Stay tuned.

I’m now officially an engineer! Here is a photo of diploma.

Honestly, I’m shocked at how few problems I saw with the layout. The image I used for testing is just the pads and silk screens for all the components, no traces or vias or anything like that. I’ve snapped a quick photo with my Droid for you all to look at, it’s below.

From Balloon Controller

1. Geiger Counter set up to run off 5V boost converter. This guy has an output pin that pulses whenever the Geiger tube detects a decay particle, I have that feeding an interrupt on my controller and I’m locally maintaining a count, this makes the Geiger Counter’s onboard USB/Serial and AVR stuff a bit excessive, but I didn’t know when we decided on the board it had such an output. Since I’m not splitting the workload with the Counter’s AVR I don’t need another serial interface.
2. Barometer initializes and transmits pressure and temperature data properly.
3. Humidity and External Temperature Sensor works with the ADC and has been roughly calibrated.
4. Diode used as a fixed voltage reference to self calibrate ADC reference voltage.
5. GPS reconfigured to send only the NMEA/GGA sentence, every five seconds. This is much easier to cope with, and I have the sentence successfully parsing on my controller.

That’s the big stuff!, I’ve added some more photos to the album I linked to in the other post (see link at top). Now…. sleep. I’ll post more soon. Launch is Saturday morning!

• Atmel ATmega328P microcontroller – this is the brains for the entire system, this communicates with the various sensors and forms a telemetry packet to be logged to an SD card and transmitted over 1200 bps packet radio link.
• Venus GPS with SMA Connector – this will be used to track the location of the payload while it climbs with the balloon and descends back down on the parachute, both for logging where the data was taken and for recovery. It will be used as the primary altimeter until we reach it’s effective ceiling of 11 miles.
• Antenna GPS 3V Magnetic Mount SMA – this antenna is actively powered by the GPS unit to increase the receptivity of signals, and comes with a 5 meter cord terminating in the SMA connector the GPS requires.
• MEMs Barometric Pressure Sensor – SCP1000 Breakout – this will be used to measure the air pressure during the travel up to determine altitude as a back up to the GPS. There is also a high accuracy temperature sensor integrated in to this sensor used to monitor the external air temperature.
• Humidity Sensor – HIH-4030 Breakout – this will be used to measure humidity outside the payload along with the barometer’s temperature and pressure readings to complete the trifecta of atmospheric data.
• Geiger Counter – this is my personal favorite, this SparkFun device is powered by another AVR, and counts radiation ticks. Out of the box, it draws power from USB and and sends the counts back over it. However the ISP headers are made available, and I intend to modify the SFE example code to send the data over the SPI port, accessible through those headers. I also intend to use those headers to deliver power.
• Polymer Lithium Ion Batteries – 6Ah – this monster of a battery pack will deliver power at a nominal 3.7V to all devices including a step-up regulator to provide the needed 5V for the Geiger Counter. It should also have plenty of juice left over when the pack lands so it can be found well before it runs out of power.
• Bionics TinyTrak3+ – this is a TNC (Terminal Node Controller) used to modulate serial data in to 1200 bps packet radio traffic. We’re replacing the stock firmware with DataTrak custom firmware to make the device a general purpose TNC, instead of it’s original purpose of directly operating with a GPS.

Likely we’ll be adding another component or two before we’re finished, and it should be a pretty impressive little package when it’s done. Now that you’re familiar with the gear, let’s go over what was accomplished at the work session yesterday.

First, I got a head start before everyone arrived and soldered up the headers on the GPS receiver, then I plunked it down on a breadboard with a BBUSB breakout board and the antenna and started testing. I was able to chuck the antenna out the door and get a cold fix in just over 30 seconds. Then I whipped up a perl script really fast to parse the output and give me awesome feedback. Yes, awesome, that’s a scientific term. Then everyone started arriving so we split up the work. Mark, Phil and Jonathon took turns soldering things up, including the other breakout boards and I made a sandwich.

