Andrew Cowan
Senior Member
Any M2, X, X1 or X2 PICAXE is capable of i2Cfast (so 400kHz).
Project log: Ducted fans hovering craft for air photography
- Thread starter avner
- Start date
Well that's certainly a good basis on which to proceed.
So you're starting from scratch, by yourself, building a PICAXE based self-stabilising quadcopter (with separate PID loops for pitch roll and yaw) that flies outdoors and carries its own weight plus a valuable >1kg payload and auto landing capabilities, that can be commanded in those 3 axes by an external remote system?
I will (just) watch this thread from now on with great interest - I have zero experience of this sort of thing and only a tad more than zero knowledge of the challenges to be overcome. Good luck....
Last edited:
I'm actually building much the same thing myself - a modified quadrotor, though not for heavy lifting (and let's face it, in the world of quads lifting a 1kg camera is considered heavy lifting).
I know the PICAXE isn't powerful enough to do the PID loops for a quad all by itself. My approach is to have 4 PICAXE's, 3 to monitor sensors and perform calculations for each axis (accelerometer and gyro) and 1 master to take the data and actually control the motors (collect data via I2C).
I'm still working on finding the algorithms for a flight control unit to see if it's possible to break the code/calculations up into smaller pieces, which is necessary if I want to use multiple PICAXE's to share the math workload.
I know the PICAXE isn't powerful enough to do the PID loops for a quad all by itself. My approach is to have 4 PICAXE's, 3 to monitor sensors and perform calculations for each axis (accelerometer and gyro) and 1 master to take the data and actually control the motors (collect data via I2C).
I'm still working on finding the algorithms for a flight control unit to see if it's possible to break the code/calculations up into smaller pieces, which is necessary if I want to use multiple PICAXE's to share the math workload.
WHO's got plenty of ArduCopter parts on the way ? 
not because i give up, not because i changed my mind, and not because i think it's impossible,
BUT because i just thought for a moment, and I realized that i dont have enough experience in software to do it.
AND I PREFER spending time on doing the aerial photography instead of spending YEARS on trying to develop the software.
ANYWAYS, dont be too harsh, I'm still using the Picaxe (and lovely HM-TR) to control the camera
(at least at the beginning...)
not because i give up, not because i changed my mind, and not because i think it's impossible,
BUT because i just thought for a moment, and I realized that i dont have enough experience in software to do it.
AND I PREFER spending time on doing the aerial photography instead of spending YEARS on trying to develop the software.
ANYWAYS, dont be too harsh, I'm still using the Picaxe (and lovely HM-TR) to control the camera
(at least at the beginning...)
I would be very wary of flying a brick as a payload and using 433MHz. right in the middle of the Amateur Radio 70cm band. 25mW versus 400W and not a lot of hope when some keen type asks, "Is this frequency in use please", and receives no reply, and proceeds to communicate with his pal at great length. Wiping out your control signal in the process.
Ever had your central locking not respond in a built up area? Someone testing a transmitter or your sense of humour perhaps . . .
Typical max powers found are nearer 50W, but more than enough to cause you grief.
Ever had your central locking not respond in a built up area? Someone testing a transmitter or your sense of humour perhaps . . .
Typical max powers found are nearer 50W, but more than enough to cause you grief.
I reckon you can get this working.
Case in point, last year someone posted a very similar thread on another forum for a processor starting with P, and within a few months there were prototypes around.
Re stability, the first helicopter I ever bought took off and went sideways. It was harder than balancing a broom handle on your hand. But the new models that keep coming out, ie the ones with counter rotating blades, are inherently stable. You can set everything up for a hover, look away, eat some lunch, take a drink, and look back 10 seconds later and it will still be hovering roughly where it was before.
That is the sort of stability you want. I especially like the counter rotating chinooks, as they have 4 fans, and the newer ones have gyros as well.
So if it is stable and has some onboard smarts, like "hover and slowly descend" if there is RF interference, then it could end up more reliable than many RC planes. Especially the jet models that travel with a scale speed faster than the speed of sound. Now those ones you *do* have to keep looking at all the time!
Make it stable in hover, gyro for direction, and then use a picaxe to tweak the controls.
Maybe buy some toys and learn about flight, how to control for wind, stability vs forward speed, how to control when the controls are reversed when you are flying towards yourself.
