Saturday, August 13, 2011

Using an External Button & LED with NETMF

In continuing with the BlinkyLEDNETMF for Electronics Noobs series of tutorials, today I’m going to demonstrate how to connect a button/switch to a Panda-II NETMF microprocessor and how to respond when the button is pushed.  Since I need to do something when the button is pushed, I thought this would be an opportune time to also show how to connect an LED to the board and light it up also.

Nearly every NETMF basic tutorial out there starts with a very simple program that uses the on-board button to light up the on-board LED.  That’s just great and is absolutely fantastic the first time you write a NETMF program.  However, as soon as you start working on a real project you soon discover that the on-board button & LED have very limited uses since they can’t be exposed outside of your project enclosure. Then the question comes up – how do I add more buttons?  How do I light up an LED on my project box?  How do I use more than one button or LED?

I’m going to demonstrate how to do this in the most basic situation – the one that will work for 95% of all hobbyist’s needs.

Panda-II Connections

For this project, you’ll need a momentary button, an LED, a breadboard, one 10K ohm resistor, one 100 ohm resistor, and some wires.

Like most things, there are many ways you could wire this up.  This is just one.  What I did was to first plug my button into the breadboard and wire one side of the button to the ground line (very top row on the breadboard).  Then, I connected a 10K ohm resistor (brown, black, orange) to the power (5V) line (2nd row on the breadboard) and used it to connect power to the other side of the button.  The 10K resistor prevents us from creating a “short” which could fry the electronics by allowing too much current to be pulled through the electronics.  Next, we put a 100 ohm resistor (brown, black, brown) on the power side of the button then connect the other side of the resistor to the analog 0 pin on the Panda-II.  This resistor prevents too much current from getting into the board and damaging it.

Panda-II Connections.Fritzing

Now that we have the button wired up, we’ll wire up the LED.  Since only a very small amount of current flows through pins of the Panda-II board we can connect the anode/positive side (longer wire) of the LED to the D5/PWM5 pin and wire the cathode/negative (shorter) side of the LED to our ground line.  That’s it for the wiring.

I do want to note here that if we were wiring the LED to an external power source that we would need to add a current limiting resistor to one side.  To determine the correct size resistor, you can use an LED calculator.

Now that we have the hardware all wired up, it’s time to write some software that responds to our button being pressed by lighting the LED.

In Visual Studio, we create a new console project and add the following references:

FEZPanda_II_GHIElectronics.NETMF.FEZ
Microsoft.SPOT.Hardware
Microsoft.SPOT.Native
mscorlib

In our Program.cs file that is automatically created, replace what is there by default with the following.

using System;
using System.Threading;
using Microsoft.SPOT;
using Microsoft.SPOT.Hardware;
using GHIElectronics.NETMF.FEZ;
 
namespace ExternalButtonAndLedTutorial
{
    public class Program
    {
        // Tell the CPU we're going to use digital pin #5 to send a signal to the LED.
        static readonly OutputPort _led = new OutputPort((Cpu.Pin)FEZ_Pin.Digital.Di5, false);
 
        public static void Main()
        {
            // Tell the CPU we're going to receive an analog signal from analog pin #0.
            var blueButton = new InterruptPort((Cpu.Pin)FEZ_Pin.Interrupt.An0, true, Port.ResistorMode.PullUp, Port.InterruptMode.InterruptEdgeLow);
            
            // The second parameter in the line above ("true") tells the CPU that we want to use the "glitch" filter.
            // If you disable this line then run the program, you will notice in the debug windows that if you are not very fast
            // at pushing the button that the event will actually get fired multiple times.  We only want the event to fire once.  GHI gives
            // us this property that we can set to put a boundary around our first event and prevent redundant events from being fired.
            // Adjust the number of tick
            Cpu.GlitchFilterTime = new TimeSpan(TimeSpan.TicksPerSecond/2);     // 1 Tick = 100 nanoseconds
 
            // When the button is pushed it will signal an interrupt in the CPU which will in turn cause an OnInterrupt event to be thrown.
            // Our BlueButton_OnInterrup function will be called when the interrupt/event is fired.
            blueButton.OnInterrupt += BlueButton_OnInterrupt;
 
            while(true) Thread.Sleep(10);
        }
 
        private static void BlueButton_OnInterrupt(uint port, uint state, DateTime time)
        {
            _led.Write(true);       // Turn on the LED
            Thread.Sleep(1000);     //  wait 1 second...
            _led.Write(false);      // Turn off the LED
 
            Debug.Print("Blue Button Pushed! " + DateTime.Now.ToString());
        }
    }
}



I’ve chosen to document the code as comments in-line, but if anything isn’t clear please leave a comment and I’ll clear it up.


