Beta Details - dmgoedde

[dmgoedde]

This is finally the split-off thread from the 3-axis low cost UAV thread that describes Beta testing details. I plan to use this to handle details of the "frozen" list of beta testers of the "AttoPilot", non-disclose form, and user manual plus Q&A forum.

First off, as promised today I have pics of the first beta to be sent out. Another last-minute addition is provision for lithium coin cell for much faster GPS lock. See the last photo.

[dmgoedde]

Thank you for overwhelming response! I was hoping to 4 or 5 bete testers, the tally is 19!

If I missed you and you requested today or before, please contact me so I can add you to this list, and sorry that I left you off the list shown below.

Icebear was the first requestor - and I hope he will gain confidence in the AttoPilot for a "Super lighthouse mission" - perhaps a sweeping tour of all the lighthouses in his area. So, the first Beta (pictured above) will go out to Icebear.

List of beta testers:

Icebear (requested Aug 19, 2007, 12:04 AM)
helitron
docphi
ios
rcdoma
Badllarma
dalbert02
Vince53
Workshop
Kilrah
Tom Harper
Tahoe Jmfc
tazdevil
dentompie
Rockeyes
LoopForEver
rimb05
UAVeez
DHDSP

[workshop]

Icebear is first at a bunch of things... He must get up early. :D

BTW... jparisse is now Workshop ('cause I'm done traveling and home at the workshop!)

[_helitron_]

Great, absolutely the appropriate selection is the 1st beta for Icebear :) !

Cheers,

//Erwin

[dmgoedde]

Icebear is first at a bunch of things... He must get up early. :D

BTW... jparisse is now Workshop ('cause I'm done traveling and home at the workshop!)

Thx! I Edited the list

[icebear]

Wow Dean - I've been away for a few days and I amazed by the speed you are keeping up!
I am honoured to be on your list and to test the first beta unit!:D
Puts some pressure on me to test it thoroughly...!

Looking forward to more good news.

/Icebear

PS. Jeff is right, I flew my Viking UAV with a new motor this morning at 01.00 AM EST - but OK, that's 7 AM here :o...

[Badllarma]

One thing we me on the list I assume the Instruction manual will be
"UAV for Dummies" :D
Really looking forward to giving this ago just remember the instructions need to be idiot proof :) Other wise I'll be lost. Documentation is always a big thing BTW in other items I've beta tested.

[Tahoe Jmfc]

I love the amount of channels, I need the 6th channel to control the light system on my UAV. Great work!

[dmgoedde]

I love the amount of channels, I need the 6th channel to control the light system on my UAV. Great work!

Well, it is 6 inputs from the Rx, but only 5 out from AttoPilot (beta version). My reasoning for having 1 more input than output is that one of the inputs is just an aux channel that you use from Tx to control Auto vs Manul flight. The other 5 are intended that the autopilot "steps in" between the Rx and the servos to replace the human pilot. I have since then changed that vision, and added flexibility through settings.txt on SD card for user-configurable assignment of the inputs and outputs... for more flexibility.

I actually do have some extra unused IOs on micro, could use them to drive two more servos, or listen to 2 more Rx channels, or 1 more of each.

So these are the things I'm relying on the beta testers for feedback. I will aggressively fix problems and make enhancements then jump on to market ASAP with a well-tuned product as reasonable price.

[Tahoe Jmfc]

I really want to eventually be able to control a few switches through a ground station so that users can interact with the plane, giving them the ability to turn on and off lasers, lighting, and possibly move a camera around that will be dispayed on a large screen for everyone to see. But I know this is a ways off. I was thinking of just using a separate reciever on a different channel to do this but its just more weight.

[Tahoe Jmfc]

If there is a way when creating the two way interface in the future to add features such as sensors to monitor temps, volts of 2 different battery systems, rx/bec current, GPS to PC and PC to GPS and Video in/ out. Here is a product I just ran into coming out that has alot of great great features built in, its called OSD (on screen display) which displays text over a video stream with the sensor data.

http://www.rangevideo.com/index.php/aerial_video
scroll down to the bottom of the page to see a mock up of their product

[_helitron_]

Here the development thread of the mentioned OSD

http://www.rc-cam.com/forum/index.php?showtopic=1915&st=0

and here of another nice OSD comming in the near future

http://www.rcgroups.com/forums/showthread.php?t=682985

and here in one of my posts a video with FredericG's (his nick here on RCGroups) OSD in action that I use

http://www.rcgroups.com/forums/showpost.php?p=8030421&postcount=55

and here another good one (no two :D )

http://www.dpcav.com/xcart/home.php?cat=24
http://www.dpcav.com/xcart/home.php?cat=23

//Erwin

[dmgoedde]

I really want to eventually be able to control a few switches through a ground station so that users can interact with the plane, giving them the ability to turn on and off lasers, lighting, and possibly move a camera around that will be dispayed on a large screen for everyone to see. But I know this is a ways off. I was thinking of just using a separate reciever on a different channel to do this but its just more weight.

Future, but again even the current Beta is forward-compatible to those features.

Right now I'm addressing the main problem of "simple low cost 3 axis autopilot", ESPECIALLY as I percieve that the only commercially available systems are difficult/buggy and require lots of fiddling to tune the gains, PLUS aren't that cheap. This is why I have stressed 3 KEY attribute of AttoPilot: 1) manual flight training of control gains via least-squares fitting of data and write of file to SD for subsequent autonomous flight, 2) block of female pins with direct interface to Rx for less wire clutter, 3) transparent 3V/5V compatibility, 4) all things related to using standard micro SD card for waypoints and configuration files.

The beginner should be able to easily enter the world of UAV flight and at low cost. More advanced user will enjoy the same benefits and be able to focus on the missions and not the hardware.

[docphi]

I would suggest two receiver more outputs for a head-tracker. Just thinking about FPV and OSD! :D

[_helitron_]

Is it really necessary to route pan/tilt through the autopilot ?

//Erwin

[Vince53]

Ok, I'm a little confused now. What are the five outputs from the AttoPilot?

