latest project

Posted by: pca

latest project - 25/03/2005 16:57

So, what's this?



pca
Posted by: andy

Re: latest project - 25/03/2005 17:04

Avionics and auto pilot ?
Posted by: Ezekiel

Re: latest project - 25/03/2005 17:12

Is the can on the vertical board a yaw rate sensor? One of the pinouts is labeled RPMx2...and it does say AUTONAV SENSOR ISSUE 1 up top.

Hmmmm.

-Zeke
Posted by: sn00p

Re: latest project - 25/03/2005 17:26

The black device at the far end looks like a pressure sensor, guess you'll be using that for altitude.

The vertical board might well have some sort of accelerometer on it, if that's the case then theres probably another one mounted on the board elsewhere. I did some work with accelerometers a while back and that's the configuration we used.

It's probably none of the above though!
Posted by: pca

Re: latest project - 25/03/2005 17:55

Damn, given away by my own documentation

Yes, it's an autopilot/telemetry/data logger widget for model aircraft. I was a little bored a few weekends ago, and decided to take all the different odds and ends of boards for my fly by wire project and make one overcomplicated unit

Specs are:

CPU board:

180MHz 200 MIPS ARM9 processor
64MB SDRAM
4MB Flash
1 low/full speed USB host port
1 low/full speed USB slave port
1 RS232 serial port
1 MiniSD card socket
onboard 3.3/1.8V linear PSU
runs from single +5v supply

Sensor board 1:

10 buffered 5v hardware PWM servo ports, standard futaba wiring
8 buffered 5v compatible servo inputs, standard futaba wiring
2 buffered 5v compatible RPM sensor inputs
2 TTL serial ports, both 3.3v/5v compatible
3.3v I2C port
5V I2C port
64KB non-volatile I2C parameter storage
differential pressure sensor (airspeed indication)
absolute pressure sensor (barometric altitude)
battery voltage monitor input
battery or motor current sensor input
onboard temperature sensor
5 uncommitted 0-5v analog inputs (external temperature sensors, etc)
3 axis accelerometer (lateral and vertical translation, from 1.2G to 18g max)
3 axis gyro (roll, pitch, yaw, from 75 deg/s to 300 deg/s max)
1200 baud modem transmitter (telemetry feed via audio channel of video tx)
optional 10/100 ethernet on secondary board

overall system dimensions 2.2 x 3.2 x 0.8 inches, weight including minisd card 42g

Admittedly, it's twice the size and weight it should be, but it's only a quick prototype. Generation 2 will be quite small.

A couple of friends are writing the autopilot software to go with it. There is a little GPS unit and a single-cell LiPoly PSU as well, all up the system currently weighs about 85g. I'm going to try to hack together some software over the weekend to at least log the altitude and G readings, so I can have a play during the good weather.

I'm curious to see how many G my camera plane pulls in a tight vertical loop, you can see the wings bend rather alarmingly so I'm thinking 3+ at least.

There's a bigger picture of the thing here , and one of the better aerial photos here.

pca
Posted by: mlord

Re: latest project - 26/03/2005 00:50

So, how much each if I wanted six of them by mid-May 2005?

Thanks
Posted by: canuckInOR

Re: latest project - 26/03/2005 03:25

Embedded programming workshop, again?
Posted by: mcomb

Re: latest project - 26/03/2005 04:41

Damn, and I thought you where finally going to release the empeg autopilot add on board. I guess I'll have to keep doing it the old fashioned way
Posted by: pca

Re: latest project - 26/03/2005 08:05

Well, I though of adding a good DAC to the thing, so it could play MP3s while flying (sort of whistling while you work), but didn't have the board space

Actually, one thing I have been thinking about is making an addon for the board that would drive a couple of three-phase brushless motors. The drive method allows you to run the coils at audio frequencies, thereby making beeps and whatnot come from the motor itself. I wondered if you could mix in audio data to the drive signals, and superimpose music on the motor output! Since my latest plane design is a twin-motor design, you could even have stereo

Maybe I should patent the idea.

pca
Posted by: SuperQ

Re: latest project - 26/03/2005 14:24

hehe. you're the type of person we need to make the empeg mk3
Posted by: JBjorgen

Re: latest project - 28/03/2005 14:36

Quote:
hehe. you're the type of person we need to make the empeg mk3


uhh...you do realize that patrick designed the mk1 and mk2 right?
Posted by: andym

Re: latest project - 28/03/2005 14:48

Quote:
Quote:
hehe. you're the type of person we need to make the empeg mk3


uhh...you do realize that patrick designed the mk1 and mk2 right?


With a bit of help from Hugo!
Posted by: TigerJimmy

Re: latest project - 29/03/2005 15:37

What gryros are you using? Have you thought about how to handle gryro drift? Most of the small and inexpensive roll-rate sensors I know about have a fairly high drift rate, which makes them suitable for short flights only. I haven't worked on one of these systems for about 10 years, though, and I'm curious if this is still an issue.

Sounds like a great project!

Jim
Posted by: pca

Re: latest project - 29/03/2005 17:38

The board is set up to use either the Murata ENC-03MA/B or ENC-03JA/B gyros (300 deg/s) or the Tokin equivalent (90 deg/s). I am also investigating using some new epson ones, and a three axis rather pricey one I can't so far get much data on.

I also have the option of the SSS solid state ring gyro, which has much lower drift rates and less noise, but they are horribly expensive compared to the piezo type. Much more accurate though. I've actually got half a dozen I was given a while ago as samples, but they're something like £250 each in small quantities which is a bit much for this current project. They're also considerably larger and heavier.

The drift rate on the murata ones is higher than is ideal but usable, and can be compensated for to a degree by cross-correlating with the accelerometers, gps, and horizon detecting autolevel system I'm using.

I just managed to get WAAS (EGNOS in europe) working on the little GPS modules my system uses, and the DOP is as low as 1.2m, evenfor altitude, which almost makes the barometric sensor redundant.

What I'd like to do is find a source of elevation data for the UK at around 10m resolution so I could upload a local height map into the autopilot before a flight, allowing it to avoid the ground without some form of active sensing, but all the high res data I can find is much too pricey. I'm just wondering whether the height data in the Microsoft flight sim UK terrain map addon is both accurate enough and extractable enough for this purpose.

pca
Posted by: g_attrill

Re: latest project - 29/03/2005 22:10

Quote:
What I'd like to do is find a source of elevation data for the UK at around 10m resolution so I could upload a local height map into the autopilot before a flight, allowing it to avoid the ground without some form of active sensing, but all the high res data I can find is much too pricey.

Have you looked at Memory Map? (www.memory-map.co.uk) The elevation data is a separate .qed file. I'm not sure of the resolution or format but it might be useable. They have sample downloads which include a qed fle.

Gareth