Make sure and set your project properties to deploy to the USBizi device then run it.  If you’ve done everything correct then when you push the button the LED should come on for half a second then turn back off. 


Now that you have one button and one LED working,  get creative!  You can reproduce this many times and as many different ways as your imagination will allow.  To use more buttons or more LEDs, just use any of the other pins and be sure to specify in your code when you define your buttons or LEDs which pin they are connected to and you’ll be fine.  Put these new skilz to work and be sure to post your results in the comments!

NETMF for Electronics Noobs Tutorials

NETMF Logo
I’m starting a series of tutorials for people just getting started with .NET Micro Framework (NETMF).  My target audience will be people with an intermediate level of experience with C# but very little if any experience with the electronics side of things.  There is plenty of good info out there regarding using C# and there are several good tutorials on getting started with NETMF but in my opinion most of these assume the user is already somewhat experienced with electronics.  This is where I hope to bridge the gap.  I will not go in-depth on either the C# (NETMF) or the electronics side.  I will show you just enough to help you get started having fun with NETMF and then point you to other sources where you can learn more.
Although in most of my tutorials, I will use the GHI Panda-II microprocessor board the techniques can directly be applied to netduino or most any other NETMF device with minor code changes.
I’ll update the following list of topics as I add posts (I’ll try to add one weekly).  So, bookmark this page and follow my blog if you want to keep up.  If you have any specific topics you would like to see covered, please feel free to email me at ian (at) houseoflees.net or post to the comments below.
TUTORIALS:
QUICK TIPS:
FURTHER DISCUSSION
  • TinyCLR.com – great place to ask questions regarding anything about the GHI Panda-II microprocessor and NETMF.
  • netduino.com – great place to ask questions about the netduino microprocessor and NETMF.
  • netmf.com – the official NETMF community forum
NETMF

Monday, August 8, 2011

Basic XBee on NETMF

As I wait for parts for the Omnicopter, one of the things I’m getting started is the XBee radio communications between the remote/computer and the copter.  My initial design will attempt to use XBee radios for all control of the copter as well as in sharing diagnostic & control data between the base station [remote & or PC] and the copter.  I fear that XBee may end up not being responsive enough and have enough throughput to accomplish all things but that’s where I’m going to start before attempting traditional R/C radio techniques.

Xbee Series 1I want to be able to control the Omnicopter both by a remote control I’m building based on a NETMF microcontroller and with software on a PC.  The most basic of things is to make sure that the NETMF microcontroller I’ve chosen for the copter, a GHI FEZ Panda-II, can communicate with my PC. 
For the Panda-II I have bought and assembled an Adafruit Protoshield and an Adafruit XBee Adapter.  I already have some XBee Modules (series 1) from a previous project.  So, I’m using them for now.  Once I’ve proven the copter can fly and I’m brave enough to push the range limits of the XBee series 1 (300 ft.) then I can simply buy new XBee Pro radios and drop them in place of the existing radios to get a range of 6-15 miles (line of sight).
To test communications, two things are required – an XBee connected to the PC and a NETMF microprocessor with an XBee and the necessary software on both sides.

Setting Up An XBee on NETMF

An XBee radio communicates using basic RS232 type of communications.  All that’s required to connect it to the Panda-II is to give it 5V power, ground, and transmit & receive signals.
XBee Test on Panda-II w Caption

For a better view of the wiring connections, I drew it up in Fritzing.
XBee Test on Panda-II.Fritzing

Note that the receiver (RX) on the XBee is connected to the transmitter (TX) on the Panda-II and vice-versa.  This may not be obvious to those first getting started.