Back to triggering a camera when a waypoint is reached; from the previous thread that would be defined by a txt file on the micro SD card. Can the "on" time for that trigger also be set from that file? An example would be to reach a waypoint and have the camera triggered for 15 continuous seconds to take pictures of the area.

Thanks
Vince

[Tahoe Jmfc]

These items are all FUTURE ideas, lets all focus now on what is needed to get the simple system running and tested as Dmgoedde has mentioned, and keep these ideas on the sidelines for a bit so we can focus on the main goal, a "simple low cost 3 axis autopilot" Then maybe later we could work on the development of an advanced system for the power users once this first system is running perfect and is ready for the market.

Thanks for the links also to the various OSD's

[_helitron_]

Hi guys, wrote just a short comment to this theme in the old thread

http://www.rcgroups.com/forums/showpost.php?p=8076672&postcount=116

To trigger the camera via the autopilot was an excellent idea of course.

Cheers,

//Erwin

[docphi]

Not necessarily through the autopilot. You'll need an extra two channels on the receiver end though.



Is it really necessary to route pan/tilt through the autopilot ?

//Erwin

[_helitron_]

Exactly what I meant docphi :) , pan/tilt servos are directly connected to two spare channels of the R/C receiver without any connection to the autopilot.

//Erwin

Not necessarily through the autopilot. You'll need an extra two channels on the receiver end though.

[Rockeyes]

I suppose its about time we stepped back a bit and wait until the units arrive and start testing, but I am pondering on the unit to rx interface method. Although I can see the logic, I can also see a couple of problems that may well require the use of a link / loom.
As I understand it, four channels required normal flight control plus one for Aux giving the ability to switch from manual to auto and No6 being a spare. So if the RX sits on top of the unit thus plugging into each other would this not be restrictive to the size, shape and connection layout of the RX? Would it still be possible to access any spare channels on a larger receiver?

[UAVeez]

Right now I'm addressing the main problem of "simple low cost 3 axis autopilot.

Gentlemen... Just want to point out that this is the proper way to hijack a thread. :D

Love it!

[docphi]

LOL! Yes, of course. I was stuck on the fact that the beta unit shows a 5 channel rx attached. It's like being a kid in a candy store. I'm just jumping the gun before even getting my dirty paws on a unit. :D


Exactly what I meant docphi :) , pan/tilt servos are directly connected to two spare channels of the R/C receiver without any connection to the autopilot.

//Erwin

[_helitron_]

LOL! Yes, of course. I was stuck on the fact that the beta unit shows a 5 channel rx attached. It's like being a kid in a candy store. I'm just jumping the gun before even getting my dirty paws on a unit. :D

:D :D :D

//Erwin

[_helitron_]

As I understand it, four channels required normal flight control plus one for Aux giving the ability to switch from manual to auto and No6 being a spare. So if the RX sits on top of the unit thus plugging into each other would this not be restrictive to the size, shape and connection layout of the RX? Would it still be possible to access any spare channels on a larger receiver?

You're right Rockeyes but it's not absolute necessary to connect the Rx in that way, short adapter cables between Rx and autopilot are also possible of course to get access to spare channels.

//Erwin

[cawink]

I love the idea you have here, it looks like you really know what you are doing.

Are you accepting pre-orders now? :D


c

[dmgoedde]

I suppose its about time we stepped back a bit and wait until the units arrive and start testing, but I am pondering on the unit to rx interface method. Although I can see the logic, I can also see a couple of problems that may well require the use of a link / loom.
As I understand it, four channels required normal flight control plus one for Aux giving the ability to switch from manual to auto and No6 being a spare. So if the RX sits on top of the unit thus plugging into each other would this not be restrictive to the size, shape and connection layout of the RX? Would it still be possible to access any spare channels on a larger receiver?


Betas will have short adapter plugs provided for physical flexibility in mating to any Rx, plus you would then be able to swap channel inputs (not stuck in order 1,2,3,4,5)

Again, I'm striving to make the first hobby autopilot that is very simple to use , effective without fuss, and a neat clean installation. I pondered wether or not to just "provide" a Rx with it, or even if I wanted to integrate a small 4ch Rx in the package, like the Berg stamp4.

right now for betas I have the block which you can use if it suits your Rx, but you can also employ the provided short servo extensions for flexibility.

I got you covered! :)

[dmgoedde]

You're right Rockeyes but it's not absolute necessary to connect the Rx in that way, short adapter cables between Rx and autopilot are also possible of course to get access to spare channels.

//Erwin

Helitron - those short servo extension adapters will be provided with each beta!

[Vince53]

With the rapid adoption of 2.4 gHz radio gear and the incompatability of the various systems, receiver integration would be difficult. When almost everyone was on 72 MHz that would have worked great, an off the shelf working package.

Vince

[dmgoedde]

Is it hijacking a thread if I started the thread? The though below sort-of ties into the discussion going on...

Future-ish question for all of you: how would you feel if the there was a future version of this autopilot that DID have a tiny 4 channel high sensitivity Rx literally integrated inside the package with place to insert your specific crystal?

For people wanting a simple autopilot 3 servo control flight, it would REALLY cut down on clutter...having basically what looks like a normal-sized 10 gram Rx that just happened to have GPS and IMU plug and a block of 5,6, or 7 servo outputs?

For more advanced users, they could either buy the unit that doesn't have integrated Rx, or they could use the integrated model and happen to have an additional Rx in the plane to use channels 5 and up, just that the additional Rx probably wouldn't have any servos plugged into channels 1-4?

I am not spending time on anything other than getting the betas made and firmware top-notch, but do want to put some of these ideas out there as they come into my head so the discussion can start.

[dmgoedde]

With the rapid adoption of 2.4 gHz radio gear and the incompatability of the various systems, receiver integration would be difficult. When almost everyone was on 72 MHz that would have worked great, an off the shelf working package.

Vince

Don't the 2.4 GHz units provide normal PWM 1-2ms signals to the servos? If so there should be no issue. My autopilot just reads the pulses provided by Rx that normally go straight to a servo.

I'll admit I don't yet own a 2.4 GHz system.

