Replies to This Discussion

Permalink Reply by John_Mac on May 29, 2009 at 2:56pm

Bill
This was the link for the green board I had up

http://www.sparkfun.com/commerce/product_info.php?products_id=9038

but when I refreshed it, it was the red board...so don't know what is going on. Now can't find the green board (I swear it was there when I wrote you!)

Thanks for clarifying. If you have any contact info for the heli vibration issue, i would appreciate it and try to follow up.

Best
John

Permalink Reply by William Premerlani on May 29, 2009 at 3:29pm

John,
There was a discussion of the vibration issue for applying the LISY300AL gyros in helicopters in this thread.
Bill

Permalink Reply by Chris Anderson on June 5, 2009 at 7:20pm

David,

We'll have the Z sensors within a week. You were just unlucky enough to catch the system in an inventory gap.

• ▶ Reply

Permalink Reply by William Premerlani on June 5, 2009 at 9:00pm

David,

I am going to answer your side question first about the advantages of using ailerons versus rudder for control and navigation, and answer your other questions later...

The DCM firmware that Paul Bizard and I developed for the UAV DevBoard makes it trivial to navigate and stabilize with the ailerons, because it produces accurate and fast estimates of roll, pitch, and yaw attitude. Once you have the 9 direction cosines, it is just as easy and simple to steer with the ailerons as with the rudder. The advantage of using direction cosine control of the ailerons and elevator over rudder for navigation and stabilization is that it produces tight, level, stable, authoritative banked turns with neutral-stable planes. (It will work equally well with stable planes as well.) Several folks (Paul Bizard, Scouser, and "Sid") have rated the performance of "AileronAssist" as excellent. Sid commented that it must have artificial intelligence. (It doesn't) In one of Scouser's flights of a twin motor plane, one of the motors conked out, and the firmware very nicely compensated.

During a turn, the use of the direction-cosines plus ailerons and elevator for combined navigation and control makes the plane "turn like its on rails".

You mentioned the use of a separate rc gyro. I prefer not to require a separate gyro, since there are already three gyros on the UAV DevBoard, it should be possible to get all the rotational information you might need from them without the added size and weight of a separate gyro.

Also, there was a personal reason why I developed the AileronAssist firmware to use ailerons only...my Endurance does not have a rudder!!! (It only has ailerons and an elevator V-tail, but no rudder) So, my logic went something like this:

If an airframe does not have ailerons, it is likely to have a rudder and an elevator. For those sorts of airframes, I developed "MatrixNav" firmware, which does use the rudder for navigation.

For an airframe with ailerons, it might or might not have a controllable rudder, but in either case it is a safe bet that the ailerons can be used to turn the plane. Therefore, to cover either case with the same firmware, and to not tie up a channel for the rudder, I decided to use the ailerons instead of the rudder for steering. My next project is going to be altitude control, which will require throttle control. This will require the use of all of the PWM I/O channels of the UAV DevBoard to control my endurance as follows:

3 PWM output channels: ailerons, elevator, throttle
4 PWM input channels: ailerons, elevator, throttle, and command (mode selection)

It will work equally as well on other sorts of planes with ailerons and elevator, without being tied into the rudder.

Regarding your other questions about whether you should use the UAV DevBoard or the ArduPilot, based on your questions, the short answer is that I think that Chris Anderson's ArduPilot is better suited for what you have in mind. I will give you a longer answer tomorrow.

Best regards,
Bill Premerlani

• ▶ Reply

Permalink Reply by John_Mac on June 15, 2009 at 1:54pm

Bill

Couldn't find a good place to put this, but.....

As you know I want to use your board for both my Apprentice AC and my TREX Helli. I think I have a good idea how to modify your code for a 3 servo swash. My biggest concern are the gyros. As you know, the gyros on the Red Board are suspect for the frequencies generated by Helis. Your Green Board gyros appear to be better suited.

Is there any way to get a Green Board? If you want to PM me, feel free.

Thanks
John

• ▶ Reply

Permalink Reply by William Premerlani on June 15, 2009 at 2:18pm

John,

Good question. I thought someone would ask sooner or later. Now I feel guilty about letting down the heli pilots out there. Sorry.

I know there were a few buyers of the Green Board who wish they had waited to buy two Red Boards for the price of one Green Board. So, I wonder if there is a market for used Green Boards? The question is what price would make it worthwhile?

Anyone out there want to sell a Green UAV DevBoard to John?

-Bill

• ▶ Reply

Permalink Reply by William Premerlani on June 6, 2009 at 11:19am

Hi David,
Here are the answers to the rest of your questions. But first, let me explain why I think you probably want to use Chris Anderson's AruPilot. It sounds to me that you would like a ground station to set way points and to track your flight using GoogleEarth. Chris has that well-integrated into his system. My interests are more in the direction of helping novice pilots avoid crashing, and to make their flying easier and more enjoyable.

Now, on to your questions:

- airspeed sensor

There is no provision on my board for an airspeed sensor. IMO, with a good IMU, an airspeed sensor is not needed. My next project is going to be altitude hold using only the hardware on the UAV DevBoard, plus GPS.

- throtle control

I am going to do that next, as part of altitude hold.

