Biggles' Elder Brother - Moderator Posted March 31, 2016 Author Share Posted March 31, 2016 A clubmate took a couple of photos and a short video of the hand launch from this afternoon's maiden. So, first in the afternoon sunshine.... Then, mate Colin sends her away,... BEB Quote Link to comment Share on other sites More sharing options...
john stones 1 - Moderator Posted March 31, 2016 Share Posted March 31, 2016 Looks prettier every time I see it BEB I may look into fitting a camera to a model myself this year, will have to look into the cost first though.... John Quote Link to comment Share on other sites More sharing options...
Ryan Doig Posted March 31, 2016 Share Posted March 31, 2016 That looks like the making of a great model beb. I'm liking it the more I see it 👍🏻 Quote Link to comment Share on other sites More sharing options...
Ryan Doig Posted March 31, 2016 Share Posted March 31, 2016 Posted by john stones 1 on 31/03/2016 01:09:46: Looks prettier every time I see it BEB I may look into fitting a camera to a model myself this year, will have to look into the cost first though.... John I really like my bixler 2, it is a stable and versatile plane. Good stable flight but can be infinitely modded for almost anything you want really. Quote Link to comment Share on other sites More sharing options...
cymaz Posted March 31, 2016 Share Posted March 31, 2016 Well it would be nice to see the landing as well BEB ! Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted March 31, 2016 Author Share Posted March 31, 2016 Well, I didn't take the video and that's were that one ends! But I can tell you that the landing was not dignified affair! The problem again was the undercarriage - as soon as that nose wheel touches the ground its "up and over". Unless there is a significant improvement when the grass is cut (and frankly I'm doubtful) the undercart will had to either be extensively modified or go! Having no undercarriage isn't really an issue for a working model like this - you have to remember that many of the places where we have to operate there isn't a suitble take off strip - we have to hand launch and belly land in the tall grass. This is one of the major reasons for the popularity of multi-rotors with professional UAV pilots - less demanding on take off and landing facilities. Incidentally - I forgot to mention - but she flew straight and level with the control surfaces centred and no trim input at all! So while that down-trust may look a bit odd it clearly works. I'll get an incidence meter on the wing and tailplane, it would be interesting to see how its rigged. BEB Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted March 31, 2016 Author Share Posted March 31, 2016 OK time for some "back of a fag packet" calculations about power and weight etc Considering yesterday's set up for the maiden: AUW = 2.3Kg (approx 5lb), Prop 11x5.5, Battery 4s 4250mAh. Testing that on the wattmeter tells us that the set up gives 325W (21A) at full throttle. So that's a power to weight ratio of 64W/lb. The first thing to note is that while this power to weight figure is quite low you don't get that impression flying the aircraft. It went away from a moderate hand launch very positively, climbing out well and maintained straight and level flight on just under half throttle - probably around 100W. [This figure is based on the rule of thumb that half throttle is typically 1/3rd full throttle power]. This tells us that the Skua has a very lift-efficient wing whch can probably sustain level flight on a little over 20W/lb! Now the final configuration of the aircrafy will be somewhat heavier. So we need to work that out. Replacing the 4250mAh battery with the 10,000mAh planned for the final system, and allowing 330g for: camera, auto-pilot, videoTx and cables, gives us a estimated final weight of 3Kg (6.6lb). Given that we were quite satisfied with the performance in the maiden we should attempt to maintain the same power to weight ratio of around 64W/lb, much more than this and this model would be over-powered to be honest. This gives us new target max power output of around 420W, which we should get from around 28A or so. This would give us an estimated cruise power consumption of around 140W (again about 20W/lb roughly) which would require a current of about 9A. Giving us a endurance of around 1 hour on the 10Ah battery. Not bad. So, the question is: how do we get the power output up by over 30% to 420W from its current figure of 320W? Answer, we change the prop. We need a prop that will draw more power from the system - that will mean a bigger diameter and/or pitch than our current 11x5.5. But here we hit a problem. We can't increase the diameter because the 11" prop already only just clears the transparent obsevation dome - anything bigger would hit the dome. In fact its actually uncomfortably close and it would be preferable if the prop could go down to 10" diameter. But if we take the diameter down to 10" (but still need a 30% plus increase in power) we are going to have to increase the pitch dramatically - maybe trying something like a 10x10 [I have tested a 10x7 for example - but that only gave a