Sternås robots without functions

I almost just finished my minisumo, janu siers. Here is a quick video showing some of its features.

I have these past days been working with a new programmer. By combining two projects (usbtiny and avr-cdc, both based on v-usb), it should be possible to make a device which programs avrs and speak uart using only an ATtiny45/85 and a couple of resistors. The usbtiny-project is an avr-programmer which I have succesfully implemented (really easy actually, given the nice instructions in the link at the bottom!). It uses the Reset-pin as an I/O-pin so using an ISP-programmer is only possible once. Since I don’t want to build a new programmer every time I test new software and since I have access to an AVR Dragon at CRF I thought I could use the HVSP-functionality to reset the fusebits. However, there were lots of trouble. When I first tried to use it, I didn’t have the correct port the Dragon. After some googling I found that the correct port is “usb”. I first tested with the ISP-header, and it worked fine. Then when I tried the HVSP-profile avrdude sent out some wierd errorcode I’ve never seen before (unknown status 174 or status 171) and apparantly no one on the internet had seen it (at least not what google could find). It also sent out the same no matter if the chop was connected or not so I double-checked the connections and they seemed right. So I started to search and tried to upgrade avrdude (I ran 5.8 first) and found out that building libraries for OS X is hard (or, at least building Avrdude with usb support).  So I ended up installing Macports and with the simple command in the Terminal

sudo port install avrdude

avrdude 5.11 with usbsupport (and about half of the internet it felt like) was installed. But even with the new avrdude it still didn’t work. However, now avrdude actually made sense of the error code. It said something about “TARGET_BOARD_NOT_POWER”. So, I thought that I needed to “trick” the Dragon. So what I did was to solder a small jumper and connected VCC to the VTARG pin on the ISP-header and this did the trick!

So, if you have trouble using the dragon_hvsp with avrdude: hope you’ll find this blogpost, because I couldn’t find any post saying what I needed. For the connections of the AVR dragon hvsp-header, please refer to the dragon datasheet (link below). Also check the datasheet for your chip to find out which pins correspond to the HVSP-pins (it usually says under “Memory Programming” in the datasheet, pretty far down).

The terminal string I used for speaking to the ATtiny (I use Eclipse to program, so this one is just for testing that the programmer works and has connection to the chip):

avrdude -cdragon_hvsp -Pusb -pt85

Also, som reference links for various sources I used:

USBtiny programmer

V-USB homepage

AVR-CDC uart bridge

AVR dragon datasheet

Macports

For some reason, I was contacted by a person from the board of my masters program, and she wanted to write an article about the success of my robot. I answered her questions, and it resulted in an article, published both at the Chalmers Inside pages and the IT-university page. The article is in Swedish.

http://www.chalmers.se/insidan/SV/aktuellt/nyheter/artiklar/chalmerist-kom-tvaa-stor

http://www.ait.gu.se/aktuellt/nyheter/fulltext/fran-mikrorobotar-till-stora-industrirobotar.cid1031603

I got a package from Farnell today. However, the packge did not contain what I ordered, or what is displayed on the outside…

If someone is missing their AD8231, this is where they are. They probably will light up given enough voltage…

LEDs?

Here is a movie compilation from Robotchallenge in Vienna. It shows the CRF robots in action.

Sorry I haven’t updated in a couple for some time, but I’ve had lots of things to do. One change I did to the hardware of the robot was to add an ultrasonic sensor, which makes Nrvnqsr probably the only microsumo in the that has an ultrasoinc sensor. The sensor is called Pong, and was developed by Victor Påsse from CRF.

Anyway couple of weeks ago, I and couple of people from CRF went to Robotchallange in Vienna, the biggest competitions for robots in Europe. Inofficially, it was also the European Robotics Championship.

Anyway, even though my robot had never competed against another robot of the same class, it performed very well. During the whole competition it lost only three matches (one in the qualifying round, one in the intermediate round and the final). However, my robot actually won against the robot that it lost against in the final, so it might actually be the best micro sumo in Europe!

Anyway, Nrvnqsr is now more or less a finished project and a very successful one. I will probably make some small changes to it until Robotchallange next yea, but that is mostly programming.

The hardware development is soon coming to an end. I have the robot up and running right now. I need to redo the tires however, since they didn’t turn out quite as good as I hoped. But I’m pretty satisfied with how the plow turned out. I can even get it to sit stationary when the robot is not running and make it fall down by running rapidly.

Nrvnqsr is almost done now :). I have soldered the bottom board and made connections between the top and bottom board. I have also manufactured a test wheel to check the size of it (reality never coincides with the computer model). I have also decided not to have wheels going all the way into the motor-holder, since I want to have the batteries there (as seen in one of the pictures). I will instead move the motors a little further in, so the wheels will be about 8-10mm wide.

Another thing I’ve made is something to mount the plow on. It’s made from a small hinge, which I picked apart, modified almost all parts, and put back together again. The plow holder is actually soldered to the lower circuit board. Since I have a large ground plane there, it’s actually quite sturdy. Now I only have to make a plow.

If I’m lucky, the robot will be running within a couple of weeks (if Aske ever shows up so he can help med cast tires). Iäm note entirely sure about the placement of the batteries as well.

Anyway, pictures:

 This christmas I made a quick project as a present for a friend of mine. It’s a small blinking music player. It’s based on an ATtiny25, one the smallest Atmel microcontrollers (5 I/O-pins, 2kB flash, 128B RAM). The name Gulgran is origins from a typo. From the beginning, this was supposed to be in the shape of a christmas tree, and Gulgran is a typo of Julgran, which is the Swedish word for christmas tree.

The tunes are played by a simple sequencer. They are stored as a list of tones (frequencies) and a duration time. It can only play one tune at the same time, since there was no time build something more advanced (also, I thought that his sounds good enough).

More info is found in the description of the video below.

http://www.youtube.com/watch?v=bQtM1JLQft8

There has been some great progress over the past two weeks. First I somehow managed to etch a circuit board correctly (after only 4 tries and lots problems such as the etching equipment was broken, I designed parts wrong, I’m terrible at etching etc etc.) But, now I have a working and soldered circuit board.

And today I (with big help of Victor Påsse) CNC-milled out a holder for my motors. It turned out pretty well, but the last rows of the milling weren’t really that good, probably due to the cheap milling tools. Now I’m gonna drill some holes in it to make it stick. And also clean the CNC mill.

Here are a couple of pictures (here you go P1r)