IMU on a Bioloid

[JonHylands]

So, I've been experimenting with this Bluetooth 6 DOF IMU that I borrowed from the guys I'm doing some work for.

http://www.sparkfun.com/commerce/product_info.php?products_id=754

Its a neat little device, although too big for MicroRaptor, but small enough to fit inside the Bioloid humanoid I'll be building for them.

http://www.bioloid.info/tiki/tiki-index.php?page=BrainEngineering+BrainBot

The IMU is pretty cool, and the results are very interesting - I wrote a little graphing package that is integrated into my GUI, so I can spawn a graph after doing a sequence of movements, to see in more detail what happened.

http://www.bioloid.info/tiki/tiki-index.php?page=MicroRaptor+Screenshots

Anyways, thought you guys would be interested in seeing some of the initial results. I'm not doing anything with the IMU data at this point other than logging it.

- Jon

[pepperm]

I love the robot raptor videos and the GUI you have written.

Are you planning on making the code available at any point?

Keep going and keep updating us. I will try and keep an eye on the TikiWiki more often.

Mark

[JonHylands]

Hi Mark,

I don't know if I will be making the code available or not. I will have to see where it goes, and how my code ends up going.

If you want to stay updated, and you have an RSS reader, you can generally stay updated by watching my blog: http://www.huv.com/blog. I personally can't stand any of the RSS readers I've found, but Google has a really nice one built into their home page, and since its my home page, I see it every day, many times a day.

- Jon

[DerekZahn]

Hey Jon, nice looking work! I considered getting that same IMU for Bing but decided to build my own instead because the bluetooth wasn't very useful for me (in fact I just finished building a second Bing IMU for obscure and silly reasons). I'm a little surprised by the magnitude of the high frequency variations in your graphs, what do you think the source of that is?

Oh, I asked them for the source code to the IMU because if I got it I was going to add some low pass filtering and maybe kalman filtering to the firmware. Here's a link to it if it will be useful to you:

http://www.sparkfun.com/datasheets/Acce ... 20v3_1.zip

[JonHylands]

Hey Derek,

Yeah, I would never buy that IMU myself - its way too big, and way too expensive.

This is one I will be building, hopefully very shortly:

http://www.bioloid.info/tiki/tiki-index.php?page=5-Axis+Bus+IMU

The really big variations in the graph are when the robot starts to fall, and I catch it and put it back in balance.

I've already got the source for that IMU, thanks

- Jon

[billyzelsnack]

If you're going through the trouble of building a daughterboard, why not just buy the gyro and accelerometer chips and just stick it on a single board.

[JonHylands]

Because the components are tiny surface-mount chips, and I'd always prefer to buy stuff like that rather than build it. I don't have much choice when it comes to the ATmega168 I will use, but at least I don't have to solder the IMU chips and support components.

Besides, it fits in a smaller horizontal footprint with stacked boards, so it fits better than having one larger board.

- Jon

[JonHylands]

So, I ordered the parts from Digikey today for my bus-IMU. After talking to the guys at the Brain Engineering Lab, we decided it would be better to have the full six axis IMU, so I am adding a third board with an ADXRS-300 on top.

So, this will be a 6-axis IMU, with a built-in Kalman filter (assuming I can get it working), that plugs into the Bioloid bus for both power and data. It will be a sensor on the bus like an AX-S1, although it won't be in as convenient a form for attaching.

It will have an on-board ATmega168 for interfacing to the Bioloid bus (at 1.0 Mbps).

Size-wise, its going to be 31 mm long x 20 mm wide x 22mm high. It will only have one bus connector, so it will have to be at the end of whatever daisy-chain you add it to.

http://www.bioloid.info/tiki/tiki-browse_image.php?imageId=112

I will be publishing the full specs on the wiki when I am done, including the source code for the ATmega168, and the full parts list for putting one together.

I still have a bunch of work to do on this, but I will hopefully have it done in a few weeks...

- Jon

[JonHylands]

So, the whole process is taking much longer than originally planned (what else is new), but progress is being made...

Last night I put together a prototype board for the 5-axis IMU board plus a sixth rate gyro:

http://www.bioloid.info/tiki/tiki-index.php?page=6-Axis+Bus+IMU

I need to build a six-pin cable, and once I have that done I can hook that board up to my ATmega168 prototype board, and start playing with this.

The hard part of this whole exercise is the software, and I have had some progress with that - I can send and receive data over the hardware UART, I can flash LEDs, I can read the A/D converters. I've got the control table laid out for both the foot pressure sensor and the IMU, and I'm starting to work on some of the utility code around that - reading and writing to/from the control table and EEPROM, etc.

- Jon