puny 275W]. But even if a 10x10 prop could deliver something like 420W, such a large pitch prop is not a good idea on this aircraft. We will almost certainly have to hand launch this most of the time (in actual operations). Hand launching a 3Kg model is not to be undertaken lightly! To avoid having to perform a hernia inducing, javelin style, toss into high speed flight we need a model that can accelerate away well - a large pitch prop won't do the business here. To get good accelleration we want a smaller pitch, not a large one. So, what's the solution? Well we are facing the same delemma that faced the generation of Mitchel and Camm - too much power for the prop! They solved it by going to multi-blade props - so must we. The best bet here would be a 10" diameter 3-blade prop. What pitch? Well that's a bit of a guess, 10x6, 10x7, 10x8? Possibly. There is not alternative really to buying a couple of 3-bladed props and trying them out. Which is what we shall do! BEB Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 1, 2016 Author Share Posted April 1, 2016 Right some 3-bladed props of 10" diameter and a range of pitches ordered, let's see what they produce on the wattmeter. To do basic FPV all you really have to have is a camera, a videoTx/Rx pair and someting to view it all on - pair of goggles or a screen. That will work, but it has major limitations - some of which, such as loss of orientation and positional awareness, pose potential safety issues. We greatly benefit from a bit of clever electronics to help with this. In my case the particular bit of clever electronics will be this... An EagleTree Vector system. At heart, the Vector does two jobs: 1. It is an On Screen Display (OSD) unit. It is actually a very powerful OSD; the user is able to specify exactly what information is displayed. This is useful as different information is often useful in different operational circumstances and having everything present just leads to a clutterd display and potential pilot overload! Another respect in which this is very good OSD is that it comes with a wide range of sensors such as; battery cpaacity used in mAh, temperatures, barometric altitude, air speeds, GPS (which as well as giving position can also estimate ground speed and altitude), magnetic compass etc. Further it can interface with many of the latest receivers and pick up RSSI. So, this is an OSD with a great potentail for providing a selection of really useful, mission critical and safety enhencing information. 2. But there is more...the Vector is in reality a Flight Controller - just like the FC's we find in multi-rotors. Indeed it can be used as an MR controller. As an FC it s again very powerful. Whilst most FC's have half a dozen or so flight modes the Vector has well over 20! Some of these mimic functions we get from other systems, so for example there is a flight mode that mimics the presence of a 3-axis stabilising Rx gyro. There are flight modes that invoke straight and level (altitude holding) flight, there are ones that create "loitering" flight in a circle around a set GPS point. And, very importantly, there is a Return to Home (RTH) flight mode which can; display an arrow on the OSD pointing in the direction of "home", display a compass bearing to fly to get "home" or even autonomously fly a leg "home" all on its own. This is an important safety feature. To get some of the operations this aircraft will need to do passed the safety threshold and profesionally acceptable in terms of risk assessment, we must have features such as this. You might be thinking "surely all this automated flight tales the fun out it - don't we build these things to fly them? Where's the fun if the aircraft can basically fly itself?" These are good questions but it has to be remembered that this sort of aircraft is not just about flying - its designed to do a job. That job might be photography for a hobbist or surveying for a professional, but the point is, it isn't just flying, its flying for a purpose. Actually often you need the flying to "take a back seat" and if the aircraft can "look after itself" for a few minutes while you focus on getting the photographs/data so much the better! That is the value of these flight modes. So, how ar we going to do this? Well this is going to be a very complex aircraft indeed. I've found over years of doing these sort of things that a step-by-step approach is best, building it up and testing each step before you go on to the next. In my view this is much to be prefered to the "throw all the technology at it at once" method which, when it then doesn't work, you haven't a clue why and what exactly isn't doing what it should! So, the next step will be to just get the Vector running as a Flight Controller on which we can invoke simple flight modes which don't require GPS such as stabilsation modes that depend only on the accelerometers and gyros. Then add the GPS and invoke more advanced flight modes such as loitering etc. All of this with no video, still flying VLOS. Once all that is working only then we will add in the video and a simple OSD. BEB Quote Link to comment Share on other sites More sharing options...