[dmgoedde]

Ok, I'm a little confused now. What are the five outputs from the AttoPilot?
Thanks
Vince

Per what you configure in the settings.txt on SD card, the 5 outputs are whatever you want them to be.

For a simple Rud/Elev/Throttle plane that happens to have a camera trigger with servo, you would assign 3 of the outputs to control rudd, elev, and throttle, and another output to control the camera shutter servo. Take your pick of which servo outputs do what.

BTW - there are no label stickers shown on the beta. your betas will look a bit more "professional" with logo and descriptive stickers.

[workshop]

Good Ideas....
RS232 Uplink capability (to add Maxstream or Aerocomm module) for GPS Downling and/or programming uplink or better: tilt and pan gimbal control
GPS OUT Header - to plug data into OSD
Mission Data on SD card - Pop in SD Card - Fly the plan on card - no laptop necessary to reprogram AutoPilot on ground.
Event triggering on Waypoint -


Not so Good...
Intergrated RX - too many RXs out there - you'd limit the appeal with a onboard RX
Intergrated OSD - same here... there is not one OSD that suits everybody


Jeff

[Mark Harris]

Problem with integrated RX is.. what if you want to use PCM, if so.. do you want PCM1024 or PCM that JR offer? or other PCM? too many different kinds of receiver... and then there is the "how many channels do you need" question. Some full featured UAV's may use 9 channels.. some may only use 3.

[Tahoe Jmfc]

Integration of Receiver

The integrated RX could be good for entry level, but you are going to narrow your customers, everyone is on different systems and with the new 2.4 stuff there is another problem, Im still using my really nice Hitech remote and do not plan on upgrading at all because i have 2 of these remotes with trainer cords. The 2.4g systems needs more time to take hold a bit first so that the market share of your customers are using the new frequency before integration of one or the other can take place without cutting out a large group of potential customers that do not have 2.4 yet.

Also I don't know about you but I have had 2 RXs go bad over the last year and have had to send them back to the manufacturer. Do you want to deal with this if your receivers start to fail or glitch? Although I love how integration would save space, weight and reduce clutter inside the aircraft.

Mark Harris has a great point in that there are too many competing formats of receivers out there. Choosing just one of the 4 types could limit your customer base, time will tell what type of reciever will dominate, but that will take time to see what happens to the 2.4g system and how hard and fast it takes root in the RC community.

I know that alot of guys at my shop in Auburn California have changed to the 2.4g system, but they get all there controllers at cost so the expense for the switch is not that bad for them. I will have to pay full price to switch, this is not very appealing for me because I have too many college loans right now and every dollar counts. But ultimatly i see myself moving to the new format if it is all that it is made out to be.

By the way, once you get the stickers on the unit and have a finished product i will do my photo magic and get you really nice images for your marketing. Having Pro imagery of your product really makes a difference, creating a quality perception to investors and customers.

[Tahoe Jmfc]

How many Channels

9 channels on the deluxe model might be a bit nice or complete overkill, using them to control other feature on the plane like landing gear, lights, camera, etc would be nice, but the only reason to route these feature through attopilot would be for the range issue.. Conventional controllers will not reach the range that a Wireless modem would reach and would be a reason to integrate more channels into the Attopilot so that the additional controls could be activated/deactivated from a computer groundstation with the additional range available with the wireless modem. but what are the rules in the USA about UAV flight?

[Tahoe Jmfc]

Future Advanced AttoPilot Configuation Ideas
Dont Forget we still need to focus on the first beta project, testing and completion for market, this is just ideas to throw about.
As for the Future advanced product key features, here are some ideas for integration into the Attopilot
1. Return to home if battery is getting low
2. Additional channels
3. 7 or 14 mile 2 way wireless modem integration
4. OSD or downlink of critical information to the ground station i.e. temp, volts, GPS and bearings, camera link
5. Dynamic GPS update to Attopilot
6. Software that displays maps and Sat imagery with current GPS location of plane, and planned flight path along with flight path record and display on map.
7. The one thing that is really key to making this unit appealing is the voltage display link to ground station. or if someone is on a gas system then a sensor that transmits low fuel warning back to ground station..

Everyone please take this list and start expanding and adding model numbers of parts that you may recommend, like modems, antennae’s, gimble mounts for various camera sizes, temp and voltage sensors, etc.

We should be helping develop a complete platform for Dean that includes every accessory that the customer may be interested in purchasing from Attopilot, from video cameras to still cameras, to sensors, to modems, to accessory switches, to gimble mounts and so on.

Being Beta testers and Dean being a one man show we need to help make this project become a reality for him by supplying him with as much information as possible - hope this will help get the Nerds

[_helitron_]

Helitron - those short servo extension adapters will be provided with each beta!

Great dmgoedde ! Beside to your Rx (if it's available for 35 MHz!) I've planned to test also with my Futaba PCM Rx.

Regarding the planned integrated Rx, not to forget the different frequencies, we need 35 MHz here in the EU for example.

//Erwin

[workshop]

If this is a technology development discussion, the above question throws the conversation WAY off topic...

Rules and Laws are too fluid to tie your hands this early. :rolleyes:

Jeff

[_helitron_]

Per what you configure in the settings.txt on SD card, the 5 outputs are whatever you want them to be.

For a simple Rud/Elev/Throttle plane that happens to have a camera trigger with servo, you would assign 3 of the outputs to control rudd, elev, and throttle, and another output to control the camera shutter servo. Take your pick of which servo outputs do what.

BTW - there are no label stickers shown on the beta. your betas will look a bit more "professional" with logo and descriptive stickers.

By the way, a question to the servo outputs dmgoedde, is the autopilot able to control two aileron servos ? I mean without Y-cable of course, that means two aileron channels into the autopilot from the Rx and two out from the autopilot to the aileron servos (necessary for flaperons for example).

//Erwin

[dmgoedde]

Problem with integrated RX is.. what if you want to use PCM, if so.. do you want PCM1024 or PCM that JR offer? or other PCM? too many different kinds of receiver... and then there is the "how many channels do you need" question. Some full featured UAV's may use 9 channels.. some may only use 3.

