Author Topic: Electric's at WC (Read 3007 times)

Dennis Toth · « **on:** July 28, 2010, 06:23:06 PM »

Guys,

Looks like Igor Burger is 4th after two days. Is he flying the electric with A123 battery? Any other electrics at the WC.

Best, DennisT

Robertc · « **Reply #1 on:** July 28, 2010, 06:44:57 PM »

I think Igor said there are 9 electric at the world's in F2B.

Howard Rush · « **Reply #2 on:** July 28, 2010, 09:43:23 PM »

They shortchanged Igor by 60 points, which should put him in second.

Keith Renecle · « **Reply #3 on:** July 28, 2010, 11:20:28 PM »

Quote from: stuntguy13 on July 28, 2010, 06:23:06 PM

Guys,

Looks like Igor Burger is 4th after two days. Is he flying the electric with A123 battery? Any other electrics at the WC.

Best, DennisT

Igor is flying with his 6-cell A123 battery pack. The 3 Czech guys also use Igor's system, and Igor is using his own active 2/4 break type of system as well. The leading people right now is Richard Kornmeijer, and he has two models with Igor's sytem, but decided to use his old faithful ST 60 model that he is more familiar with. Anyone go a spare ST 60?? S?

John Cralley · « **Reply #4 on:** July 29, 2010, 08:45:32 AM »

Quote from: Howard Rush on July 28, 2010, 09:43:23 PM

They shortchanged Igor by 60 points, which should put him in second.

Humm, too bad.

How was he shortchanged??

Just curious.

Howard Rush · « **Reply #5 on:** July 30, 2010, 01:15:12 AM »

They recorded the score incorrectly. They fixed it today.

Igor Burger · « **Reply #6 on:** August 02, 2010, 09:51:47 AM »

I already wrote it somewhere, but it belongs here, so I will repeat the story again with more details.

I had my model from Landres 2 years ago which placed 2nd. Now I did not have that luck, a placed 6th. It is model originally for piped LA.46 built for Sebnitz 8 years ago, and later converted for electric. Weight is 1750 including 6 cells A123, Spin 66, APCE 12/6 and now I newly tested Scorpion 3020/710. But it goes very hot, I found that those hot days it was over 100deg celsia. Yes as Keith wrote, I used active control with gyroscope. I use very conservative setting, so it was not "hearable", but anyway I was comfortable especially in wind. May be that is why it placed 1st in first final round flown in wind up to 9m/s. actually biggest open problem is tendency of oscillations during regulation if system is set too strong.

3 Czech flyers had also electric, they started last year with my combination from Landres which was 6s A123, spin66, APCE 12/6. But Jiri vejmola and Rado Dobrovolny changed A123 to 6s lipos 2250 because of weight. They were new, so we do not know life time, and also they had little longer charging time (they have only one battery and charging in the model as I do with A123). They all had also active system using part of my software but converted to accelerometers by Kamil Meisl. It is simpler / cheaper solution with little less sensitivity because of too large centrifugal force and limited resolution of A/D converter. Actually biggest open problem I see is gravity which makes troubles, resulting after all corrections to acceleration at descending, especially in wind – at least that what I saw from outside, I did not fly it yet.

2 Hungarian flyers also copied combination of 6sA123 and spin 66, but using also accelerometers for active control by Kamil.

Polish flyer Kristian Borzecky has complete system from me from Landres, without active control, unfortunately he decided to put handle to ground after accident on very low quality grass. He thought that model cannot fly so tried to catch him, but unsuccessfully, so model crashed after short flight. I cannot understand how someone can put handle to the ground with running motor.

And then we had Japanese and Spanish model with classic 4 lipo setups and 13x4.5 pushers both also well flying.

Dean Pappas · « **Reply #7 on:** August 02, 2010, 08:15:39 PM »

You did well, Igor. We are pleased and proud for you.

So, do I understand correctly that some folks are using a solid-state accelerometer, with the axis of sensing aligned with the crankshaft, to control slowdown and speedup? That has an ugly instability that will show its ugly head just as the gain approaches the "desired" setting. Don't you agree?

Regards,
Dean

Igor Burger · « **Reply #8 on:** August 03, 2010, 04:51:25 AM »

Yes, it is 3-dimensional accelerometer, but for now he uses only one, span wise axle. It has influence from gravity and it makes troubles, that part suppressing gravity is not so easy to do well and they had as far as I know 3 different software versions.

Accelerometer oriented in direction of crankshaft does not make sense too much, because it can measure only aerodynamic forces in that direction (thrust, drag) and so it should control the thrust on base of measured thrust, what is not very logical … it works to some extent, but it cannot recognize flying up and flying down (accelerometer oriented in that direction does not “see” the gravity, means does not know “up” and “down”), so it completely fails in wind.

Dennis Toth · « **Reply #9 on:** August 03, 2010, 10:18:21 AM »

Dean/Igor,

With the accelerometer in line with the shaft it would control us to constant ground speed which pretty much would add power as you hit the wind and hold back going down wind and when pushed. Problem could be oscillations of on-power - off-power when the ship gets pushed by either the wind or gravity. Maybe the system could be set to a positive bias where it can add power when the ship slows but limits the shut down to a small amount or zero. Also could a dampener circuit be used to smooth out the on/off as the ship gets close to the "hold speed"?

Best, DennisT

Howard Rush · « **Reply #10 on:** August 03, 2010, 01:46:18 PM »

I presume that Mr. Borzecky refused the active control because he feared putting himself in the right half plane.

Igor Burger · « **Reply #11 on:** August 04, 2010, 01:16:46 AM »

Quote from: stuntguy13 on August 03, 2010, 10:18:21 AM

With the accelerometer in line with the shaft it would control us to constant ground speed which pretty much would add power as you hit the wind and hold back going down wind and when pushed.

Yes, it will work in level, you can add power when you see that prop brakes, but not in verticals. In Verticals you will need just oppside action, add power if prop makes thrust. Accelerometer does not "see" the gravity, it can sense on forces from wing, prop, lines etc. So you never know (from value from accelerometer) where is up and where is down.

It is not so easy, we work ~3 years and we do not have really good result yet

Dennis Toth · « **Reply #12 on:** August 04, 2010, 08:17:34 AM »

Igor,

I'm not sure I understand how the accelerometer is working. I thought it provides a variable resistance circuit that could act like a trim pot that sense change of state. How are you adding the accelerometer to the control loop? It seems that the accelerometer would sense the slow down of the plane as the nose goes up and the ground speed slows in the climb and would sense the speed up as the nose goes down and the ship increase speed? Wouldn't you be able to have it trim the set rpm setting say 10% rpm on deceleration and reduce it say 5% when it accelerates? This would be a true 4-2-4 simulation.

Best, DennisT

Dean Pappas · « **Reply #13 on:** August 04, 2010, 10:10:32 AM »

Hi Dennis,
It's all in the "HOW" and nobody has nailed it yet.
Igor and Erik Janssen have had maybe the best success, and even they will say that there are conditions under which the yaw rate and/or accelleration feedback hurt rather than help.
And yet it moves!

Howard's inside joke about certain eastern Europeans buying tickets for the wrong seats in the plane just proves that we are deep into techno-geek territory.

Regards to All,
Dean

Igor Burger · « **Reply #14 on:** August 04, 2010, 11:14:22 PM »

Accelerometer proviveds voltage depending on mechanical forces applied to its body (acceleration regarding its weight). Means accelerometer oriented in direction of its shaft, will sense prop thrust, or model drag. It means if you fly level at constant speed, you will see "0" on its output. If you hit the wind and that wind will slow the model, you will see "-something" value and if you fly down the wind, you will see "+ something". I think that is what you wrote before. So yes, you can use it to maintain constant speed in level flight. Means if you see "-" on output, you have to ADD power.

But accelerometer does not see the gravity. It can see only force acting against the gravity if you support accelerometer by something acting against the free fall. But nothing hold the model in air, beside aerodynamilac forces. And now: if the prop pulls model up, you will see "+" value on output and you have to ADD power. It is just opposite I wrore before.

So that is what I am trying to say, it is not so easy, there are many ways to do that and everything is question of good software and good software is question of long term debugging

Howard Rush · « **Reply #15 on:** August 05, 2010, 05:22:12 PM »

It's also a question of understanding the physics. There is a small group of people who deal with this stuff and the arcane mathematics that go with it for a living. You folks who think that electronics is the salvation of stunt might think differently when those guys take over, particularly if you let them augment the pitch axis. Don't say I didn't warn you.

Dean Pappas · « **Reply #16 on:** August 05, 2010, 09:29:43 PM »

Well Howard,
From what I gather pitch is safe: the FAI Technical Sub-Committee is being diligent on the matter.
AMA will eventually do what is sensible, but I genuinely think there is no rush.
The first guy who shows up with a fly-by wire elevator in his Toyota Stunter is gonna laughed at, then lynched.
After that, we'll all take turns evaluating the dear departed's efforts, and we will discuss our analyses over steak and beer.

Arcane mathematics ... yeah, that's the word I was thinking of!
later,
Dean

John Witt · « **Reply #17 on:** August 06, 2010, 07:51:21 AM »

Quote from: Howard Rush on August 03, 2010, 01:46:18 PM

I presume that Mr. Borzecky refused the active control because he feared putting himself in the right half plane.

Howard,

Would that be the s-plane?

John Witt

Howard Rush · « **Reply #18 on:** August 06, 2010, 12:38:25 PM »

Quote from: Dean Pappas on August 05, 2010, 09:29:43 PM

I genuinely think there is no rush.

No Rush? Don't count on it.

(edited to clarify joke)

Larry Cunningham · « **Reply #19 on:** August 09, 2010, 09:07:54 AM »

I'm wondering about the accelerometers - long ago Analog Devices had an ADXL series, with 1 or 2 axes, and you could use two of these chips in a perpendicular arrangement fit into about a 1" cube. You applied +5V (nominal) and these devices could give you three modulated outputs. Eventually they became fairly inexpensive and you could order them from places like Digikey. I think they even sold a prepackaged 3-axis version in a cube.

I visualized a magic box which would accept all three axes of force data and emit a motor speed control value. Something like three 5-bit A/D converters creating a 15-bit value to access a 32K byte "ROM" lookup table. Of course, you could compute externally and then load virtually anything into the lookup table.

Sounds good, but 3-axis acceleration data is not quite adequate - as Igor mentioned, there's the 1G gravity component, which gyres about as the model navigates on the flight hemisphere. To accurately factor it out, you need to "know" the orientation of your model.

Also, does gravity need to factor out, at least completely? After all, it IS affecting what the model needs to fly against in its pattern.

Fly-by-wire, I don't really see a problem, either in implementation or why competitors would need to be lynched for using it - it's already permitted under rules we've had forever.

And let me correct you guys, that's not the right half of the S-Plane here, it's the right half of the S-Sphere.. ;->

Many things are workable in theory. In practice, few actually are. I personally think the baby gyros are the slickest idea for modulating motor speed. In the past, I had simply assumed they were too exotic or expensive, and I was ignorant of how they were configured for helicopters.

Anyway, it's all interesting stuff, isn't it?

L.

"Slight not what's near, when aiming at what's far." -Euripides

Igor Burger · « **Reply #20 on:** August 09, 2010, 02:50:31 PM »

Larry, actually we have both - 3d accelerometers and also gyros available as small chips, so it is TECHNICALLY very easy include them in small board together with timer.

But you are right, actual small heli gyros allow direct hooking with timer and ESC without any line of code. It will work somehow, unfortunately there are many troubles, for example throttle can run out of allowed values, it can stop motor etc. Additionally heli gyros has some "too clever" tricks included in the software which are may be good for helis, but not usefull for us, for example different sensivity (gain) at different frequencies (stronger action for quicker movement).

So After all it is all question of software. We can have whatever we want - also full movement sensing with 3 gyros and 6 accelerometers, but then comes that proper unanswered question ... and NOW?

... we still do not know what is that proper motor run which we want, and we do not know how to get it by our limited software capacity (processorwise and also programmerwise

).

I tried so far 3 different sensing systems (power sensor, gyro, accelerometer) with lot of different handling and I still do not know how to get that what I do not know what is it

John Witt · « **Reply #21 on:** August 09, 2010, 02:51:02 PM »

Larry, I had enough trouble with just two axes...

John Witt

Larry Cunningham · « **Reply #22 on:** August 09, 2010, 06:33:00 PM »

Igor: nothing has changed since I retired from electronic engineering - everything is STILL only a "SMOP" - a Small Matter Of Programming".. ;->

And I'll bet most EEs are still doing just what I did - learn just enough programming to write a small program which proves to the programmer that MY hardware IS working! Heheheh.

I know that the REAL killer gyros for aeronautics and space are made with loops of fiber optics - solid state, per se.

And the little solid state ADXL devices actually have tiny silicon structures which physically distort and alter capacitance to modulate a VCO. I'm assuming some similar ideas are applied to make the tiny solid state gyros.

It's a wonderful world, I'm not keeping up with it very well. But aren't you worried about nanobots graying out the entire planet? Probably you guys will poke a black hole with the Large Hadron Collider first anyway. ;-)

L.

"There is no sadder sight than a young pessimist." -Mark Twain

Igor Burger · « **Reply #23 on:** August 10, 2010, 01:27:45 AM »

It is boring if everything works well, we must implement some trouble making points to make some fun and place where we can play with something.

Bill Little · « **Reply #24 on:** August 10, 2010, 04:10:02 AM »

Hi Igor,
Congratulations! A great job at the WC's.

Now, this going to sound dumb........ What are the actual pieces you use in your set up? Motor, battery, ESC,

? I am not asking for brand names, just what parts make it up including necessary computer items.

Thanks!

Igor Burger · « **Reply #25 on:** August 10, 2010, 04:45:10 AM »

Hi Bill,

Basic setup for those inspired by my model from 2008 and 2009 (means my, Czech and Hungarian):

Battery 6cell of A123
ESC Jeti Spin 66 with governor
Timer my, programmed by Jeti Box as well as used for programming of Jeti spin - it speaks about RPM instead of imaginary power %
Motor AXI 2826/13
Prop APC 12x6 tractor

So as you can see this set does not need anything with computer, USB link ETC. Advantage is programming by simple 4 button box with 2 line display.

Another thing you will need is charger and some source of electricity. I have two chargers, one is FMA charger with balancer which I use for maintaining, balancing measuring. And second is Polish Pulsar without balancer, which I use on field for quick charging. Both have 300W power (charging 6 cells by 10A) and they can charge battery in 15-18minutes. You can easily live with only one, but I am better if I have everything twice or backuped.

For field power supply I use LIFE 12V battery used as replace for car battery. They have 20, 40, 80 Ah, I use that smallest, which fits to my field box and I can charge from it 4 times. Means I can do 5 flights during one session. That battery is rechargeable in 1 hour. As a backup I used battery made from old A123 cells which also makes 12V battery (4 life cells make battery chargeable and dischargeable like usual lead acid car battery)

And now actual modifications:

I used active regulator with gyro. Beside one windy day set very conservative, almost flat (almost constant rpm). And I tested lighter motor Scorpion 3020/780. But I will need to go back to AXI, because I found that the motor is very hot after several flights in hot air, I measured 100deg cesia (boiling point of water) while 15g heavier AXI goes to only 60-70 deg.

Czech guys replaced A123 cell by lipo 2250mAh before WC, because of weight. It is lighter almost 100g, but we do not know life time yet (A123 typically last 400-600 flights) and another disadvantage is longer charging. I was able to fly every 3rd flight, and it IS advantage in such stressy situation.

Czech and Hungarian guys used active regulators with accelerometer from Kamil Meisl. Also programmed by Jeti Box (he uses part of my software). Also set very flat, because it is not fully debugged yet.

Polish guy has standard setup from me, it flew well, but he crashed and flew borrowed Yatsenko model.

I am not sure with other models from Japan and Spain, but as far as I know they used classic 4 cell lipo setups. Model from Spain had AXI motor and CC ESC.

Bill Little · « **Reply #26 on:** August 10, 2010, 04:50:56 AM »

Thanks Igor!

I like the *no need for computer on the field*

Bill

Alberto Solera · « **Reply #27 on:** August 13, 2010, 12:33:31 PM »

Hello

This is the model from Spain (mine

)
http://stunthanger.com/smf/index.php?topic=17710.0

I am anxious to learn more and start playing with programs, and I really appreciate your words, thanks.

Larry Cunningham · « **Reply #28 on:** August 13, 2010, 03:45:57 PM »

Lovely model, Alberto. It was lovely "in the bones" as well.

Best regards,

L.

"When cheese gets its picture taken, what does it say?" -Stephen Wright