Arduino Part 6: Batteries

A few weeks ago, I picked up the Arduino Uno that I've had for a while, and started thinking about what I should build with it. I decided on modifying an old toy rover that I had, and making it wirelessly drivable... but with 2 way audio and a camera feed. single-frame camera-broadcast.


This is the sixth part of several posts. I will be explaining about how I set up the power supply of my rover.

Links:
Part 1: Initial Plan and Reverse-Engineering
Part 2: Preparing the Rover Chassis
Part 3: Motor Control
Part 4: Bluetooth!
Part 5: Mounting
Part 6: Batteries
Part 7: The Test

To start with, here are some facts about the situation:
  The rover is only designed to hold three AA batteries (4.5 volts).
  According to my multimeter when monitoring the long-disconnected-and-very-bored rover brain, the motors are designed to get the full 4.5 volts from the batteries.
  Arduino requires, at the very minimum, a stable supply of more than 5 volts.

This makes it more difficult than I originally thought, which was: "Let's just hook up a 9-volt!" Browsing Google search told me otherwise...

I wasn't about to go looking for how to use a switching regulator, so plan B was to use AAs. Three for the motors and four for Arduino means seven batteries total. That's even more than I remember RC helicopter controllers and the Lego Mindstorms NXT using.

Here's Plan B-and-a-half:

1. Test to see how few batteries can successfully power the motors (Two)
2. Add a single battery to the rover (For a total of four)
3. Have a third wire coming from the middle of the battery chain, thus using all of the batteries to power the board while two of those are also powering the motors when necessary.
4. Do extensive tests with my multimeter to ensure that the voltage never drops below 5 volts. If it does, fiddle with the circuitry until it's stable.

After the test was set up, I found that voltage dropped drastically at the moment I started the motors. Voltage would then very slowly increase as the motors weren't requiring as much power to continue spinning. I added the biggest capacitor I had (1000 μF, also referred to as 1mF). This made good progress, but the capacitor was discharging too quickly. This delayed, but hardly reduced, the voltage drop.

To combat this, I added a 1KΩ resistor between the power / capacitor leads and the Arduino Vin wire. This was a mistake because then, while tests showed satisfactory results, trying to power the Arduino didn't work. The resistor was choking Arduino, and my multimeter couldn't tell because it has a very high internal resistance.

My next attempt was to add the resistor in series with the capacitor, leaving the power wire directly connected to Arduino's Vin pin (Or, at first, the multimeter's lead). This was successful. Charging the capacitor wasn't noticeably affecting results, and voltage would slowly drop, but not to the fatal 5V level, when spinning the motors.

The final test went very well. Arduino turned on. Power was stable. (Manually) activating the motors didn't break anything, and Arduino didn't die when I kept the motors going for several solid seconds. Success!

To get the extra battery into the rover chassis, I had to take a AA holder from another kit, solder it up, and tuck it into the Rover. This would, unfortunately, require me to dismantle the rover in order to change all of the batteries, but it's a personal project. It doesn't need to be super simple, especially since I've already got the Arduino sitting directly on top of the other batteries.

More tips... They just keep coming...

1. Do the math! Connecting random batteries would only have gotten me an overpriced blue paperweight.

2. Know your electrical engineering! The long-standing Kerbal convention of "More Power!" would have lead me to use more batteries rather than a capacitor. This would be less efficient and more likely to overheat the Arduino.

Edit: 7/6/17: Updated list of links

Arduino Rover Part 4: Bluetooth

A few weeks ago, I picked up the Arduino Uno that I've had for a while, and started thinking about what I should build with it. I decided on modifying an old toy rover that I had, and making it wirelessly drivable... but with 2 way audio and a camera feed.

This is the fourth part of several posts. I will be explaining about how I connected my phone to Arduino via Bluetooth.

Links:
Part 1: Initial Plan and Reverse-Engineering
Part 2: Preparing the Rover Chassis
Part 3: Motor Control
Part 4: Bluetooth!
Part 5: Mounting
Part 6: Batteries
Part 7: The Test


The rover was coming along nicely, however I still couldn't control it remotely. Options for communications were:

A. WiFi: Hosting a webpage on the rover, and inputting commands from a computer or phone. Advantages: Higher bandwidth, range while in free WiFi. Disadvantages: would only work in free WiFi, and would require a special "shield" that goes on top of Arduino.

B. Bluetooth: Connecting via phone and driving with an interface. Advantages: Works without WiFi, doesn't require special shield, uses Arduino's built in serial capabilities for bit-by-bit communication. Disadvantages: Lower bandwidth than WiFi, lower range while in free WiFi zone.

Let's compare:

WiFi

Bluetooth

Bandwith

Flexibility

Control quality

Weight and size

Cost

Cool factor (0.25 points)

Sorry, WiFi, but I think Bluetooth wins 4.25 to 1.

This means that a continuous video feed is not really an option. The highest serial 'baud' that Arduino can muster is 115200. If I was to try to send a 640 by 480 pixel video feed of 10 frames per second (using grayscale instead of color), that would be 640 * 480 * 10 = 3,072,000 necessary bytes per second. That's about 3 megabytes per second, 26.66 times what is available. Kind of disappointing. On the bright side, I'll still be able to press a button on my phone and receive a single picture... over the course of 2.66 seconds.

I looked online to see how people usually do Bluetooth connections with their Arduino and discovered that the most popular Bluetooth module was the HC-06. After some research, I discovered that it was capable of using a higher baud rate than Arduino could handle, but had a lower default. It also ran on the standard 5V that Arduino outputs... Not bad. There were also guides everywhere, so the HC-06 it was.

It's $9.49 on Amazon with free 2 day shipping (Amazon Prime only). When I got mine, I think there was a 50% sale.

While waiting for it to come in the mail, I had to slap together some Android code. I downloaded this guy's example and set about reverse-engineering it. There was a disturbing number of references to Chihuahuas. I've been working on my own app for a while, so all I had to focus on was the Bluetooth code. Please note that the project folder is formatted for Eclipse. I had to manually copy the source code into Android Studio (I only took part of it).

Modifying the code was a big task (Android code terminology ahead!). The first thing I did was delete the Chihuahuas. Then I ditched the layout and created my own (one that will be easier if I'm looking at the rover rather than my phone). I changed the search term for the paired devices from "HC-05" to "HC-06". Then I set up the controls to broadcast a single-byte command every quarter of a second depending on which buttons were pressed, rather than as the buttons were pressed. I added another activity for pre-driving setup (custom search term, more stuff to be added later). I also worked on cleaning up and custom-formatting the code.



Then there was the Arduino code to design. A quick Google search revealed that it's as simple as making a SoftwareSerial on the proper pins. I didn't bother downloading anything for it. The complete code sends a message through the HardwareSerial to my computer every time it receives a command.

On the day of projected arrival, I discovered that the relays had come early, and didn't notice the big yellow bubble-protected envelope containing the HC-06 (facepalm). I realized my mistake in the late afternoon...

On the hardware side of things, it's pretty simple to set up.

Here's where the pins go:
GND --> Arduino's GND
VCC --> Arduino's +5V
RXD --> Voltage Divider to TX on Arduino (More on this in next paragraph)
TXD --> RX on Arduino (Don't send these to the 0 or 1 pins. These are used to communicate with the computer)

After doing my research, I noticed that some people skipped the voltage divider, and some people regard it as necessary. Both groups say that it works for them, but the RXD pin is technically not supposed to be exposed to the full 5V of Arduino's pins. Here's a link to the page I learned most about how to do it.

Looking into my inventory, I don't have any 20KΩ resistors. That's ok. I'll just use two 10KΩ resistors in series.

Plugging everything in and verifying hardware connections... No bluetooth connection. More troubleshooting:

Switching RX and TX pins: Nope.
Staring at it to use the Force: Why??
Checking on Bluetooth menu in phone settings: Getting somewhere. It sees the HC-06.
Tapping on it and inputting the default password (1234): Bingo! Now the phone and Arduino are "paired" together!
Trying again: It works! Every quarter of a second while pressing a button on my phone, Arduino sent a message of "Received right command" or "Received forward command"!

I wonder why the Chihuahua code didn't automatically pair with the device... Probably doesn't want to accidentally pair with random devices (smart).

I think I'll wait to post downloads of my own code until the post where the rover is driving successfully (Sure, I'll do it sooner if the comments ask me to). I encourage anyone wanting to, to try and make their own code. I only downloaded the Chihuahua code to figure out how to create and maintain a bluetooth connection (I'd been struggling). Anyone with the same problem can go ahead and get the Chihuahua code. The Arduino code is so simple it shouldn't be necessary to download.

Part 5 will be about my search for ways to mount my Arduino to the rover.

Tips jar is for you, not me!:

1. Be conservative. It wouldn't have been very fun to kill the bluetooth module via overvoltage.

2. Do it yourself (if you can). It's very important to learn what's going on under the hood, and the best way to do that is to make it yourself. If you're stuck, you can get some help, but still do your best to understand.

3. Don't give up troubleshooting!

Edit: 7/8/16: Updated list of links
Edit: 8/25/16: Updated list of links
Edit: 7/6/17: Updated list of links

Arduino Rover Part 3: Motor Control

A few weeks ago, I picked up the Arduino Uno that I've had for a while, and started thinking about what I should build with it. I decided on modifying an old toy rover that I had, and making it wirelessly drivable... but with 2 way audio and a camera feed.

This is the third part of several posts. I will be explaining about my struggles to power the default motors with my Arduino. Prepare for math!

Links:
Part 1: Initial Plan and Reverse-Engineering
Part 2: Preparing the Rover Chassis
Part 3: Motor Control
Part 4: Bluetooth
Part 5: Mounting
Part 6: Batteries
Part 7: The Test

Now that the rover was gutted and is ready for installation, the next thing to do is configure a circuit to power the motors. It can't just be plugging the leads directly into the board. Prior to most of my research, I knew there were problems associated with taking too much current from the Arduino. After my research, I had a much better understanding and could eventually power motors through Arduino.

The maximum safe current from an Arduino output pin is 20 mA. That's 0.02 Amps. Ohm's law states that Current = Volts / Resistance. Using basic algebra to get resistance by itself, we get R = V/A. If we plug in the output values of the Arduino (5 volts, 0.02 A max), we get R = 5/0.02 = 250Ω minimum resistance.

I downloaded the Arduino IDE to my Linux drive (Yes, I use Linux and I love it) and started refamiliarizing myself with the IDE and programming language, which is a variant of C. I'm very familiar with Java (As is obvious from my Minecraft mods), so the learning curve wasn't too large.



I looked up a motor-control circuit, and found this. I recreated it with a motor from an electronics kit.

Measuring the motor's resistance, I got 100Ω, which is fine because the transistor from the same kit was protecting the Arduino. Then I hooked up one of the rover's motors...

Nothing happened.

I tried again: Nothing.

I left the blink program on and listened to the gearbox: Nothing.

Then I touched the transistor...

and promptly pulled my hand back. It was really hot.

I looked up the data sheet for that particular transistor and as far as I can tell, it's got a maximum current rating of 200mA. Measuring the resistance of the motor with my multimeter, I got... holy cow! 0.7Ω! That's supposedly pulling 7.14 A based on 5V/0.7Ω. I tried doing some math to find the right resistor to do the job. I got about 30Ω. I didn't have any of those handy, and figured that if they were designed for use with 4.5 volts from 3 AA batteries and 50Ω didn't work with 5 volts, then this transistor wasn't going to work.

Guess what. 2 100Ω resistors in parallel didn't work...

The headache intensified.

Finally, I gave up trying to use transistors. I went searching online for some relays. Luckily, I found some really good ones, the HE3621A0510, that were just what I needed (Hooray for Capitalism!)

They are rated for 5V, and have 500Ω resistance, thus requiring 10mA, and they only have 4 pins... That's exactly what I needed.

Here's the datasheet as of May 6th.

Due to the fact that it's the 510 version, it has an internal diode, (which in my relief I neglected to learn more about).

They came in the mail a day early, and I forgot to look in the giant yellow bubble-wrappered envelope for the bluetooth module (I'll talk about that in part 4).

I mapped all of the resistances with my multimeter. The two outer pins (Shown as green in the picture) had no conductivity, to be connected when electricity flows through the inner two, which had 515Ω. I plugged it into my project, and loaded the blink script...

Nothing. I got nervous, verified connections...

Nothing. The headache from the transistors returned...

I looked back and researched the exact model of relay... Mentions something about a diode.

I researched the Diode Suppressed Coil: A method of protecting delicate electronics from current surge caused by collapsing magnetic field in the coil. Thus, the relay only works one way. Here's more information on how relays and DSCs work.

I switched the direction of the relay, and... Poof! It worked perfectly! I've marked the negative and positive leads in the picture with black and red, respectively.

That concluded my motor-control headaches.

Part 4 will be about bluetooth serial communications and connecting to an Android phone (with a custom app)

Now for more DIY tips:

1. Do your research! I could have saved myself a lot of headache-hours and a burnt finger by just learning more about the transistor and relay I was trying to use.

2. Use Linux (Preferably Mint)

Edit: 6/1/16: Updated list of links

Edit: 7/8/16: Updated list of links
Edit: 8/25/16: Updated list of links
Edit: 7/6/17: Updated list of links