Next, we’ll need a NETMF C# program to install on the Panda-II.  Visual Studio 2010 (VS2010) Express or better can be used to compile and deploy the program.  GHI has a great tutorial on getting started if this is your first time to create a NETMF program.  So, I’ll jump straight to the code.  I’m using the MFToolkit open source code by Michael Schwartz to simplify the code required considerably.  An alternative assembly that is somewhat easier to understand and seems to satisfy basic XBee needs is Grommet.  Since I expect this project to get much more sophisticated over time, I’ve decided to stick with the MFToolkit. 

Once you download the MFToolkit (or Grommet) code from Codeplex and create a new NETMF C# project, you can simply copy this code into your project to have the NETMF device send a “ping” every 5 seconds.  The ping is basically the word “Hello” plus the current time in Ticks.  The following code is not meant to be “production ready”.  It’s just enough to test that the XBee is working.
public static void Main()
{
    var ledBlinkerThread = new Thread(BlinkOnboardLed);
    ledBlinkerThread.Start();

    var pingThread = new Thread(SendPingWithMFToolkit);
    pingThread.Start();
}

private static void SendPingWithMFToolkit()
{
    var xbee = new SerialPort("COM1", 115200, Parity.None, 8, StopBits.One);
    xbee.Open();
    xbee.DataReceived += XBeeDataReceived;

    while (true)
    {
        string output = "Hello " + DateTime.Now.Ticks;
        byte[] bytes = System.Text.Encoding.UTF8.GetBytes(output);
        xbee.Write(bytes, 0, bytes.Length);
        xbee.Flush();
        Thread.Sleep(5000);
    }
}

static void XBeeDataReceived(object sender, SerialDataReceivedEventArgs e)
{
    int bytesReceived = ((SerialPort)sender).BytesToRead;
    var bytes = new byte[bytesReceived];
    ((SerialPort)sender).Read(bytes, 0, bytes.Length);
    var received = new string(System.Text.Encoding.UTF8.GetChars(bytes));
    Debug.Print(received);
}
Note that the second parameter of the SerialPort constructor in SendPingWithMFToolkit() specifies the baud rate of the XBee radio.  By default, the radios will be set to 9600.  I have changed mine to 115200 prior to this writing.  If this is your first attempt with XBee then you should change this parameter to 9600 or whatever your actual baud rate is if yours have been changed to something different.  The important thing to know is that both radios must be set to the same baud rate and the software on both sides is configured likewise.

At this point you should be able to compile and run your code from the Panda-II without issue and it will start sending it’s “ping”.  Also, although I am using the Panda-II for this post just about any Arduino shield compatible NETMF device such as the netduino should also work w/o any change to the code or wiring.


Setting Up An XBee on a PC

Setting up an XBee on a PC is a bit simpler since all we really have to do is connect the XBee to a breakout or dev board with a USB cable and install the X-CTU software by Digi.  No reason to write our own code for this test since there already exists this great utility that does everything we need to test & configure an XBee.  I am using Windows 7 Ultimate (x64) but the steps are basically the same for any OS.

XBIB-U-DEV w caption

I have a MaxStream (now Digi) XBIB-U-Dev (Rev 3) XBee development board from a previous project that I’m using, but you can use just about any XBee adapter including the one from Adafruit that we used on the NETMF.  On my board, I’m able to plug a USB cable directly into it.  On most of the basic adapters, you will need an FTDI cable to convert the serial communications to USB.  You will also need to locate and install the appropriate drivers for your OS.

imageOnce you have your hardware all figured out, we need to download and install the X-CTU software.
After you install X-CTU, run it and it should come up with something similar to this except the Baud setting will probably be at 9600.  Set the baud rate to match that of your modem and hit the “Test/Query” button to see if X-CTU can connect to your XBee radio.  Assuming you can connect and all is good then we’ll now see if we are receiving the “ping” from our NETMF device.  Click on the “Terminal” tab and wait 5 seconds.  You should see some text appear in red on the screen and more to be added every 5 seconds.

If we have this then we know we are properly sending from the NETMF device and receiving by the PC.  Next thing to test is that we can send from the PC and receive from the NETMF device.