One thing such toys are useful for is getting a feel for power to weight and flight times. Lifting 1kg is going to be very hard. I'd be suggesting a payload not more than a few tens of grams. Indeed if I were building a picaxe controller, I might be tempted to build it 'dead bug' style to save on the weight of a PCB.
Case in point, last year someone posted a very similar thread on another forum for a processor starting with P, and within a few months there were prototypes around.
Re stability, the first helicopter I ever bought took off and went sideways. It was harder than balancing a broom handle on your hand. But the new models that keep coming out, ie the ones with counter rotating blades, are inherently stable. You can set everything up for a hover, look away, eat some lunch, take a drink, and look back 10 seconds later and it will still be hovering roughly where it was before.
That is the sort of stability you want. I especially like the counter rotating chinooks, as they have 4 fans, and the newer ones have gyros as well.
So if it is stable and has some onboard smarts, like "hover and slowly descend" if there is RF interference, then it could end up more reliable than many RC planes. Especially the jet models that travel with a scale speed faster than the speed of sound. Now those ones you *do* have to keep looking at all the time!
Make it stable in hover, gyro for direction, and then use a picaxe to tweak the controls.
Maybe buy some toys and learn about flight, how to control for wind, stability vs forward speed, how to control when the controls are reversed when you are flying towards yourself.
One thing such toys are useful for is getting a feel for power to weight and flight times. Lifting 1kg is going to be very hard. I'd be suggesting a payload not more than a few tens of grams. Indeed if I were building a picaxe controller, I might be tempted to build it 'dead bug' style to save on the weight of a PCB.
Last edited:
I too would have reservations ........
I have been working on a Lost Model Alarm based on the Hope RFM42 and 28x2\20x2, its working but I have had to introduce code which detects whether the RF module has reset itself during transmit.
It may be down to layout of course, its currently breadboarded (PCBs due soon) but its erratic, sometimes its fine for hours at a time whilst at others it resets during transmit every few seconds.
Yes- sadly others have had similar experiences too,although my own trials (mostly with the TTL versions) have been pretty trouble free. I've found them ideal for general PICAXE work however, and continue to consider them worthwhile, but point out that a wide range of alternatives are now available.
Perhaps refer my 2008-9 Hope resource page => http://www.picaxe.orconhosting.net.nz/hoperf.htm. The model plane fuel mixture case study makes good reading- the plane will still be controllable even HOPE sourced data is lost. Stan
Perhaps refer my 2008-9 Hope resource page => http://www.picaxe.orconhosting.net.nz/hoperf.htm. The model plane fuel mixture case study makes good reading- the plane will still be controllable even HOPE sourced data is lost. Stan
Last edited:
Thanks for that useful info.
I stick with the RFM42 for now, it is small and cheap after all. If a PCB version proves unreliable, then maybe I will change my mind.
What alternatives could you suggest, has to be small, allow direct TX of data (so you can generate audio tones), programable power output, programmable wake up or int timer, low power sleep mode.
I stick with the RFM42 for now, it is small and cheap after all. If a PCB version proves unreliable, then maybe I will change my mind.
What alternatives could you suggest, has to be small, allow direct TX of data (so you can generate audio tones), programable power output, programmable wake up or int timer, low power sleep mode.
srnet: A question everyone asks... Factor in trade offs between cost,size,package/pinouts, Tx power,supply voltages, Rx sensitivity,current drain,frequencies, wake on signal/sleep & ESPECIALLY configuration. And you also want (RTTY?) tones!
Google the latest offerings from Shenzhen for starters - firms such as HAC, Yishi, Appcon, Sure, Dorji are continually rolling out new offerings (with many "badge engineered"). Don't forget EU sourced Radiometrix & Texas Instrument ChipCon etc.
The Chinese Appcon units shown below are quite appealing alternatives to the Hope offerings, but config. hassles have apparently arisen with some users. My trials have been OK however. Stan.
Google the latest offerings from Shenzhen for starters - firms such as HAC, Yishi, Appcon, Sure, Dorji are continually rolling out new offerings (with many "badge engineered"). Don't forget EU sourced Radiometrix & Texas Instrument ChipCon etc.