These are great points for a more advanced system - however I need to stress that the goal here is successful Beta of base-level unit featuring easy tuning of flight for Rud/Elev/Throttle planes w/ polyhedral wings. Apparently there is a lot of frustration out there that no-one yet sells a truly user-friendly system with any serious capabilities.

Let's get past beta, we can later upgrade firmware on the beta units to handle all sorts of things including IMU.

Also, everyone please realize (as has been stated before) that this beta does not use IMU, thus can't do true wing-levelling (I understand that U-NAV claims "wing levelling" but they also disclaim that there system does not employ IMU, thus indicating why people report wing-rocking and issues with wing-levelling on U-NAV system). I will gaurantee performance of the Beta on polyhedral wing trainers like the HobbyLobby Miss2, or powered glider type planes. All testers need to understand this.

[dmgoedde]

Not so Good...
Intergrated RX - too many RXs out there - you'd limit the appeal with a onboard RX
Intergrated OSD - same here... there is not one OSD that suits everybody

Jeff

I wouldn't "only" sell integrated Rx unit, but is proposed as one product for sale. Could come in really handy if you want to be the first person on your block with a 5 ounce park-flyer UAV, because in a SMD version of my autopilot I can envision <1 ounce for entire Rx+Autopilot+GPS+antenna. Again, this might be a niche product that suits a sub-group of enthusiasts, but I would never make this my only product.

I never considered an integrated OSD, I'm aiming for the most general system I can provide to work with array of people's equipment on the stated type of airframe.

[dmgoedde]

Integration of Receiver
By the way, once you get the stickers on the unit and have a finished product i will do my photo magic and get you really nice images for your marketing. Having Pro imagery of your product really makes a difference, creating a quality perception to investors and customers.

YES! I had that thought as I took a look at your website and was totally amazed. The photos of people look so real and not staged at all. Very nice.

Professional pics would be great, and we could negotaite payment by hardware trade of ca$h. But, this is another thread

[dmgoedde]

Future Advanced AttoPilot Configuation Ideas

As for the Future advanced product key features, here are some ideas for integration into the Attopilot
1. Return to home if battery is getting low
2. Additional channels
3. 7 or 14 mile 2 way wireless modem integration
4. OSD or downlink of critical information to the ground station i.e. temp, volts, GPS and bearings, camera link
5. Dynamic GPS update to Attopilot
6. Software that displays maps and Sat imagery with current GPS location of plane, and planned flight path along with flight path record and display on map.
7. The one thing that is really key to making this unit appealing is the voltage display link to ground station. or if someone is on a gas system then a sensor that transmits low fuel warning back to ground station..

Everyone please take this list and start expanding and adding model numbers of parts that you may recommend, like modems, antennae’s, gimble mounts for various camera sizes, temp and voltage sensors, etc.

We should be helping develop a complete platform for Dean that includes every accessory that the customer may be interested in purchasing from Attopilot, from video cameras to still cameras, to sensors, to modems, to accessory switches, to gimble mounts and so on.

Being Beta testers and Dean being a one man show we need to help make this project become a reality for him by supplying him with as much information as possible - hope this will help get the Nerds

Wow! You sound as enthusiastic as I am. We can focus on all of this LATER! :D For beta testing I'm really just requiring that all bugs get flushed out and fixed on basic functionality, because nothing else matters much until we have 20 beta testers confirm that all problems are hammered out. I have 100% confidence that this Beta approach is the correct approach and that when I execute on all issues found by betas, will have a ultra rock-solid platform, plus all 20 betas will have in their possesion latest firmware with only $150 invested.

[Rockeyes]

I think that a very small and simple all intergraded three or four-channel unit for a park flyer a/c has appeal. To have a very small machine that could fly a pre-programmed route would be great fun and a great way for all ages to learn. It may be an idea, depending on your market, if such a unit waypoints distance could be controlled (locked) to provide a degree of safety.

The other end is the end (and maybe somewhere in the middle) is where someone is looking for a good, reliable, simple to operate autopilot that does not cost the earth and has the potential to grow with his or her requirements. Being a relative new guy here and not having the electronic background as a lot of the users of this forum. I can categorically say this will always be a very narrow market until one does not feel you need a PhD in electronics just to get started.
Now if I could find simple and concise web source that I could have a read up, see some sample videos of the equipment and how it can all integrate. Would be marvellous. If I could access this and purchase online various components to assemble in either ready made or in kit form I would be ecstatic. That would also go for A/P, FPV and OSD’s etc. Plug and play :)

Sorry about that :)

Could the AttoPilot use PCM type receivers? I have had glance back but cannot see the answer……….Thanks

>Bill

[workshop]

"(I understand that U-NAV claims "wing levelling" but they also disclaim that there system does not employ IMU, thus indicating why people report wing-rocking and issues with wing-levelling on U-NAV system)."

The double negative is throwing me off, so I'm not sure what you meant. I have a couple of UNAV autopilots and the wing rocking is an indication of a poor setup. Once properly installed the PicoPilot is VERY adept at pitch and roll leveling.

Jeff

[Badllarma]

integrated Rx unit sound really good, anything the average bod (me :) ) can get stick in and fly has got to be a good thing (also good for sales as well I guess ;) ).

Although I'd need the extra channels to use my cameras for shooting stills be great if I had a ship with a simple video setup (camera facing forward only no downlink etc) then the simple version would be ideal.

[small_rcer]

I believe that you are better off without an integrated Rx. For EU use there are several different frequency ranges used. Even in North America, you can use 72Mhz, 2.4GHz, 27MHz, 53MHz and positive or negative shift along with AM and FM. As well the range and sophistcation of available versions is much greater than you could respond to. Just look at the Berg 4 and 7 Channel Rx. Available with side or end connectors, with or without case. 1 Brand, 2 Rx with 8 variants.

A basic assumption should be that we are talking about a user community that is already familiar with and / or already has RC gear for their jurisdiction. What the community is looking for, regardless of the jurisdiction, is the autopilot that they can connect to their existing RC radio and plane.

I would think that you should not consider a built in RX any more than you should consider shipping a complete ready to fly plane. Focus on the Attopilot which is outside the skill set of many of us, while we focus on the skill set of building and flying, which skillset we do (we hope) have.