First make sure that we are running our program on the NETMF device through the VS2010 debugger and not just on the board itself.  You’ll notice on our NETMF code that we are capturing XBee received data with the XBeeDataReceived() handler function.  Anytime data is received, it will be simply printed to the Output Window.

imageTo send a packet of data from the PC we hit the “Assemble Packet” button in the X-CTU Terminal screen.  We’ll send a simple “Hello World” packet.  Type that in the textbox and hit “Send Data”.  You should see “Hello World” printed in blue text in the X-CTU console and in the VS2010 output window you should see the same text also printed there.

Mission accomplished.  We have established the most basic of communications and know that all of our hardware is setup properly and is working.  Now the challenge is for you to figure out how to best utilize this power in your project by designing the data you will send and how to handle it properly when it is received.

If you are still having issues with hardware or software, you should start by checking the forums at any of the following sites.
Good luck and happy transmitting!

Sunday, August 7, 2011

Omnicopter - Getting Off The Ground

So, I've decided to start up another .NET Micro Framework (NETMF) challenge after taking almost two years off from the tiny framework.  Some things have changed since I last used NETMF and I'm having to re-learn some things.  Sadly, one thing that hasn't changed so much is the lack of abundance of documentation on some of the basic things required to get some very basic electronics goals achieved with the framework.  So, as I start on this next project I plan to document some of the things that are either not already documented well somewhere else.

Unlike many NETMF developers I do not have a degree in electrical engineering.  I'm a software guy.  In all disclosure, I did spend three years studying EE at TTU before deciding to change to computer science so I do have a little more understanding of electronics than your average bear but I've never practiced it professionally and a lot of the language used in the NETMF forums and literature is a bit alien to me.  So, as I work through this project and figure some of these things out I'm going to try and explain them as simply as possible to others such as myself.  I believe the popularity of NETMF is only going to grow and one of the biggest set of new users is going to be software developers similar to myself and young roboticists.  Both groups I believe will benefit.

So, a brief intro to the project that I've decided to tackle.  I'm going to build a multi-propeller helicopter (i.e. quadcopter, quadrotor, hexacopter, octocopter, etc.)  In the first stages, it will start out as a quadcopter but I plan for that to change as progress is made.  More about that when the time comes.  Why build a quadcopter?  Uh...they're really cool!  But beyond basic flight, I have a lot of AI related ideas that I want to work on and the copter will be a great platform for developing those applications.  Equally important is that there really doesn't exist yet a quadcopter built on NETMF that I'm aware.  It seems that most people shy away or give up quickly because NETMF isn't a "real-time" platform and that it can't send commands quick enough to enable stability of a copter.  I hope to prove them wrong.  Maybe they'll prove me wrong but I don't think so.  If they do then I'll be sure to document in painful detail why NETMF isn't a valid platform for copters.

I don't want to give up too much too soon on my plans for the project since I don't even know for sure that it will ever even leave the ground.  So, for now there are two parts to this project:
  1. Build a NETMF controlled quadcopter
  2. Build a NETMF based remote control.
I've decided to name my copter the Omnicopter - the copter with many propellers and many functions.
 
Unfortunately, it seems the only place you can really get the copter motors and other R/C parts is from Hong Kong (HobbyKing.com).  So, there is a 4-6 week wait for parts. Ugg.  I've ordered the motors and other R/C parts and they're on their way.  While I wait for them to arrive, I'm doing some basic things with parts I can either acquire locally or order from the U.S. that arrive in a few days.  I'm building the basic frame for the copter & figuring out some basic wireless communications and protocols that I want to use and playing with LEDs when I get tired of doing the real work.  I'll cover all these things in future blog posts.  Let's go have some fun!

Wednesday, May 25, 2011

Customer Service - done RIGHT!