Once I’d been suitably fed I started working with Phil and Mark on rigging a UART/SPI bridge together out of the AVR and the breadboard. For the uninitiated, a UART (Universal Asynchronous Receiver/Transmitter) is a basic serial port, the kind old PDAs use, and SPI (Serial Peripheral Interface) is a different type of serial port that is used with embedded systems. The principle difference is that SPI is synchronous, meaning that there is an included clock signal, and the data is transmitted and received one bit at a time every time the clock pulses. A UART on the contrary sends data at a known rate with no clock signal, and transmission and receiving do not have to run simultaneously. It also tends to be more error prone because of different clocks in the devices communicating drifting slightly. Computers don’t generally have SPI connections, but USB UARTs and regular UARTs are pretty common, hence why I wanted to rig this bridge. I set up the firmware so that it simply repeats every byte it receives from the UART out the SPI port, which in turn reads in a byte through SPI and repeats it back out the UART. I ran in to several bugs with this, and ended up working on it through out the night to get it fully working, but around 3 am I got it up and running and then wired in the barometer. Using the datasheet, and a button on the breadboard controlling the Chip Select line on the the barometer, I manually communicated with the barometer over a terminal program using my bridge. I was sucessfully able to initialize, configure and read data back from the barometer. Around this point, I decided it was finally time for sleep .

That was kind of long winded and techy, so I’ll also give my loyal readers a quick summary of what was accomplished:

1. GPS, Barometer, Humidity sensor and TinyTrak3 soldered up.
2. GPS tested successfully with active antenna
3. Simple parser used to format GPS data nicely
4. ATmega328P development/test environment set up on breadboard, running off USB 3.3V power. Available USB serial port and flash programming connection. This is a little messy and temporary.
5. With Mark and Phil’s help I’ve written a UART to SPI bridge program for the AVR and used it to test the barometer. I also intend to use it to interface to the Geiger counter.
6. Got the barometer working, haven’t done any in depth testing with it yet.
7. Took lots of photos of the prototyping set up and the barometer test setup

As I said, I took lots of photos from the UART/SPI bridge test setup and the barometer test setup, you can see them here:

Balloon Controller

EDIT: I noticed I had some of the terminal output from the GPS test still around, so I thought it might be interesting to show:

----------------------------
Time: 234745.213
Latitude: 4439.6364 N
Longitude: 07500.0351 W
Altitude: 166.3 meters
Satellites: 07
HDOP: 1.1
Accuracy: 3.3 meters
----------------------------

----------------------------
Time: 234746.213
Latitude: 4439.6364 N
Longitude: 07500.0351 W
Altitude: 166.3 meters
Satellites: 07
HDOP: 1.1
Accuracy: 3.3 meters
----------------------------

----------------------------
Time: 234747.213
Latitude: 4439.6364 N
Longitude: 07500.0351 W
Altitude: 166.3 meters
Satellites: 07
HDOP: 1.1
Accuracy: 3.3 meters
----------------------------

Yeah, yeah I know it’s been done before, I didn’t even do it right (the LED is hanging from test clips over an unsealed bottle of water) but I just wanted to see roughly what it looked like, I was killing time. Since then I’ve sketched out an idea for a permanent version of this, might make some nice gifts for geeky friends.

Without further ado, the photos:

Bottle Light

]]> http://www.mattkrass.com/?feed=rss2&p=934 0 http://www.mattkrass.com/?p=932 http://www.mattkrass.com/?p=932#comments Tue, 30 Mar 2010 07:02:21 +0000 Matt Krass http://www.mattkrass.com/?p=932 A friend (Tyler, KB1HCY) recently dug out his old Gumstix single board computer and expressed a desire to see if it still worked. So, also being curious, I invited him and Pat Wilbur (KC2OXM) over to try it out at my place. For the record we’re using an unknown model of Gumstix with an Audiostix II daughterboard and an Ethernet daughterboard I’ve forgotten the name of. All are visible in the pictures.

First we tried powering it off Mini-USB. Upon connecting it to Gojira by USB cable we took note of some Green LEDs on the board that implied power. Not noticing any new devices on my Linux install, we first assumed it was dead. We then decided maybe the USB interface was not sufficient to power it (at this time we were just guessing, hadn’t yet found any documentation on this older Gumstix). So we connected a small barrel plug to it and used test clips to hook it up to my HP power supply, set for 5V. The LED got a bit brighter, and a USB device showed up. It was a USB networking system. We later determined the extra power had nothing to do with it, the Gumstix is just slow to start up. Though it did lead to some funny dances of connecting and disconnecting power sources trying to recreate the condition later, not realizing the system wouldn’t be ready immediately after power cycling! D’oh.