There is a conference coming up in our city which I hope to attend to gain insite into the legal, political and market potential of UAVs. See:

www.defsecatlantic.ca

Regards

Jim H

[dmgoedde]

Could the AttoPilot use PCM type receivers? I have had glance back but cannot see the answer……….Thanks

>Bill

As long as they send pulses to servos that are 'normal'; i.e. they send a 1,000-2,000 microsecond logic pulse at about 50Hz (the 50Hz is not critical)

[Tahoe Jmfc]

I will be building a new plane for this project and you mentioned the hobby lobby miss 2. It has a pretty small wing span for what I ultimately want to do. Do you have any other suggestions for planes that could be a standard platform to test from so you have good test results? I want something in the 2 meter area to begin, moving up to a 2.5 - 3.5 meter later.

http://rnrproducts.com/airframes/gliders/xbxc.htm
this is one that is used for another uav project

here is the project
http://soaring.goosetechnologies.com/

[_helitron_]

By the way, a question to the servo outputs dmgoedde, is the autopilot able to control two aileron servos ? I mean without Y-cable of course, that means two aileron channels into the autopilot from the Rx and two out from the autopilot to the aileron servos (necessary for flaperons for example).

//Erwin

No answer for me, dm :o ?

//Erwin

[UAVeez]

I just ordered my Miss 2. Should have it in a few days. :D

[ios]

By the way, a question to the servo outputs dmgoedde, is the autopilot able to control two aileron servos ? I mean without Y-cable of course, that means two aileron channels into the autopilot from the Rx and two out from the autopilot to the aileron servos (necessary for flaperons for example).

//Erwin

I'm also interested in the ability to use the extra servo outputs for mixing independent servos. I've got several planes with this type of configuration. I use the Hitec Eclipse 7 transmitter which can be programmed to independently control ailerons servos, flaperons , elevons, and v-tail. Other useful options are Aileron-Rudder Mixing.

I'm initially planning to test the Attopilot on the pictured glider, but with Elevon mixing, I could probably fly it in my tailless delta plank (pictured) which I've almost finished.

[_helitron_]

Nice planes ios :) !

My 1st testplatform, the motorglider Cularis needs normally anymore servo outputs because she has not only two aileron servos but also two flap servos, so it's possible to mix ailerons to the flaps to get more roll agility. Due to the lack of servo outputs (need two for pan/tilt) I've in the moment both flap servos connected together with a Y-cable and use it only for the flap function.

//Erwin

[dmgoedde]

No answer for me, dm :o ?

//Erwin

Erwin - it is not part of current firmware, but that IS actually one of the reasons I have a .txt file on SD card that is hand-editable for assignment of the Rx inputs and servo outputs from AttoPilot. What is involved with what you specifically ask, is the ability to reverse one of the two aileron servos...pity to have to do that mechanically when software could do it easily.

Let's focus on that after intital beta testing flushes out any functionality bugs I may have missed, OR is this question critical to your ability to Beta test the AttoPilot?

[dmgoedde]

I'm also interested in the ability to use the extra servo outputs for mixing independent servos. I've got several planes with this type of configuration. I use the Hitec Eclipse 7 transmitter which can be programmed to independently control ailerons servos, flaperons , elevons, and v-tail. Other useful options are Aileron-Rudder Mixing.

I'm initially planning to test the Attopilot on the pictured glider, but with Elevon mixing, I could probably fly it in my tailless delta plank (pictured) which I've almost finished.

Implementing elevon mixing won't be any harder, per-se, than implementing two independent aileron outputs. You'll probably have to wait for IMU add-on and firmware re-vamp to fly that flying wing! :eek:

[dmgoedde]

I'm also interested in the ability to use the extra servo outputs for mixing independent servos. I've got several planes with this type of configuration. I use the Hitec Eclipse 7 transmitter which can be programmed to independently control ailerons servos, flaperons , elevons, and v-tail. Other useful options are Aileron-Rudder Mixing.

I'm initially planning to test the Attopilot on the pictured glider, but with Elevon mixing, I could probably fly it in my tailless delta plank (pictured) which I've almost finished.

Nice models! The motor glider is exactly what I have in mind for non-IMU AttoPilot.

[_helitron_]

Erwin - it is not part of current firmware, but that IS actually one of the reasons I have a .txt file on SD card that is hand-editable for assignment of the Rx inputs and servo outputs from AttoPilot. What is involved with what you specifically ask, is the ability to reverse one of the two aileron servos...pity to have to do that mechanically when software could do it easily.

Let's focus on that after intital beta testing flushes out any functionality bugs I may have missed, OR is this question critical to your ability to Beta test the AttoPilot?

Hmmmmm, which functions controls Attopilot now in automatic mode ? Rudder (for waypoint sequencing) and elevator (for altitude hold) only or ailerons also ? As you told in the beginning, Attopilot does no wing levelling in the moment, so it's not absolutely necessary to control ailerons, am I right ?

//Erwin

[ios]

Guys,

I’ve tested the RCAP3/ALTHold on this glider and it’s a great platform. I do realize that stating my desire to fly the beta on the plank was adventurous, but admittedly I wanted to shift the discussion from camera integration (which is a valid application), to the more fundamental issues of flight stability and control of the aircraft by the Attopilot. If through beta testing we can achieve good results in this, we can widen the available platforms we can use for our desired applications.

Thank you for your nice comments. I’m really looking forward to beta testing and providing some useful feedback.

Nick

[_helitron_]

I'll also wait for the beta then we'll see the possibilities.

//Erwin

[dmgoedde]

Hmmmmm, which functions controls Attopilot now in automatic mode ? Rudder (for waypoint sequencing) and elevator (for altitude hold) only or ailerons also ? As you told in the beginning, Attopilot does no wing levelling in the moment, so it's not absolutely necessary to control ailerons, am I right ?

//Erwin

In auto it controls: Rudder, Elev, Motor.