These days it’s very rare for me to have a customer service experience that truly is satisfying.  In fact, this is the first customer service experience I’ve had in my lifetime that I felt the need to put pen to paper (well, fingers to keyboard…) and tell the world about it.  It’s certainly still possible to occasionally get good customer service when going to a local store such as Lowes and approaching someone face-to-face.  But, what I’m talking mostly about is online or over-the-phone customer service.
All companies today are doing everything possible to cut costs and one seemingly obvious way to do this is by utilizing more and more technology to streamline customer service and eliminate as many people (that have to be paid wages…) out of the process as possible.  Usually, this just means that we as customers have to deal with reduced quality in service due to shouting at the phone to make it understand what we want or answering the same questions multiple times.  Why does the automated system make me type in my account number and then not share that info with the real person when I finally get to talk to them???  Maybe companies are saving money but they ARE NOT providing a better customer service experience.
logo-logitechOne company that I recently had an experience with has figured out how to use technology to properly streamline customer service to produce a service that is truly superior than the old way of just talking to someone over the phone is Logitech.
Harmony900We recently built a new home and as part of the design we decided to put all the A/V equipment in a closet out of site.  The problem with this design is that it requires that you have a RF remote that can work through walls.  When we made this decision I was unaware at how few options there exist for universal remotes that have this capability and how expensive the ones that do exist can cost.  After much research, I decided to go with the Logitech Harmony 900.  It is a truly wonderful remote that I highly recommend but that’s a review for another blog post…
So, we had the remote for about a month before the screen turned white and would not operate any longer (WSOD – white screen of death?).  I researched all the internet posts about it and some had been able to fix it with a BIOS upgrade but this wouldn’t work for me since the remote was also no longer detected by my computer.  All of the programming and configuration of this remote is done through your computer and the settings are all saved on Logitech’s website.  A nice feature that I didn’t fully appreciate at the time…
So, after exhausting my abilities and what little info was available on the internet regarding how to fix the problem I dreadfully went to the Logitech Customer Support site one night at around 8:30PM CST and filled out a short form that basically asked for the serial number and a description of the problem.  I submitted it and expected to get an automatic email back telling me to do everything that I’d already done and to start a dreadful back and forth “conversation” that would eventually end a month later with me having to send the remote to them and having to pay to get it fixed.
Did I mention that I bought this remote “used” on eBay for about half the retail cost?  Yea, that’s why I was dreading this whole thing so much.  I knew for sure that as soon as they asked where I bought it from that any warranty would be void and I would then be stuck having to pay to get it fixed.
To my surprise, about an hour after I submitted the form I got an email back from a Logitech “Customer Care” person (thanks Devin!)  The email did of course do as I expected and suggested that I try upgrading the firmware.  However, what I did not expect was that it also said the following…
[copied from the actual email from Logitech Customer Care]
If the firmware update does not fix this problem, then I do apologize for the inconvenience, but there does appear to be a hardware issue with your Harmony remote. Before I can submit your warranty claim, I would require some additional information. I would kindly ask that you reply back to this email with the following information:   1. Full Name: 2. Street Address (No PO Boxes please): 3. City/Town: 4. State: 5. Zip Code: 6. Country: 7. Phone: 8. Harmony remote PID (Found within the battery compartment): 9. Harmony remote P/N (Found within the battery compartment):   In addition to the above information, we will require digital photos of your Harmony remote's LCD screen. When submitting your photos, I would ask you to follow the guidelines outlined below:   1. Please place a piece of paper beside/underneath the remote with your reference number 110519-001728. 2. Set digital camera to Macro. 3. Ensure your Harmony remote is powered on. 4. Take 2 pictures. One with the remote powered on and the other with the remote powered off. 5. Take images from twelve inches directly above the remote, featuring the LCD display. 6. Submit the digital images as a jpeg. 7. Ensure the file size is on the smaller side, but the digital image allows us to view the remote's LCD screen. 8. Ensure the filenames are short and in one-string. i.e. remoteimage.jpg remoteimage2.jpg 9. Submit the digital images as an attachment with your reply and do not paste the image inline. 10. Please don't use the flash.
harmony900_whitescreenNote that they never asked where I bought the remote and they are making excellent use of technology by having me send them pictures of the problem rather than having me pay to ship the item back to them for them to confirm that it’s doing what I said it was.  So, I took the pictures and sent the info as requested.  I then expected to get the email back that said “it appears that you actually have a problem, please send the remote back to us and we’ll fix it or send you another one.”  Note the case.  This is what I got back…
Dear Richard, Thank you for contacting Logitech Customer Care about your Universal Remotes Your Return Merchandise Authorization number is: RMA 110519-001728 Your order has been submitted for processing today. You will receive the tracking number in 1 -2 business days We do not require you to return the defective unit. Please note that your original defective unit has been disabled from our system. Any update on this product is not possible. You can use it with the existing configuration until you receive the replacement product. Kindly dispose of the unit after you receive the replacement product. If you have any questions about your RMA or this process, please reply to this e-mail. Thank you for your patience.
WOW!  They’re just going to send me a new remote.  Easy as that.  No 500 questions or paying the shipping fees or anything.  They’re just going to stand behind their product and provide true customer service.  That is what I call “Customer Care”.  Also, notice again the wonderful use of technology.  Because all of the settings for the remote are kept on their website and that’s the only way to program the remote, they are able to use that technology to disable my old remote and therefore prevent people from taking advantage of the system and re-selling damaged remotes or faking a good remote as a damaged one.  Because they can do this they’re saving me the cost of shipping and them the cost of handling the remote once its returned.  A truly wonderful solution.
But wait…there’s more.  So, 2 or 3 days later I get a box in the mail from Logitech.  I thought, “wow!  that was fast!” and proceeded to open the box expecting to find just a remote.  Again, they exceeded all my expectations.  Inside was a full retail boxed product – remote, charging base, battery, power cords, USB cord, RF repeaters – everything.  So, now I have an extra battery and charging dock if those should fail.  The old RF repeater doesn’t seem to work with the new remote so I had to swap that out.  However, I’m not convinced that I couldn’t get it to work if I should ever need to utilize an extra repeater.
So, that’s my experience that I wanted to share and spread some of the love I recently received from Logitech.  Although, historically, I’ve not been big fan of Logitech (I always buy Microsoft mice & keyboards) that has changed.  I’m now their biggest fan and will be checking out their other products first when I need more electronics.
My hope in writing this is that other companies will take notice and follow their example and try to focus more on trying to produce better “customer care” solutions that provide a better experience (and more loyal customers) and fewer “customer service” solutions that actually serve the company’s bottom line more than they serve the customer.
Go Logitech!
[8/26/2011 UPDATE]
Unfortunately, the replacement remote quit working and I had to go through this procedure again.  It was just as easy the second time around and I have nothing but wonderful things to say about Logitech support.  Unfortunately, I don't know that I can continue to sing praises about the Harmony 900 remote.  When it works it's the best thing since sliced bread.  But, it seems to have some quality issues in the electronics and for what it costs it should either be much more reliable or have a much longer warranty.  Unfortunately, there really aren't any other options for RF universal remotes in the same price range.  If you know of any, please link to it in the comments.