- wireless telemetery

There are spare pins on the UAV DevBoard that you could use to interface with wireless telemetry. There is a spare serial port on the CPU that is available. You would have to do your own firmware and integration, I do not have any immediate plans for wireless telemetry.

- aerial photography

There are spare pins on the board that you could use to interface with a camera. I do not have any immediate plans.

Regarding future plans for this board, once I finish with altitude hold, I am going to work on making the navigation and control work upside down, sideways, etc.

This is an open source project, so I expect others will develop more firmware. As we speak, Scouser is working on way points.

Some other folks are working on VTOLs, including helis and a Harrier jump jet. I hope to give them a hand.

Best regards,
Bill

• ▶ Reply

Permalink Reply by jonny on June 6, 2009 at 1:53pm

Hi Bill
You say that the dev board controls the airframe with aileron elevator only, what if it were fitted to say a high winger ( not equipped with frize ailerons ) that was quite stable, in a turn, this kind of airframe would require rudder input to keep it properly coordinated, otherwise the airframe would want to yaw with its nose in a skywards direction, the aircraft would probably complete the turn, but it would have slowed and lost altitude considerably due to drag induced by the adverse yaw, in the worst case this could cause the airframe to stall, how does the dev board cope with this?

Kindest regards
Jonny

• ▶ Reply

Permalink Reply by William Premerlani on June 6, 2009 at 3:00pm

Jonny,

You are absolutely right, I had not considered that case. So far the two planes I have written firmware for and used for testing are a Gentle Lady with a rudder, but no ailerons, (MatrixNav firmware) and an Endurance with ailerons, but no rudder (AileronAssist firmware). Those two versions of the firmware will work quite well with a wide variety of other airframes. MatrixNav is well suited for high-dihedral and rudder, AileronAssist is well suited for a straight wing with ailerons.

I didn't mean to imply that the UAV DevBoard could not control a rudder, only that given a choice between ailerons or a rudder for steering, I prefer ailerons.

The hardware design is locked in, but there is a great deal of flexibility in the firmware, its always possible to implement more features.

What do you suggest I do in the case that you describe? The hardware has 4 PWM inputs, 3 dedicated PWM outputs, and spare pins that could be configured to provide an additional 4 PWM outputs. So the UAV DevBoard could control 7 channels, but it can listen to only 4 channels.

By the way, I am not sure what you meant when you said:

"...what if it were fitted to say a high winger ( not equipped with frize ailerons )..."

First, I think you meant to say "frise ailerons".

But I am not sure what you are saying. Are you saying that with a high winger it would be better to have frise ailerons, and then I would not need to control the rudder?

Could you please elaborate on using (or not using) frise ailerons with a high winger?

Best regards
Bill

• ▶ Reply

Permalink Reply by jonny on June 7, 2009 at 4:08am

Bill

Thank you for the spelling lesson, I did indeed want to say " frise " ailerons!
Frise ailerons were one of the earliest designs to overcome adverse yaw and are designed in such a way that the leading edge of the aileron protrudes into the airflow a slight ammount to counter the drag of the opposite wing and reduce adverse yaw, we all get around this somewhat by positioning the fulcrum of the servo arm in such a way that the up aileron travells more than the down aileron, but on a wing only equipped with a single servo this is not possible. Personaly I am not aware of a high wing design equipped with frise ailerons, I bought one of the first (mk2) UAV DevBoards available with the intention ( in the future ) of fitting it to an 86" piper cub with a very flat wing, this airframe requires constant rudder input in a turn and I was wondering how I would configure the autopilot to fly this type of aircraft
I have only been learning C programming for six months ( specificaly so I can use the UAV DevBoard ) so I am not yet in a position to make any suggestions regarding programming, this was the reason for my question

Thank you for the quick reply
Jonny

• ▶ Reply

Permalink Reply by William Premerlani on June 7, 2009 at 6:38am

Jonny and Scouser,

Here is what I suggest for making coordinated turns with the UAV DevBoard:

Combine the aileron control firmware from AileronAssist with the rudder control firmware from MatrixNav. I think the combination would do a great job with coordinated turns using ailerons and rudder, since right now the firmware works rather well with only ailerons or rudder.

You will have to reassign the channels and servo reversing switches. Right now, both programs use the same two PWM channels and reversing switches.

There are enough PWM channels for you to control rudder, elevator, and ailerons, but you will have to manually control the throttle, the same way that is presently done with both AileronAssist and MatrixNav.

If you would like, I could write the firmware for you, but I would not be able to do full flight testing, since I don't have a plane with rudder and ailerons. I could do walkarond testing. I think after that it would be easy enough for you to do the flight testing and adjust the control gains.

I am planning on developing altitude hold firmware, which ideally will use throttle and elevator, but I think it would work reasonably well with elevator only, if you keep the throttle high enough.

Best regards,
Bill

• ▶ Reply

Permalink Reply by jonny on June 7, 2009 at 11:10am

Bill

That sounds just the ticket! I would not have a hope of writing that much of the program so this would be greatfully appreciated, allso the Alt hold would be very usefull when it becomes available
I've had the board scince January so I think it is time to make a start!
I quite enjoy all the adjusting and flight testing so I am looking forward to it

Kind regards
Jonny

• ▶ Reply

• ‹ Previous
• 1
• …
• 11
• 12
• 13
• 14
• 15
• …
• 56
• Next ›
• Page