Rudder - waypoint sequencing

Throttle - controls altitude (which is one reason I stated the type of airframe to be used with firmware loaded when I ship them)

Elevator - Right now elev stays "locked" at last position you had it at when passing control from manual to autonomous. It is important that you get plane trimmed well for elevator before switching to auto flight. The Miss2 (being an old-timer type) flies just fine like this.

BTW, I see this elevator scheme as a limitation, which is why the Betas have ultra-sensitive barometer SCP1000 in the package. During beta testing I'll roll-out two key firmware upgrades: 1) altitude control via baro data rather than purely GPS, and baro to be cross-referenced against GPS over long timeframes, 2) use of baro in pitch control and coordinating elevator with turns, which allows more aggressive turning.

Currently you could leave ailerons unused in some fashion (unplugged servos?), though it would be necesary to ensure they stay "locked" in position close to nuetral trim, unless you wanted to control them by R/C by bypassing the AttoPilot for just ailerons.

[dmgoedde]

All - more hardware arrived today. See attached photos.

Also, I won't request payment from anyone until I finish and flight-test your actual AttoPilot, and will be showing pics of them as-finished with all the hardware. I'll be shipping them in groups, not waiting until I get all 20 finished.

Again, thanks for your interest and help everyone!
Dean

[_helitron_]

In auto it controls: Rudder, Elev, Motor.

Rudder - waypoint sequencing

Throttle - controls altitude (which is one reason I stated the type of airframe to be used with firmware loaded when I ship them)

Elevator - Right now elev stays "locked" at last position you had it at when passing control from manual to autonomous. It is important that you get plane trimmed well for elevator before switching to auto flight. The Miss2 (being an old-timer type) flies just fine like this.

BTW, I see this elevator scheme as a limitation, which is why the Betas have ultra-sensitive barometer SCP1000 in the package. During beta testing I'll roll-out two key firmware upgrades: 1) altitude control via baro data rather than purely GPS, and baro to be cross-referenced against GPS over long timeframes, 2) use of baro in pitch control and coordinating elevator with turns, which allows more aggressive turning.

Currently you could leave ailerons unused in some fashion (unplugged servos?), though it would be necesary to ensure they stay "locked" in position close to nuetral trim, unless you wanted to control them by R/C by bypassing the AttoPilot for just ailerons.

Great Dean, now it's also clear for me :D ! Thanks.

The reason why I'm so interested in the ailerons is, the Cularis motorglider and also my new motorship, the Challenger II, need two from the Rx separately controlled aileron servos. Necessary for example for differential aileron control or for flaperons (both ailerons 2 - 3 mm down for example).

Ok, that means I simply don't connect both aileron servos to Attopilot, that's all. Am I right :) ?

By the way, in the meantime I've planned to make the connections on the plane and programming in the Tx in that manner, that the IR-stabilizer AP2000i controls ailerons and elevator, when I switch from manual to automatic mode, due to the fact that both birds (Cularis, Challenger II) have very less dihedral.

Cheers,

//Erwin

[Tahoe Jmfc]

Let me say Dean, the package is coming together beautifully, very clean wire connect for RX.
I’m sure you have already thought this through but just in case here are ideas
1. shrink wrap the USB connector
2. I'm worried about any sort of interference from the receiver being installed above the Attopilot, you prolly have a answer for that.


Geez, it is a nice and small package

I graduate from school this Sunday and cant wait to get started building next week. Any thoughts on a good test platform to build?

[_helitron_]

Wow, very impressive and professional work, Dean ! And great pics too !

//Erwin

[_helitron_]

Another hint I forgot before, Dean, if you've no access to 35 MHz Rx as we need here in the EU please don't send a Rx with my package. I go then with my Futaba Rx.

Thanks,

//Erwin

[dmgoedde]

Let me say Dean, the package is coming together beautifully, very clean wire connect for RX.
I’m sure you have already thought this through but just in case here are ideas
1. shrink wrap the USB connector
2. I'm worried about any sort of interference from the receiver being installed above the Attopilot, you prolly have a answer for that.


Geez, it is a nice and small package

1) I assume you mean shrink wrap the 4 naked pins. I don't have any sockets to go around the 4 pins, and shrink wrap could be too tight. You could wrap a piece of masking tape over the pins until it is time for a firmware upgrade. That is the only thing those pins are used for, no routine access.
2) Like any other R/C installation, you'll need to do a range check. I haven't had interference problems. Are you referring to Rx getting messed up, or the attopilot? I have let planes fly at least 1500' away, then taken manual control no problem. I have never seen the AttoPilot act glitchy either. It might be different with the wide array of equipment that people will employ, but again do a range check.

[dmgoedde]

Great Dean, now it's also clear for me :D ! Thanks.

The reason why I'm so interested in the ailerons is, the Cularis motorglider and also my new motorship, the Challenger II, need two from the Rx separately controlled aileron servos. Necessary for example for differential aileron control or for flaperons (both ailerons 2 - 3 mm down for example).

Ok, that means I simply don't connect both aileron servos to Attopilot, that's all. Am I right :) ?

By the way, in the meantime I've planned to make the connections on the plane and programming in the Tx in that manner, that the IR-stabilizer AP2000i controls ailerons and elevator, when I switch from manual to automatic mode, due to the fact that both birds (Cularis, Challenger II) have very less dihedral.

Cheers,

//Erwin

Yes you are correct about aileron connection regarding the AttoPilot. I think you are right on the money with use of AP2000i for right now, and it should work.

[dmgoedde]

Another hint I forgot before, if you've no access to 35 MHz Rx as we need here in the EU please don't send a Rx with my package. I go then with my Futaba Rx.

Thanks,

//Erwin

I had no plans to include a R/C Rx with the betas. the Berg6 was just a stand-in model for that one photo.

[Tahoe Jmfc]

Just to make sure I am ready, do i need any other electronics like IR sensor or anything? Is it just RX, Attopilot with gps, speedcontroller with bec, motor, servos, battery for motor and bec

Also what software will i use to plot my course? I have a Garmin PC based Gps software with either street or topo maps, do i just need coordinates from like google earth? or a handheld gps?

[_helitron_]

Yes you are correct about aileron connection regarding the AttoPilot. I think you are right on the money with use of AP2000i for right now, and it should work.