Wednesday, February 2, 2011

Oracle View Columns Info

Today I had a need to be able to list the columns and their data types for a view that exists in Oracle.  Since I know TOAD displays this info in a grid in the schema view, I figured the easiest way to solve this problem would be to setup a trace and see how it does it…  Here are the results in case you or I need this again in the future.
select 	cols.column_id		
        ,cols.column_name as name
        ,nullable
        ,data_type as type	
        ,decode( data_type
                ,'CHAR'         ,char_length
                ,'VARCHAR'      ,char_length
                ,'VARCHAR2'     ,char_length
                ,'NCHAR'        ,char_length
                ,'NVARCHAR'     ,char_length
                ,'NVARCHAR2'    ,char_length
                ,null) as nchar_length
        ,decode( data_type ,'NUMBER' ,data_precision + data_scale, data_length ) as length
        ,data_precision as precision
        ,data_scale as scale
        ,data_length as dlength
        ,data_default  
        ,' ' comments 
        ,data_type_mod
        ,cols.char_used
        ,initCap(histogram) as histogram
        ,num_distinct
from    all_tab_columns cols
where 1=1
  and cols.table_name = :TABNAME
  and cols.owner = :OWNNAME
order by column_id 

Monday, January 18, 2010

I’m a Semi-Finalist in the embeddedSpark 2010 Challenge!

embeddedSparkThe results were announced on January 15th.  I never received an email, so I assumed that I didn’t make the cut.  Thought about it this morning and decided to wander over to the site and look for any announcements and low & behold there is Mr. ianlee74 on the list!  I guess it’s time to get busy!!!  More updates to come in the near future. 
Check out the list of contestants here.  You can see my entry here .