The Chinese Appcon units shown below are quite appealing alternatives to the Hope offerings, but config. hassles have apparently arisen with some users. My trials have been OK however. Stan.
Attachments
Last edited:
Ah ... it has to be small.
I dont want the size of the assembled unit to exceed a PCB of 37mm x 27mm, the size of a 420mahr 1S lipo.
The idea (not an original one of mine) is that you tranmit a sequence of tones of reducing power and increasing frequency. Thus when listening to the tones on a scanner or UHF handie if you move away from the RF beacon, the lower power higher frequncy tones dissapear. Its a simple location technique and it works.
I dont want the size of the assembled unit to exceed a PCB of 37mm x 27mm, the size of a 420mahr 1S lipo.
The idea (not an original one of mine) is that you tranmit a sequence of tones of reducing power and increasing frequency. Thus when listening to the tones on a scanner or UHF handie if you move away from the RF beacon, the lower power higher frequncy tones dissapear. Its a simple location technique and it works.
The Appcon/Sure/Dorji modules may be up to that. I've a loaned pair here (thanks Dr_A !) that you may want to evaluate. In fact I've just been reading an Instructable that takes this to another level !
Power stepping has long been used for propagation checks on the HF ham bands of course- W6WX 14.1 MHz I recall. At UHF a distance estimating woes may arise with signals bouncing all over the place and LOS view corridors (between trees etc) showing enhanced levels.
EXTRA: Andrew Hornblow has neatly enhanced my 433 MHz "sniffer" receiver with a PICAXE read RSSI. This may be worth while investigating !
Stan. (ZL2APS)
Power stepping has long been used for propagation checks on the HF ham bands of course- W6WX 14.1 MHz I recall. At UHF a distance estimating woes may arise with signals bouncing all over the place and LOS view corridors (between trees etc) showing enhanced levels.
EXTRA: Andrew Hornblow has neatly enhanced my 433 MHz "sniffer" receiver with a PICAXE read RSSI. This may be worth while investigating !
Stan. (ZL2APS)
Last edited:
Well .. I have spent a while wandering around the woods in the local park, using a simple decreasing power rising frequency beacon and a cheap UHF receiver , you can get to within 5M or so of where the beacon is. When you get close just take the antenna off.
There are plenty of fancy and/or expensive solutions to this issue, and some cheap ones you can make yourself such as the 'handi finder'. But thats all extra junk and kit you need to lug around. Some zealots have suggested that beacons transmitting GPS data are the way to go, as if you are going to lug around a PC or similar to decode the GPS data into something useful.
Anyway back on topic a bit, my experience so far with the Hope RFM42 module is that when you run it at max power 17dbM (50mW), and especially if you pulse the output, the module sometimes initiates a power on reset, clearing all the registers, not good.
I fly Radio Control Planes, hand launch gliders mainly and planes such as this for aerial photography;
http://www.electricflights.com/
When said plane is at an altiude of 1,500ft and 1,500ft out, you ought to consider what happens if you loose RC control and it flies away.
Plenty of commercial wildlife trackers can be used;
http://www.falconryelectronics.co.uk/shop/receivers-c-36.html?osCsid=b32b6a32777c532be5ae216d889610a1
But the receivers for wildlife trackers are hardly handy to carry (or cheap) so its likley they will be left at home. A small RF beacon which you can locate using a small scanner or cheap PMR radio is something you are likley to carry with you.
And if you have a PIC\PICAXE micro you can ad the functionality of a remote control for the camera, which gives feedback using morse audio from the RC plane to the pilot on the ground;
http://www.rcgroups.com/forums/showthread.php?t=1387378
http://www.electricflights.com/
When said plane is at an altiude of 1,500ft and 1,500ft out, you ought to consider what happens if you loose RC control and it flies away.
Plenty of commercial wildlife trackers can be used;
http://www.falconryelectronics.co.uk/shop/receivers-c-36.html?osCsid=b32b6a32777c532be5ae216d889610a1
But the receivers for wildlife trackers are hardly handy to carry (or cheap) so its likley they will be left at home. A small RF beacon which you can locate using a small scanner or cheap PMR radio is something you are likley to carry with you.