Ok, thanks for the answer, Dean. By the way, I fly already with the AP2000i, so I've no extra money to spent :) .

Apropos, in the meantime my 2nd testbed Challenger II stands on her own feet already :D :

http://www.rcgroups.com/forums/showpost.php?p=8090240&postcount=71

//Erwin

[_helitron_]

1) I assume you mean shrink wrap the 4 naked pins. I don't have any sockets to go around the 4 pins, and shrink wrap could be too tight. You could wrap a piece of masking tape over the pins until it is time for a firmware upgrade. That is the only thing those pins are used for, no routine access.


I think he meant the complete "prop plug".

//Erwin

[_helitron_]

I had no plans to include a R/C Rx with the betas. the Berg6 was just a stand-in model for that one photo.

Ah, ok. I thought anywhere in one of your first postings in the old thread you mentioned something of $ 40.00 for a receiver or so. Maybe I'm wrong.

//Erwin

[dmgoedde]

I graduate from school this Sunday and cant wait to get started building next week. Any thoughts on a good test platform to build?

Well, how about start with the platform I have been using (and Icebear as well), the Miss2 ARF from Hobby-lobby.com. It is about $70 as an all-wood ARF, really good construction, takes about 6-8 hours to assemble (though I recommend using real pinned hinges, and re-soldering the linkages to the tail feathers, and Z-bends to the servos instead of supplied hardware)

My 2 cents

[dmgoedde]

Ah, ok. I thought anywhere in one of your first postings in the old thread you mentioned something of $ 40.00 for a receiver or so. Maybe I'm wrong.

//Erwin

Yep - that is what a Hitec 6 channel single conversion costs, that I got from BalsaPr.com. I quated the price to say that if I included it, I would not add mark-up just pass on the cost.

[dmgoedde]

Just to make sure I am ready, do i need any other electronics like IR sensor or anything? Is it just RX, Attopilot with gps, speedcontroller with bec, motor, servos, battery for motor and bec

Also what software will i use to plot my course? I have a Garmin PC based Gps software with either street or topo maps, do i just need coordinates from like google earth? or a handheld gps?

You need a normal ready-to-fly R/C airplane (plane, motor, speed control, servos, R/C Rx, batteries, Transmitter). The AttoPilot is a stand-alone plug-n-play solution, not a piece meal unit made of 4 or 5 seperat boards ganged together, so no IR or wing leveller is required as long as you fly it in a type of plane I specified (stable trainer with polyhedral wings + elevator and rudder control)

all instructions on waypoint generation to be included in instruction manual. To make waypoints you click out a path in GoogleEarth including altitude, then save that path file as a .kml file. You then use a Windows utility to open the .kml file with option to hand-edit altitude at specific waypoints and add holding pattern information, then the utility converts the .kml file into a file called "waypoint.txt" which you save on the micro-SD flash card, load into the AttoPilot, then go fly your mission by manually flying up safely away from the ground, then handing over control to autonomous.

[dmgoedde]

I think he meant the complete "prop plug".

//Erwin

You don't need to put shrink wrap o prop plug, but can if you want.

[Tahoe Jmfc]

What do you think the timeline will be for getting the beta to test?

I need to know when to put payment in the mail, unless you have a paypal account that i can pay with credit card

[_helitron_]

Yep - that is what a Hitec 6 channel single conversion costs, that I got from BalsaPr.com. I quated the price to say that if I included it, I would not add mark-up just pass on the cost.

Ok, got it, Dean :) !

//Erwin

[DHDSP]

Hi Dean,

great to see you keeping the beta-rollout speed up. Looking forward to see the first test results from the other beta testers! I ordered a Soarstik for this project - will have to see if I need an external leveler (IR/IMU) or not.

Btw. I saw someone from belgium using the thermopile sensors in his own project (think the pcb layout was from the papparazzi project). Maybe in the future you could think about implementing this solution for wing leveling. On the other hand, the IMU should be more reliable - but more expensive.

Have fun !

CU
Peter

[dmgoedde]

Hi Dean,

great to see you keeping the beta-rollout speed up. Looking forward to see the first test results from the other beta testers! I ordered a Soarstik for this project - will have to see if I need an external leveler (IR/IMU) or not.

Btw. I saw someone from belgium using the thermopile sensors in his own project (think the pcb layout was from the papparazzi project). Maybe in the future you could think about implementing this solution for wing leveling. On the other hand, the IMU should be more reliable - but more expensive.

Have fun !

CU
Peter

The thermophiles are a great idea, especially with the flexibility of that unused port (intended for IMU), which has 3 I/O pins to micro and various voltages and GND - in short I could put just about anything on the spare port provided it only need 3 IO pins. It is all "merely" (ha ha) a matter of programming. I actually have a few thermophile setups, but have never tinkered with them or even used them on the provided Futaba Rx. Back when I got them I didn't know as much about electronics, so they seemed like a black box. I may just have to crack them open and figure out how to interface them onto some sort of breakout board.

[_helitron_]

The thermophiles are a great idea, especially with the flexibility of that unused port (intended for IMU), which has 3 I/O pins to micro and various voltages and GND - in short I could put just about anything on the spare port provided it only need 3 IO pins. It is all "merely" (ha ha) a matter of programming. I actually have a few thermophile setups, but have never tinkered with them or even used them on the provided Futaba Rx. Back when I got them I didn't know as much about electronics, so they seemed like a black box. I may just have to crack them open and figure out how to interface them onto some sort of breakout board.

Hi Dean,

regarding thermopiles you can read also something in the development thread of the WPS-2 from MX

http://www.rcgroups.com/forums/showthread.php?t=682910

and a specialist for this thingies is also Mr. RC-CAM

http://www.rc-cam.com/forum/index.php

Edit: here an older development thread about the "artificial horizon MAHI" from Mr. RC-CAM where he has used the FMA IR-sensor

http://www.rc-cam.com/forum/index.php?showtopic=294&hl=MAHI

Btw. hardware for a X-Y IR-sensor is very simple

http://www.recherche.enac.fr/paparazzi/wiki/index.php/Sensors

pinning of FMA sensor you can also found there.