At this point we started searching for documentation with Google, after a while we stumbled on to this procedure. It worked, after a few ps/kill sessions to make Network Manager stop trying to auto-configure the USB interface. Once we had it up and running we started fighting with vi to edit the interfaces file so that the network jack (eth0 on the Gumstix) would properly start up and execute DHCP. After several failed attempts (the USBNet link seemed to fail within a few minutes), Pat suggested using cat to ‘paste’ the new file over the old one in one single burst.

His method was to use the command:

cat > /etc/network/interfaces

Which would overwrite the file with whatever was received through the STDIN. Great idea, so I took the original file, dumped via cat to the screen, in to gedit, and made the modifications. Then I logged back in over ssh and ran that command, so the system was all ready to accept the new file. Keeping in mind that file’s original configuration is what prompts USBNet to search for DHCP and is fairly essential to that link working at all, we almost bricked it, well I almost bricked it. As soon I started cat I instinctively typed CTRL+V instead of Gnome Terminal’s standard Paste shortcut SHIFT+INSERT. Instead of pasting our corrected file in place on the tiny computer, it sent the control character ^V, overwriting the entire file with that! I started panicking and quickly killed it. Knowing how unreliable USBNet had been thus far we were all dreading it dropping it at exactly that moment. It was a tense few seconds while I retyped the command and properly pasted the new file in. Immediately after that it did fail. Talk about close. Screwing that up would have meant some tricky soldering to gain access to the serial port on the Gumstix, or at least purchasing another board. However it worked out fine and on next boot the Gumstix sucessfully configured both of it’s interfaces (Ethernet and USBNet) over DHCP and we were able to play around with it for a while. All in all a successful evening, albeit tense at times.

And of course, I took photos:

Gumstix

Component Reorganization

I’ve done some more work on my projects, nothing too crazy, and I’ve started reorganizing my components to make them more accessible. I’ve been storing everything (several grades of resistors, capacitors, diodes, ICs, transistors, sockets, crystals, LCDs, etc) in these Ace Hardware compartmentalized storage boxes. This did not work out very well for a couple reasons. For starters, I cheaped out and only got two (look at the linked price, I really am cheap!) which wasn’t really sufficient. As a result I have several values of resistor just stuffed in each compartment, same for the other components. For things like sockets and groups of the same IC this is fine, but for things as varied as basic discrete components its a nightmare. Also the bags the separate values were in came badly labeled in blue marker that was hard to read. So Phil gave me an old parts drawer he had (and possibly will have another for me soon, sweet!) and I took an inventory of all of my resistors. Turns out I have about 57 unique values, and 32 drawers available. So I got some sticker labels and one of my fine-tipped Sharpies and labeled the first 28 drawers with two values each, going from the upper left drawer towards the bottom right drawer in ascending order. Then my roommate helped me chop up an old priority mail box and start making dividers with cardboard and tape, about half the drawers are done. Also I have a 29th drawer that has just a single value. Next up I’ll be reorganizing the other components, and getting more of those compartmentalized storage boxes. The resistors are the most accessed component, so the new arrangement will help immensely, and the next set of drawers will probably be for capacitors. I’m currently on break, but when I get back to school I’ll post some photos of the new arrangement.

New Phone!

My faithful BlackBerry 8330 finally bit the dust. So today I made a run to the store and after some convincing I got my hands on a Droid for the 2-year upgrade price early. I love this phone, it has already been rooted of course!

I’m using the WordPress app for it to write this part of the post in fact. Here’s a (random) snapshot from the camera to show it off. It’s pretty nice, take a look:

Next stage for Audio Mixer

Now that I’ve more or less confirmed the design works I have two possible actions:

• Try to fix the attenuation problem without using huge capacitors
• Buy the nice parts and ‘upgrade’ the prototype.

I’m leaning towards the latter because I can see if the better amp solves the problem, and worst case it doesn’t preclude me from working on the problem anyway. Once I have a working prototype with the new parts, I’ll have a PCB made and sized to fit an enclosure. I can borrow a Dremel from a friend and make a nice box for the final mixer.

Equipment Page

Another thing I’ve started but not really worked on is a page detailing all of my equipment and my workbench. This will have high resolution photos, descriptions and where I can links/prices. Look for this soon!