And if you have a PIC\PICAXE micro you can ad the functionality of a remote control for the camera, which gives feedback using morse audio from the RC plane to the pilot on the ground;
http://www.rcgroups.com/forums/showthread.php?t=1387378
I wonder if we are at the point where GPS is small enough and light enough to control a plane? Head to a pre-programmed location (the launch site), do very slow turns, then head downwards in a square shaped flight path.
I am not sure how all RC units "fail" as they go out of range, by my sailing yacht tends to send the servos right to one end of travel. So if a servo is jittering, or right at the end of travel, or maybe some other way is found to detect it is out of range, then a picaxe could take over and bring the plane back into range.
The input to the picaxe is a GPS, and the output is a range of servo signals. The algorithm is a little complex, but I don't think it would be beyond some of the newer picaxes.
Thoughts?
Its already done, albiet not with a PICAXE;
http://diydrones.com/notes/ArduPilot
Quite a few RC GPS systems with On Screen Status displays that give you (when you install a video downlink) the GPS location etc. Also there are many with Return to Home (RTH) features too. Loss of RC control is detected and the plane returns to you and circles overhead .......
Failure mode on link failure is dependant on the RC receiver, some types just go nuts as you describe, some stay where they are and others return to a 'fail safe' location, which if you set it up right can make the plane go into a gentle circles, rather than fly away in a straight line never to be seen again.
slimplynth
Senior Member
http://www.npr.org/2011/03/14/134533552/its-a-bird-its-a-plane-its-a-drone
Saw this through Hackaday... for our USA brothers... the new FAA rules to come??? Build 'em now while you still can?
Saw this through Hackaday... for our USA brothers... the new FAA rules to come??? Build 'em now while you still can?
Drones, per se, are not such an issue.
The greater issue is that these days for not very much money at all, you can remotely pilot an RC model from several miles away. The legalities of which seem to be of little concern to the advocates of the technology.
Its understandable that the 'authorities' are concerned about this ....
The greater issue is that these days for not very much money at all, you can remotely pilot an RC model from several miles away. The legalities of which seem to be of little concern to the advocates of the technology.
Its understandable that the 'authorities' are concerned about this ....
slimplynth
Senior Member
Drones to police the drones...
Here's MIT's effort on this subject:
http://groups.csail.mit.edu/rrg/index.php?n=Main.VisualOdometryForGPS-DeniedFlight
http://groups.csail.mit.edu/rrg/index.php?n=Main.VisualOdometryForGPS-DeniedFlight
Ah, so this is what you've been up to...
http://globalpublicsquare.blogs.cnn.com/2011/04/04/last-look-war-of-the-machines/
http://globalpublicsquare.blogs.cnn.com/2011/04/04/last-look-war-of-the-machines/
Sparkfun has an annual contest of autonomous (self-directed) vehicles in a spirited competition to race around the outside of their building for fun and prizes.
I went to it last year and saw model planes circle the building and land in front of the crowd based on their GPS, accelerometer and other internal and sensor info (some after flying off on their own, doing an assortment of aerobatic stunts for fun and self-calibration, then returning to the building to fly the course.) I only had to duck a couple of times for various aircraft. Not as exciting as dodging berserk computer-controlled military tanks, but exciting enough for me in my dotage. Lots of fun.
I chatted some with some of the teams and individuals in the competition. It appears that a fair bit of government and private corporation autonomous flight technology is making its way into the hobby experimenters market via the usual path, the folks who actually create it.
Micro-control experts are very busy working with new sensor and processor technologies for fun and profit, so I'd say we can expect much more "self-flying" support to come to RC flight in the very near future, and at reasonable prices. It's in the works.
I went to it last year and saw model planes circle the building and land in front of the crowd based on their GPS, accelerometer and other internal and sensor info (some after flying off on their own, doing an assortment of aerobatic stunts for fun and self-calibration, then returning to the building to fly the course.) I only had to duck a couple of times for various aircraft. Not as exciting as dodging berserk computer-controlled military tanks, but exciting enough for me in my dotage. Lots of fun.
I chatted some with some of the teams and individuals in the competition. It appears that a fair bit of government and private corporation autonomous flight technology is making its way into the hobby experimenters market via the usual path, the folks who actually create it.
Micro-control experts are very busy working with new sensor and processor technologies for fun and profit, so I'd say we can expect much more "self-flying" support to come to RC flight in the very near future, and at reasonable prices. It's in the works.
Last edited: