/wifi for a remote-controlled car or robot… all kinds of things.

I want to put some cameras outside to catch the various animals that are leaving tracks in the snow, and these things are so inexpensive that they’re aalllllmost disposable. That’s not to say that I plan on putting them in situations where they’ll be ruined, but if one gets gnawed on or water gets to it, it won’t be the end of the world.

One thing that the stock camera can’t do, though, is see well in the dark. Even with bright IR illumination, the IR filter in the camera does a good job of keeping the picture stubbornly dark. Back when the Raspberry Pi camera module first came out, I fought with the IR filters in a couple of them, so I figured I’d take a shot at removing the filter from one of the OV2640s that came with the ESP32-CAM. Turns out it wasn’t too tough, although it took some patience.

PLEASE NOTE THAT THIS IS ONLY FOR THE CAMERA SHOWN IN THE PICTURES BELOW. I DON’T KNOW IF THIS PROCEDURE WORKS FOR OTHER CAMERAS.

OH, AND ALSO PLEASE NOTE THAT I’M NOT RESPONSIBLE FOR ANYTHING YOU DO, SO IF YOU WRECK YOUR CAMERA OR CUT YOURSELF OR BURN YOUR HOUSE DOWN OR WHATEVER, THAT’S ON YOU. BE CAREFUL!

First, remove the camera from the ESP32-CAM module by lifting the latch that’s holding the connector in place. Take note of where the glue is on the lens – it shines a little more under a light than the plastic does:

After the camera is loose from the module and you’ve located the glue, very carefully and gently run a sharp knife in the seam between the lens barrel and the camera casing:

Once you’ve cut the glue, grab the lens barrel with one finger and thumb and the casing with the other and gently turn the barrel counterclockwise. If it doesn’t turn, go back and run the knife through the glue again. If it does turn, you’ll be rewarded with a view of the camera sensor…

… and you’ll be able to see the IR cut filter too:

The filter in this particular camera is held on by a very thin ring of plastic and possibly some glue. I slowly shaved the plastic away until I got right down to the filter, then very carefully pried the filter off with the knife:

After that, it’s just a matter of making sure there’s no dust on the inside of the lens or on the camera sensor, then very carefully threading the barrel back into the casing. It is really easy to cross-thread it, so take it slow and be careful. If all goes well, you should end up with a put-back-together camera:

One of the other advantages of cutting the glue and being able to thread the barren in and out is that you can adjust the focus on the camera now. Want to look at something nearby? Back the barrel out a bit. Something far away you want to see? Spin it in a bit.

Anyway, put the camera back in the ESP32-CAM and flip the lever down on the connector to secure it all. Voila, you now have an IR-sensitive camera!

Oh, one of the things that sometimes goes unmentioned when talking about making your camera “see in the dark” is that after you remove the IR-cut filter, your camera will no longer work very well in colour, particularly where there is a lot of extra IR light bouncing around (i.e. daytime). The colours will not look normal and it will seem like the picture is out of focus. You can fix that by putting an IR-cut filter back into or in front of the camera, but I just want to watch animals wander around in the dark so B&W is just fine for me.

Here’s the difference removing the filter makes. Same remote, same button (3), and same camera settings. Here’s before I removed the filter:

And here’s after:

A bit of a difference! I’m looking forward to trying this out and seeing which wavelength of IR LEDs it will be the most sensitive to.

## ESP32-Cam Antenna Workaround

A while ago I got my hands on a pair of ESP32-Cam modules. They were crazy cheap and it looked like I had a great little device that would work well as a security or wildlife cam:

It’s tiny!

It was easy to program using the Arduino IDE and the example program, but no matter what I tried changing in the program it wouldn’t connect to my home wireless network.

There’s a lot of documentation out there on the ESP32 devices, but there are parts that seem to be missing. I wish I’d bookmarked the link, but I came across the comment section of one article that mentioned a 0R surface-mount resistor that selected the external antenna jack or the on-board PCB antenna. It didn’t take too long to find, and it was set to use the external antenna jack:

That explained why it wasn’t talking to the wireless – no signal. I could fix the problem in two ways: plug in an external antenna, or move the tiny little resistor so that it connected the ESP32 to the PCB antenna. I don’t have any connector that will fit that jack, so I decided to move the resistor.

**NOTE** Doing any combination or number of the following steps will definitely, POSITIVELY void the warranty of the device. It may also increase the electrical noise produced by the unit and/or its susceptibility to other sources of electrical noise. It may also cause operational or stability problems. It may also cause the device to produce wifi signals that are more powerful than allowed by law. Proceed at your own risk.

Unfortunately, between not being able to see it very well (even with a magnifier lamp), having a soldering iron with too large a tip, and having tingly fingers and shaking hands, I ended up with this:

Not only did I not manage to solder the resistor in place, I also ruined the PCB pad that connected back to the ESP32.

By this point I was getting pretty frustrated. Actually, I don’t think I’ve been that frustrated in quite a while. Fortunately, I had another ESP32-Cam that I could apply the knowledge that I gained on this one and not make the same mistake twi-bahahahaha… yeah, I ruined that one, too.

The two of them together cost me less than \$15, but I hate throwing something out when it’s still “alive” (they both still booted and tried to connect to the wifi), and I was looking forward to playing around with them. Seeing as how they were pretty much garbage at this point anyway, I figured I’d see if there was anything I could do to get them working.

**NOTE** See previous note. Seriously.

I could see the PCB trace leading under the metal can, so I figured that if I was careful, I might be able to remove the can and find another soldering point to attach an antenna.

To remove the can, you don’t need a hacksaw or a Dremel… just a pair of very fine-tipped pliers or snippers. there is a small hole in the can in the inside corner right by where the antenna selection resistor pads are. Carefully grab it with the pliers/snips and pull straight upwards to make some room:

Slide the pliers in a bit more and lift up and away from the PCB antenna (towards the top of this picture). It may take a bit but the can end by the PCB antenna should break away:

Reposition and pull the can up and to the left – the right side and top should come free easily as well:

Reposition again and the last bit should break away with very little force:

The antenna connection point is – very nicely – right by where the antenna selection pads are, right between two surface-mount components with big pads and lots of solder:

So now to make the antenna. I used a piece of 26ga solid wire-wrap wire, but anything nice and thin will do 22ga is too big.

Cut a piece about 6.5cm long, then strip about 1-1.5mm and bend the stripped part at a right angle. Tin the wire, make sure it’s got a good coat of solder on it:

Bend the insulated part of the wire to whatever shape and angle helps you hold the stripped part in place at the new attachment point. Use the iron sparingly – if you’re heating and heating and it doesn’t seem to be melting, back off, give it a minute, and change your angle before trying again.

If you’re careful and lucky, you’ll end up with something like (or much better than) this:

Gently bend the antenna into the orientation you prefer, and you’re done!

The next thing to try is powering it up (unless you haven’t programmed it, in which case you should probably do that now). With no can and a crappy detuned wire antenna, the ESP32-Cam was able to finally connect to my home wireless network:

And the camera’s video streaming works like a charm:

I suppose that instead of hanging a wire antenna off that point, I could’ve run a jumper from the new attachment point to the pad that connected to the PCB antenna (or the antenna jack, for that matter), but a) I just thought of it, and b) I don’t mind the crappy antenna look.