ADI

Hi Guys,

I'm lookin at building a DF also. I have the electronic knowhow but lack the knowhow when it comes to selecting props, motors, gearboxes, batteries .. etc....
I'm gonna build my own mosfet PWM motor controllers and use 3 Analog Devices ADXRS300 rate gyro chips and one or two ADXL202 accelerometer chips. The whole lot will be locked together with an Atmel AVR microcontroller.
So, that's how I plan to do it, but whether it will destroy itself at takeoff or not is another thing.
I might be able to handle the electronics, but I'm a dumbarse when it comes to aerodynamics.
Spaceclam ..... did you use gyros on your DF, or are you just a master of RC ? If those photos are taken with no gyros on board, then I'm blown away. Are you really only 14 ?

Cheers ADI

mdream

A super BIG hovering hello to everybody !!!

I have been building DF copy for few months (not so actively). I have desinged the electronics for
gyro feedback control (basically a punch of mixers). Actually it has only 4 quad opamps (SMD), so
the weight of electronics is pretty low. I already build a test version of the 4 channel PWM controller
(that I also designed) and it works really well. This will be included directly to the circular "motherboard",
similar shape and look to the original DF PCB construction.

So as ADI (btw, big hello ADI, your project sounds very very similar to mine), I will use propably ADXL202
accelerometer chips to provide the angle feedback to the mixer circuit. First I was designing the circuit to the
base of AVR chip, but after I realized that I really don't need more than few mixers to do the job and have almost
the same weight, I dropped that Idea.

I made the frame from aluminium, the dimensions are exact copy of DF, the weight is 70 grams.
I calculated the weight of electronic components and the weight will be somewhat 100 grams, this will
also include the receiver (which I'm going to build by myself too). I have 4 280-motors that weight 43
grams each. Now I'm in weight of less than 700 grams with the battery, but I'm trying to take it under
600 grams with few mods.

Anyway, I'm in still in process to build the "motherboad". I just need some time to finish the project.
If you guys want, I can send some pictures of the going project.

ADI

Hi mdream,

Good to see I'm not alone here.
I've got 2 X ADXL202Es but haven't had a play with them yet. Still waiting on ADXRS300 gyros to arrive. From what I've read about the accelerometers, they don't respond fast enough, but the gyros apparently do. Like I said, I haven't experimented with either yet, so I'll be interested to see how you get on just using accelerometers. Are you going to try direct drive or geared motors ?
Pictures would be good to see.

Cheers ADI

Spaceclam

ADI, yes i was flying the draganflyer with gyros onboard, but unless there is wind they don't do anything. the slight movements are far too slow, and few for the cheap gws peizos to compensate, so it jsut keeps it steady in wind, and nothing more. There was no wind on that day. The gyros only have a gain control and no control gain, so i have to set it low enough that it doesn't over correct in a reasonable amount of wind and cause the plane to "hunt" and negate control imputs. When the gain is that low the gyros don't really do anything unless the craft is vionaltly pitched or rolled. It is actually not that hard to fly. the cg is about 2 inches below the rotor disks and it works well that way. Yes, i am really 14.
Cheers

mdream

ADI,

I have heard that accelerometers might be too slow. Well, everything is still open, I've designed the circuit to have a voltage controlled feedback, so actually I can experiment with any feedback devices (different accelerometers, gyros...), though, it would be great to find the "right" one immediately at first buy, without speding a fortune. Yep, I will propably also need something like ADXRS300 (propably the best choice), my circuit also has the yaw feedback input option.

In my circuit I have comparators for each function (yaw, pitch and so on). It actually compares the stick position to the real position and I can leave the feedback devices off (gyros, accelerometers) if I want, of course, it will be very difficult to fly

And finally answer for your question, I'm using four 280 motors with my selfmade gear, after some consideration, I choosed 6:1 gearing.

Have you (general) guys ever thought about the effect of the battery ?? I think that it's something worth to experiment too. The lower you place it,
the more stability increases, though this will also decrease the effectivity of movements.

BTW, ADI, where you got those ADXL202Es ?? Here in finland it's impossible to find one !!

Spaceclam

i did. it is important, especially if you run low impedance nicads for cost reasons. those low impedance cells are amazing, (1300 mah capacity, 36 amp output possibly more) but they weight a ton. my battery weighs 10 oz.

FHHuber

OK... all this has got me thinking... (now I'm dangerous...[&:] )

How about I try making a mini version... using servos that have been gutted to act as both the motors and the ESC's? I have a few S-148's with busted cases.

I have the GWS micro RX and I think my computer radio can do the mixes.

I know how to feed the servo motor power with a different battery from the RX to give me an arming switch for the motors separate from the RX power switch. (just feed the red wires from the alternate switch. Y for the black to the separate battery and the RX.)

Only question is... will the servo motors wind up directly to full power or will they give a proportional speed. won't take long to find out... and as I said... I have the servos with bad pots, cases and/or gears that I have been too cheap to throw away. (they have good motors [:-] I can't toss those!)

ADI

Hi Guys,

mdream, ... yeah I was thinkin that maybe I don't need a microcontroller.
Your idea of using comparators may work really well. No code to write (Kalman filters) ughh. My reason for a micro was mainly for exponential curve gain control. I don't know .... may not be necessary.
What sort of PWM circuit are you using ? Voltage controlled NE555/Logic Level Mosfet arrangement, or different arrangement ?
The great thing about building your own, is that you can implement control of any parameter you like into the design. Gain/resolution/mixing is up to the designer.
I got the ADXL202Es from Analogue Devices as free samples, but have to pay for the ADXRS300s. I read on the net about a guy who claims to have successfully used 4 small mercury tilt switches for pitch and roll. He's got them each angled up at about 2 degrees. Not sure about the yaw problem though.

Spaceclam .... that's pretty cool, that you can fly your DF with no gyros when there's no wind. Geez, might be worth testing with no gyros first, just to get the hang of where the main stability problems are. Love to see some more closeup pics and video video would be great. Feel free to e-mail me all the pics you like.
I've got a couple of questions for you. How does the efficiency thing come into play with motors/gearing/prop pitch ? If I run the motor faster with direct drive and a low pitch prop, is that as efficient as using a geared motor with slower rotating but sharper pitch prop ?

Cheers ADI

ADI

PS: Have you guys got any links to suppliers of pinion and complete spur gear assemplies like the real draganflyer uses or alternative gearing arrangements ? This gearing business is going to be my biggest headache. The model shops here don't carry the greatest range of this sort of stuff.

Cheers ADI

Spaceclam

ADI,
The pics i shold be ablt to get by the end of the day. the actual thing has changed a bit, eliminating the large heavy rollbar and replacing it with a block of foam that extends above the electronics and other such changes. The video may take some time, and i don't have a place to host it anyways. if one of you would voulenteer to host it i will do the best i can to get it.
As far as the prop efficiency thing goes, larger props swinging slower are more efficient than smaller props swinging faster. that's why a helicopter with a 60 size engine can pull about 10 lbs or so. If i were to take off the gearbox and put on small direct drive props that kept the motor running at the same rpm, i would only get a few ounces of thrust. also, if you went direct drive than your rotor disk size would decrease. the higher the rotor disk loading, the heavier and less responsive it will be. just like a vmar arf, but worse (if you can immagine it). it would fly so heavy for the same weight that it would not be controllable. that's another reason helis have big slow rotors. they would be ver unresponsive with a smaller rotor. look at the production model of the draganfly. it only weighs about a pound, and it is swinging 11 inch props at about 800 rpm. it is very responsive becuase the rotor disk loading is so light, but that also accounts for it blowing away in the wind, even though the gyros keep it level. Stay with larger props swinging slower. i would be using a lower gearbox and a bigger prop if they made one to fit my needs. you kinda have to build the model around the props in this case because there are so few that would work in our application. for gears, motors and such, i would reccomend this site. www.balsapr.com they have all kinds of motors and gearboxes. you can buy them seperatly or in a set. they have the largest selection i have seen, but they are all gws. than again, i have found that their motors and gearboxes are pretty good. good enough that they haven ot failed in flight like all other gws electronics.
With that site in mind, you will have no headache finding the right gearbox ratio. it even gives you some data for different props giving you an idea as to what you are drawing, getting out in thrust etc with the prop you are running.
cheers

ADI

Hi Spaceclam,

Big thanks for the info and the link. Like I said, I'm a dumb arse with this stuff, so your help is much appreciated. Maybe I can help you with electronic info and stuff, if you need it. Anyway, I'm sure if we come up with a good electronic design for DF, I'll be more than pleased to share it with you.
Might save you lots of $$ on your next big monster DF - no more expensive vtails and ESCs etc.
Just to give you an idea .... the parts required to build a simple single channel ESC for 2 motors (ie. motor 1 speeds up while motor 2 slows down) would be US$10 to $20 maybe.
Do you have the software for creating mpeg video and do you have ftp software in order to upload to one of my websites (experimental) If you do get some video together, I'll set up a site just for it.

Cheers ADI

Spaceclam

No problem. I am here to help. As far as the single channel reversing escs go, the thing is, i need them to not only speed up while the other slows down, but i need them both to raise at the same time for the throttle. for a vtail mixer, the basic equasion is the master imput +/- the slave imput. + to one side, - to the other. so let's say i had a throttle imput of +5 (these are not actual values i am just using them as an example) and your slave was -2. one motor would speed up to 7, while the other to 3. however, when i increase throttle, it raises the threshold. so if i increased the throttle by another 5, i would have outputs of 12 and 8. i did not realise that you were in new zealand, i am not sure if balsapr ships overseas. nobody is a "dumbarse" here. i knew little to nothing when i started this project. all this i learned over the course of the year that it took to put this together (a lot of that time was spend sitting around waiting for money to appear which made for a great time to ask questions and think and drive my dad to the moon and back).
i do have software to create and edit video. i have adobe premier. as for the ftp software, what is ftp software? i will get back to you on that one.
however, if you can help, is it possible to make a gyro using standard pieces (not surface mount) and not weigh much? it would certainly help to be able to create a heading hold gyro with very high gain but very little control gain. i would need it to pick up minor movements, but not do a lot to compensate. i would of course need it to be adjustable. do you know of anything that i could do? i don't have enough channels on my receiver to use the heading hold function on my futaba gyro, and peizos are useless for yaw control. they just make it unstable. it gives the correction for too short of a time to make any difference in the yawing, but long enough to make it unstable while it is correcting.
any help would be appreciated
cheers

ADI

Hi Spaceclam,

The problem here I think is that you're adding the numbers instead of multiplying.

Let's start with the following, as in your example:

..........7
..........|
..........|
5------------5
..........|
..........|
..........3

7 = 1.40 times the power of 5
7 = 2.33 times the power of 3

If we bring the throttle up, then we have to increase the power to all 4 motors by the same multiplication factor and not by addition.
So let's say for example +5 = double the power, then we have to multiply by 2 which gives us the following:

..........14
...........|
...........|
10------------10
...........|
...........|
...........6


So, our power relationships remain the same as 5,5,7,3 example above
which means the DF should remain at same angle, but just speed up.

If we exaggerate the figures and 'add' 1000 to our 7 and 3 we get the following.
They're now running at virtually the same speed.


........1007
...........|
...........|
xx------------xx
...........|
...........|
........1003



But if we multiply our 7, 3 motor speeds by 143.857 instead of adding 1000 to each, then we get the following:
So, the power relationship between the 2 motors remains the same as the 7,3 example.

........1007
...........|
...........|
xx------------xx
...........|
...........|
.......431.571

This logic is just how I figure it to be. Perhaps I've got it all wrong and everything needs to be calculated exponentially due to non linear characteristics of prop lift verses motor power.
What figures would you expect to have ? Maybe the vtails aren't capable of doing what you want.
I don't know much about them. Whatever it is that's required though, can be implemented in the electronics design.
Your gyro question is something that I've got to get a handle on myself. I'll be starting with the discrete gyro ICs and building the control electronics around them with whatever gain/control requirements are needed. The factory DF does it with 3 piezo gyro chips only - and it flies pretty good and takes care of yaw problems too.
Does www.balsapr.com sell reverse props ? Would you use EPS-100C/BB - D (6.60:1) Power System or the higher power EPS-300C/BB - D (6.60:1) Power System
Would you use the highest 6.60:1 gearing ratio ?

PS: FTP = File Transfer Protocol. Used for uploading files to websites.

Cheers ADI

Spaceclam

as far as ftp software goes, yes we have it. give me a week to compile the video for you. thanks for offering to host it.
as far as the vtails go, i am not saying you are wrong, but i think it is addition. here's why
let's say your slave channel is aileron and your master channel is elevator. you have an elevon. if you give it +5% elevator, the elevator will move 5% of it's total potential upward travel. so let's say you have +5% on your elevator. then you give it +5 on your aileron too. in order for one to go down and one to come up, it will add another +5% to one side and subtract it from the other. so you would have one at 0, and one at +10%, which still averages out to the +5% elevator that we started with, while having an aileron differential of +10. i have been using vtails for a while now, and they work jsut great. you can't et the cheap ones, they are too coarse. the only one i found that works is the $50.00 omni mixer from www.veetail.com
during flight, to control the aircraft, you don't need to differentiate much between two motors to cause it to pitch/roll the way you want it to. i think the thrust it pretty linear between those couple hundred rpms because it don't really get any yaw and i don't loose much altitude by pitching/rolling. obviously i will loose a little bit because my thrust is no longer straight up when i do that.
i am aware that the factory df uses a peizo gyro for the yaw, but it was custom designed. as i am no electrical engineer and it would all be very heavy using standard parts (they use surface mount) i had to use off the shelf parts.
i currently am using the eps 300c-c gearbox which is 5.33:1 not 6.6:1. The reason is that the 6.6:1 is too low of a gear ratio and my controls get unresponsive and sluggish due to the low rpms, and i have the data for my props with the c gearbox. the 6.6:1 gearbox is the lowest, not the highest. that is because the motor will make 6.6 turns for every 1 turn of the spur gear. (the big one that is connected to the prop shaft)
balsapr.com does not sell reverse props. there are very few props out there and they are hard to come by. i would reccomend the 10x4.5 counter rotating set from www.toddsmodels.com click on props and scroll down to the very bottom. the props the draganflyer uses are designed for 800-1000 rpm and are very undercambered. you can buy them at www.spectrolutions.com but they are expensive and will not work at these higher rpms. they go into negative pitch and even flutter. you loose most of your thrust and it creates a lot of drag which ultimatly leads to higher amp draw
Cheers

ADI

Hi Spaceclam,

You're absolutely right - the vtails are obviously adding and subtracting. What I meant in my previous post was that multiplying is what I think should be used in a system like DF, especially when you need to apply master throttle.
I've found a store here in New Zealand that sells the GWS EPS300C series motor/gearboxes and Toddsmodels will ship the 10 X 4.5 counter-rotating props outside US. So it looks like I can get either the 5.33:1 or the lower geared 6.6:1 motor/gearboxes.
You've chosen the 5.33:1 to go with the Toddsmodels props, but because I'll be designing my electronics with as much gain/control as I like, do you think I'd be better off with the lower geared 6.6:1 ? Or is the 5.33:1 a better combination with these props, no matter how responsive the electronics might be ?

One of the websites I use is just a Fortunecity.com free web hosting service. Anyone can have a site for free. If you're interested, check out http://www.fortunecity.com/free-web-hosting.shtml
You only get 25Mbytes of free space though. Let me know when you've got some video though. I can throw it up on mine.

PS: What's the hole diameter size in the Toddsmodels props ?

Cheers ADI

Spaceclam

You are better off with the 5.33:1 gearbox because the 10x4.5 is too small for the 6.6:1 gearbox. It swings it slow and does not generate much thrust. Those toddsmodels props are designed to use the 3mm shaft that are found on all gws gearboxes. The 6.6:1 gearbox will work and you can regain some of the lsot responsiveness with electronics, but you still have less thrust becuase it is slower, which gives you less room to differentiate the prop speeds for control as your throttle needs to be higher to maintain the same amount of lift. Just a precaution, when you attatch the counter rotating props, they have a tendancy to come off, so here is the setup i use. First on goes the regular 3mm nut. that goes on first because it needs to be on the bottom. I fit a 3mm nylock nut into the cutout of the prop because the nut is thicker and fits better into the cutout. You can't jsut slip the prop down the shaft, so you have to screw them on together as one. Then, put two regular 3mm nuts on top. Put them as high as you can on the shaft without getting the tapered thread, and lock them together. Than screw the prop upward and lock it against the top nut once it is secure, than bring up the bottom nut and tighten it once again against the top nut. This way the prop is up about as far as it can to while being secure, but you should still try the nuts before every flight. Be sure to order a bunch of the todds models props becuase as you learn you are likely to break a lot of them. Also, if you have pets they are likely to break them. You may or may not have problems breaking them, but it is a good thing to be prepared. These are the only props that will work for this kind of application so if the line of props goes out it looks like you are pretty well done.
cheers

ADI

OK ..... 5.33:1 gearbox it is then.
Many thanks for your advice.

Cheers ADI

Spaceclam

no problem. i will probably be able to get the video to you within the week.
cheers

ADI

Hey Spaceclam,

I can get the Windsor brand Master Airscrew 1060p (10X6) reverse props here locally. They're just a pusher prop. Is this all these reverse props are - pusher props ? I realise 10x6 would make the EPS300C-C work harder than the 10x4.5

Cheers ADI

Spaceclam

unfortunatly, those are gas props. much heavier, cause a lot of current draw, possibly burn out the motors, have a large shaft diameter (about 8 mm i believe) and are designed for faster 2 strokes, so they will produce little to no thrust at your rpm. however, yes. a pusher is the same as a normal propeller, often reffered in contrast as a tractor prop. you can always try, but you will have quite a hard time securing it to your mm shaft, be sure to put fuses on your escs, and don't expect much.
cheers

ADI

Hi Spaceclam,

Ah OK .... Toddsmodels props it is then.

Cheers ADI

Spaceclam

Good luck. Any questions, feel free to ask.
best wishes

danvel

Here it is my circuit for the DF. You can see the 4 MOSFETS and a Motorola MC68HC11E9 microcontroller.
Now I need the motors, the gearboxes, the props and all the mechanical stuff.
SPACECLAM:
I followed your advice and ordered the 10x4,5 props from TODD'S MODELS.
Is your 300 motor rated 6V or 7,2V. I've heard there's a big diference in terms of current draw.
Another question for all:
Can I get nude piezo-gyros? (I mean the chip without all the RC interface) so they're cheaper and I can solder them directly to my circuit board and read the analog signal with the A/D inputs of the microcontroller? Which one would you recommend based on price and availability?

Spaceclam

the motors i use are compact 370 motors. they are gws motors so i assume they are rated at 6v, but i am running at 8.4, drawing about 8 amps at full throttle although i dont usually run them that high. one of the guys at my field made his own gyro so i assume that you can buy peizo chips, but from where i am not certain.

scubyfan

How do you guys who do not use gyros get the dealie to hover without external control? It seems a slight bit of wind w/o gyro help would flip it or toss it to the ground. [&:]