Cheers,

//Erwin

[dmgoedde]

What do you think the timeline will be for getting the beta to test?

I need to know when to put payment in the mail, unless you have a paypal account that i can pay with credit card

Tahoe Jmfc - I don't expect any payment until I get your unit finished, then will accept by Paypal, and ship the instant payment comes through. BTW, the $150 price I quoted previously is actually slightly below my cost of components, so I'm considering my extra cost as a "business startup expense". The end result is that I'll be really pushing the beta testers (well not in a pushy way) as to how it is working for them, status of the questionare, suggestions, problems, etc...

Timeline to get all beta units out the door is by 3rd week of September, however I'll be shipping units in small groupings of perhaps 5. I'm currently waiting shipment of more GPS units, and other related gadgets required to build all 20 betas.

[_helitron_]

Hi Dean,

short question only about the IMU you've in use already, is it generally possible to extend later our beta hardware with such an add-on IMU or is for this also a new hardware base necessary ?

//Erwin

[_helitron_]

Hi Dean,
Btw. I saw someone from belgium using the thermopile sensors in his own project (think the pcb layout was from the papparazzi project.
Have fun !

CU
Peter

Was Tom Pycke I think

http://tom.pycke.be/mav

//Erwin

[dmgoedde]

Hi Dean,

short question only about the IMU you've in use already, is it generally possible to extend later our beta hardware with such an add-on IMU or is for this also a new hardware base necessary ?

//Erwin
I'm happy to report that you will be able to use the IMU with the beta, provided firmware is updated. Forward compatibility!!!!!

[_helitron_]

I'm happy to report that you will be able to use the IMU with the beta, provided firmware is updated. Forward compatibility!!!!!

That's really great and good to hear Dean :) ! Congrats !

//Erwin

[wallaguest1]

http://www.rpi.edu/dept/tele/images/webgifts/applause.gif
congratulations,

[Connexxion]

I'm happy to report that you will be able to use the IMU with the beta, provided firmware is updated. Forward compatibility!!!!!


Does this mean one can choose any platform?

With an IMU there's no need for a very stable plane,am I right??

[clolson]

Does this mean one can choose any platform?
With an IMU there's no need for a very stable plane,am I right??

It all depends on the quality of your sensors and integration math. :-)

[dmgoedde]

Does this mean one can choose any platform?

With an IMU there's no need for a very stable plane,am I right??


When IMU is implemented, yes any platform will work. I have the math down pat, and software objects written plus initial (successful) tests done with the IMU on a non-stable plane, see:
http://www.rcgroups.com/forums/showthread.php?t=725773&page=2

photos of plane, did "semi-autonomous" flight with IMU performing roll and pitch levelling:
http://www.rcgroups.com/forums/showthread.php?t=725773&page=4

[Connexxion]

Thanks Dmgoedde!

Well,that's one feature most reasonably-priced autopilots don't have.

Most of them force you to use a plane that suits the autopilot instead of the mission or other preferences of the pilot.

Will the imu-unit be sold with the Attopilot or will it be a separate add-on unit?

This is just what I've been looking for!
An autopilot that doesn't need to be installed in a stable (and slow) plane,but one that can be installed in nearly every platform one would like to use.Now one can use a plane that suits the pilot,autopilot and the mission.

I'll certainly be one of your customers.

[wallaguest1]

but... if we use OSD too

how to see on screen the height read by the altimeter? we can't?

[dmgoedde]

This is just what I've been looking for!
An autopilot that doesn't need to be installed in a stable (and slow) plane,but one that can be installed in nearly every platform one would like to use.Now one can use a plane that suits the pilot,autopilot and the mission.

I'll certainly be one of your customers.

Thanks connexxion!

I would prefer to sell the IMU as an add-on, but need a simple and secure (for me) way to have firmware upgraded in the field by end-user. I have ideas for schemes to do firmware upgrade without need for cable/USB, but I have not tested or worked on these ideas.

The deal with the IMU, is that it will require its own processor to lessen the added burden on AttoPilot. IMU calculations are far from trivial, especially on a microcontroller. The IMU itself is actually more complex than the Rudder/Elev/Throttle AttoPilot. So, I envision it as a add-on that is comparable in price to AttoPilot itself. I DON'T expect, however, to have price of AttoPilot+IMU be twice the AttoPilot alone. I strongly desire to keep it in a price range accesible to many more people than the currently available systems.

[dmgoedde]

but... if we use OSD too

how to see on screen the height read by the altimeter? we can't?

Sorry - I don't quite understand the question, though I'll take a stab at answering. OSD is a ways out for now, perhaps a firmware upgrade with hardware add-on board for the Beta units. The post-beta porduction unit will feature an additional port besides IMU and GPS sockets. Current beta just has GPS and IMU socket, though because I don't have IMU ready to send out or sell, I could easily have people use the IMU socket to feed data to OSD.

Every parameter that the autopilot has could be fed to OSD. Altimeter is no exception.

[UAVeez]

I DON'T expect, however, to have price of AttoPilot+IMU be twice the AttoPilot alone. I strongly desire to keep it in a price range accesible to many more people than the currently available systems.


Hey Dean,

Don't allow yourself to take too much of a hit on these beta units whether or not the AttoPilot comes with an IMU. You are going through a lot of your free time designing, sourcing out parts, testing, and posting answers to all our questions here on this forum. You have a lot of time invested in this project so we shouldn't be too concerned about a few extra bucks here and there. I'm willing to pay whatever out-of-pocket expenses you had on my beta unit, because I feel I need to do my part in your efforts in bringing an affordable autopilot to the average hobbyist. Don't worry too much about the cost to us; this is a deal no matter how you look at it and it is still way less expensive then other comparable units out there.

Joe

[_helitron_]

Fully agree with Joe !

//Erwin

[Badllarma]

Fully agree with Joe !

//Erwin

Dito I'm in the UK so expect to pay more for postage regardless. Don't want you out of pocket dmgoedde. :)

[Vince53]

I also agree with Joe,
Vince

[Rockeyes]

Fully agree with Joe !

//Erwin


Ditto.........Me too


...Bill