cymaz Posted April 2, 2016 Share Posted April 2, 2016 BEB , your man on YouTube was using a 3 blade prop. That would be the obvious solution, as you say, the blades can't get in front of the camera...that defeats the whole object of the excersise. Did you start with two blade prop because that's the power system you had to hand at the build? Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 2, 2016 Author Share Posted April 2, 2016 Posted by cymaz on 02/04/2016 05:30:36: Did you start with two blade prop because that's the power system you had to hand at the build? Yeap, exactly. It was a case of what was "in the cupboard" that might fit the bill. BEB Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 2, 2016 Author Share Posted April 2, 2016 OK, let's make a start on the first phase - implementing the Vector in stabilsation mode. To begin I actually had to take a number of steps backwards! Step backward number 1: I had to rebind the Rx. This was because I had set up the X8R in D16 mode and as channels 1-8. This of course is perfectly fine as far as it goes, and as far as it goes is channel 8! But, one of the little featurettes of the Taranis is that, although this receiver is called an X8R and has 8 servo connections, so you might think that it is an 8-channel Rx - you'd be wrong! Its actually a 16-channel receiver with just 8 servo connections. So what's the point of that? Well you can get all 16 channels from one Rx but you have to do it via S-bus. I will want to do this - it's perfectly possible that this aircraft could require more than 8 channels. So I needed to reset the Rx to give me channels 1-16 on S-bus and then rebind. Step backward number 2: I have now to disconnect all the conventional PWM servo connections to the Rx - they will ultimately be reconnected to the Vector - and physically move the Rx. The Rx has to be moved because its currently sitting precisely where the Vector has to go. The photo below shows the Vector in its required location... The point shown via the superimposed red arrow has to be located at the aircraft's CoG. And the Vector unit has to be orientated so it is level, writing side up and pointed forward in the direction of the arrowhead on the unit located just in front of my superimposed arrow. The unit is fixed to a small square of ply that has been glued in, and is held in position via some double sided tape. The new location of the Rx - much further forward than before - is shown approximately, but the Rx is not fixed down yet. Power to the Vector is suppled via its own PSU which incorporates a current sensor that we will be able to tap into for mAh used data. This PSU is shown below.... In addition to the PSU a dedicated wiring harness is also provided to connect the Vector to the Rx, as shown below. As you can see from the photo there is only one "full three core" servo lead - the others are "signal only". As we are using S-bus to make the connection between the Rx and the Vector we only need the single three-core connection - all channels will travel along this lead - and the signal-only wires for the others are simply cable-tied up out of the way... Now we have a big decision to make - how are we going to power our servos? EagleTree are very specific - we must not power them from the Vector itself - its power is only to be used for the FPV camera and VideoTx. So, either we power the servos from the ESC's BEC or we provide a separate power supply battery. The complication is that the Vector's PSU, shown above, sits between the flight battery and the ESC - this then powers the Vector. This complicates tapping into the ESC's BEC without accidentally accessing the Vector's power system. But I will accept the complication, because I don't really like separate power supplies! I know many people favour them and believe they are safer - I don't. Folks often say that providing separate power provides redundancy - it doesn't. It would only be "redundant" if there were two supplies - there aren't. What there is is an independent supply, not a redundant supply. To me, an independent second supply is an increase in the component count and just one more system to fail! So, the ESC's BEC it will be - but how? Well, we need the servo lead from the ESC to do two different things now: 1. It has to power the Rx and the servos 2. It has to pass the throttle signal on to the ESC from the Vector unit - which in turn has got it from the S-bus output of the Rx. To do this we need to split the ESC servo cable. The easiest way to do this is with a Y-lead. This is shown in the upper part of the photo below... But the input end of the Y-lead is a male, and so is the end of the ESC servo lead! So we will need a "gender bender" to connect the Y-lead to the ESC servo cable - I made one up with my crimping tool and its shown in the lower part of the photo above. The two arms of the ESC servo lead we have now are treated slightly differently - as shown below.... The left-hand arm carries the ground and signal and will be connected to the throttle signal output of the Vector and so allows the control of the ESC. The right-hand arm carries power and ground and will be connected to any channel pins of the X8R Rx to provide the Rx and the servos with power. So, that's everything connected together that we need to connect at this stage. Next we need to do the software configuration to set up the functions of the Vector. It will be middle of next week before I can get to that as I have a busy period coming up! BEB Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 5, 2016 Author Share Posted April 5, 2016 Well, its been quite an afternoon! What was it I said, "Everything is connected together, we just have to run the configuration software", yes? Well I did that,... Everything went very well. The software confirmed that the Vector was picking up the channels from the Rx, I was able to run the "Rx Wizard" and calibrate the sticks etc. I managed to set up a combination of two three position switches to give me a choice of 5 flight modes in the following combination: S1 Pos S2 Pos Flight Mode Up Any No Stabilisation Mid Up Stabilisation – no altitude hold Mid Mid Stabilisation – with altitude hold Mid Down Loiter Down Any Return to Home I tested these and the Vector picked up the flight mode changes and set itself accordingly. So, everything was great, there was just one small problem,...the ESC wouldn't arm and none of the servos actually moved! Mmm, puzzling? So I checked out all connections - had I accidently wired something incorrectly? It didn't seem so. Maybe the S-bus isn't working as I thought - go back to parallel (conventional) connections. No joy. I wonder is a there a duff lead in there somewhere? Ring all the leads out with a meter - they are all fine. Could the BEC in the ESC have packed up on me? Strip out all the wiring and put it back like it was for the maiden - ie take the Vector, and all its sub-systems, out of the equation completely, just to re-establish that the component bits do actually work. They did. Even more puzzling. So, I decided to carefully reassemble everything - but this time "out in the open" where I could get at everything easily,....that resulted in this,.... What a mess!! But,...IT WORKED! The chain starts with the PSU,... Battery on the right, PSU in the middle, two motor power wires off to the ESC on the left. The twisted cable going up into the aircraft is carrying the power supplies for the Vector itself, the camera, the VideoTx and the microphone. To help explain how the rest is wired together here is a labelled close-up of the top zone,... The power comes up from the PSU and enters the Vector. At the top left of the Vector is the harness linking to the Rx by which the channel signals get to the Vector. At the top right of the unit are the servo connections to the rudder, elevator, throttle and ailerons. There is also a single wire coming in at the top right of the unit carrying the RSSI signal from the Rx. Finally, over on the right you can see the GPS/magnetic compass unit and the display unit (makes setting up at the field easier). All the servos now work fine. But does the stabilisation work? Well here's a short video taken with 2D Stabilisation (no hold) showing the elevator while I lift and drop the tail,....see if you can spot the deliberate error! Yes! Its correcting backwards!! No big problem - we can fix that. So, what was the original problem, why wouldn't it work at first you are wondering? Easy - I still had the USB lead connected to the Vector in set-up - and guess what,...the servo outputs don't work if you do that! yes, yes,....I am aren't I Never mind, got there eventually. So, next job is to neatly install it all again, then double check all the throws and directions (including the stabilsation corections!) before a test flight of the aircraft with the stabilisation in place - hopefully at the weekend. BEB Edited By Biggles' Elder Brother - Moderator on 05/04/2016 22:11:25 Quote Link to comment Share on other sites More sharing options...
Peter Miller Posted April 5, 2016 Share Posted April 5, 2016 AH!!! Spaghetti for supper I assume Quote Link to comment Share on other sites More sharing options...
cymaz Posted April 5, 2016 Share Posted April 5, 2016 You're going to need a bigger plane to fit all that in ,BEB Airbus do quite a large plane! Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 8, 2016 Author Share Posted April 8, 2016 Well, its all in - and not so bad once once I tied up a few cables n routed stuff a bit etc,.... Not the neatest installation in the world - but quite adequete I think. As can be seen the Rx has moved to the side and that has made room to allow me to install the display unit semi-perminantly. This is handy as it displays the status of all the compoenets - useful whilst we are still commissioning everything! The GPS and magnetic compass unit is sitting between the display and the Vector unit itself. The power harness wiring is passing down the central duct and into the lower pod. We will need to pick up on this when we add the VideoTx and the camera. I tested it all out again (typical doubter!), and it all still works! OK 3-blade prop testing time. I got a 10x7 and a 10x8.3 3-bladed prop. Here's the latter installed whilst the former "stands-by" ready to be tested: And the results were: Prop: 10x7 Current: 23.6A Power: 344W Prop: 10x8.3 C urrent: 28.1A Power: 411W So we estiimated in an eatrlier post that (for the projected final AUW) the power should be aound 420W, so the 10x8.3 looks pretty close qnd we'll try with that first.. But one thing that we will need to look out for, 8.3 is a large pitch step up from the 5.5" used on the maiden. The effect will be to make the model faster at any given rev setting and it may result in the aircraft being more sluggish on accelleration, that could complicate hand lauches - but let's hope not! Just off to recheck the throws, (direction and extent) and the the auto-stabilisation throws (for gains and extent). Then its flight number 2 and the first real test of the Vector unit itself - at least as a Flight Controller. BEB Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 13, 2016 Author Share Posted April 13, 2016 Right, the second flight of the world's ugliest aeroplane has taken place - a test of the new 3-bladed prop and the basic flight stablisation systems in the EagleTree Vector. Got down to the field early and guess what? The Park's Chaps turned up with a mower to cut the strip! Excellent, nice short grass so it would be worth another go with the undercart. I found I could just about get it to roll, but there was no steering whatsoever! So taxying was out of the question. So a helper carried it out and I tried a straight forward take off run - hoping that I'd get some rudder autority fairly soon in the process. No chance! It just went on its own sweet way and any attempt by me to steer it was about as effective as trying to nail jelly to the ceiling! OK, helper briefed to hold the fin, I gun the motor and we just "go for it". That just resulted in the nose wheel digging in (as the thrust line is much higher than the nose wheel, even more so than usual) and the whole thing nosing over. So, it was a hand launch again. (That undercart is definately coming off - its about as much use as a chocolate poker!). I was concerned about the hand launch for the reason I gave in the previous post, the relatively large prop pitch would mean that the aircraft's acceleration would be poor. Top speed would be good, but it would take a while to get there! Its a bit like trying to do a drag race start from the traffic lights with the car in 4th gear! I was right to be concerned! We got off, but only just! Whereas last time she sailed away from a hand launch this time she wallowed, lost hieght and generally gave me a hard time! I literally made it with inches to spare! OK, it was up, So how is it flying? Quite well. A bit more trimming required and all wa fine. Time to test the stabilisation. I gained height and switched in the 2D stabilisation without altitude hold. Result! It was excellent - most pleasing. I could put a deliberate pitch and/or roll perturbation in, relaease it centring the sticks and the aircraft promptly returned to straight and level all on its own. Great. I was just about to move on to testing the 2D stabilisation with altitude hold, when I started to detect a rapid drop off in power. This was surprising as I'd only been flying for less than 10 mins or so and although I only had a 4250mAh battery on board current use calculations (see above) and the fact that level flight was being achieved on far less than half throttle meant I should have a substantial endurance. But, the fact was the power was fading fast! So I did a quick turn to get it lined-up with the strip and tried to get her in. She pancaked the last 3-4 feet landing quite hard - she has a nasty stall. No significant damage done - except to the U/C but as that is coming off anyway no problem. But - what happened to the power? I checked the battery on removal - 70% left. So it wasn't that. A bit confused I decided to investigate at home. Once in the shed the culprit was located - the prop was just slightly loose. Not floppy loose - but a tacho test revealed full throttle giving me about 800rpm! A quick twist on the prop showed that it could rotate on the shaft. So that's been tightened up. Before the next outing the U/C will be removed and a skid will be added to the underneath of the fuselage also some outrunners on the wings. And I need to give the hand launch problem some thought - as I say we got away with it - but only just. She currently needs a really vigorous "heave-ho" to get off - do I just live with this? BEB Quote Link to comment Share on other sites More sharing options...
john stones 1 - Moderator Posted April 13, 2016 Share Posted April 13, 2016 Dunno BEB it's a working tool, so needs to be reliable or a days works may be lost, having said that was the prop ever giving it's best with being loose ? John Quote Link to comment Share on other sites More sharing options...
Don Fry Posted April 14, 2016 Share Posted April 14, 2016 Why not catipult it off a inclined rail, rail can be interlocking, so car pack able, and machine foot launched, and a solo job therefore. Thats the way the army launched small UAV. Have seen how to do its on the web. Quote Link to comment Share on other sites More sharing options...
Biggles' Elder Brother - Moderator Posted April 14, 2016 Author Share Posted April 14, 2016 Yes that thought had occurred to me Donald. I've been looking at this design. It seems to be popular and there are a number of threads and websites that detail building this particular design or an adaptation of it. BEB Quote Link to comment Share on other sites More sharing options...
Piers Bowlan Posted April 15, 2016 Share Posted April 15, 2016 BEB, in your second posting in this thread you estimated that with a 4s 10,000mAh LiPo the flight time might be over 2h. However by page 3 you had down graded this estimate to an hour. Your need to increase the power output to 420W (28.1A) and the necessity of using a (less efficient) three bladed prop will reduce your endurance still further. I know you stated in an earlier post that foam construction was more practical and reduce cost and repair time but do you still think that this model can fulfil its surveying role? The Skua might be light, although perhaps not so light with a 10Ah LiPo, plus all the payload, but surely low drag is equally important for an efficient design. Irrespective of the aesthetics the Skua's profile drag looks horrendous which will have to be countered with thrust/power. I hate to rain on your parade BEB but I would have thought that someone with your considerable knowledge of aerodynamics and design could come up with something much better and fit for purpose than the Skua. I appreciate that time may be an issue and the Skua may, in the end, fit the bill for what you want it to do. I don't know if you have seen this but the Gemini V2 has achieved 4h endurance despite being a fairly low tech (if larger) design. I wondered if you might come up with a 2m version for hand launching and belly landing. Incidentally they have experimented with a battery made up of Sanyo/Panasonic NCR 18650GA cells. I expect you know about these. Interesting thread, I love the technology! Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.