BEB Builds a Helicopter
|Gordon Tarling||11/03/2018 09:52:35|
|188 forum posts|
BEB - Things that I've needed the most whilst learning - feathering shaft, main shaft, tail boom, tail blades, main blades, U/C struts, perhaps one of those main gears
|1837 forum posts|
I,ve always understood that getting an experienced heli pilot to set up your heli will save time and money. Tongue in cheek do your first flight in a cricket training net with you the other side
|Biggles' Elder Brother - Moderator||12/03/2018 14:46:38|
15339 forum posts
and thanks to everyone for the advice and tips, I have ordered a small stock of spares: some main and tail blades, a main shaft, feathering shaft, control links etc. We'll see how we go with them.
Martian: yes I have a heli experienced club mate who is lined up to maiden it and check out the set up is reasonable for me and not too "lively"! Hopefully that will avert at least one potential disaster.
OK, it doesn't look like I've done much here, but in reality lots of very important little jobs have been done.
The tail rotor drive shaft has now been installed in the boom and the tail drive gears engaged. That was not as difficult a job as I feared. The rubber bush on the shaft, that covers and locates the bearing, is a very tight fit in the boom and I had fears that it would prove very difficult to insert. But with the help of a little light lubrication it all turned out very straight forwardly.
Next I went round the heli putting thread-lock on those screws which had not been fully tighten yet - such as the boom retention, tail rotor unit and UC fixings etc.
Then it was time to start sorting out the servo leads - in particular the routeing of them from servo to GPro, whilst avoiding the gears and linkages of which there are many in a very small space! Many people seem to route the leads inside the side frames, while this is obviously neater in appearance I wasn't too happy at how very close cables were coming to rapidly meshing large gears! So I have gone for an outside the frame solution with the leads cable-tied at key locations. You can see this, I hope, in the following two images:
Ultimately these are all routed to the rear to end up in the vincinity of the GPro. There appears to be sufficient slack to make the connections with the GPro, but it going to be fiddly and tight I think:
The next job is to start adjusting and attaching the control links. I'd better start figuring out how this swash plate leveller thing actually works as I will have to use it in anger soon!
|Biggles' Elder Brother - Moderator||12/03/2018 15:45:08|
15339 forum posts
An update: do you remember I ordered the original nut spinner from RS, but what was delivered was a hex driver? Well guess what's just dropped through the letter box? Yeap a padded envelop from RS with what in it I ask? That's right,.... a nut spinner!!! I can only assume that someone at RS relaised the error - I've done nothing!
Anyway, well done RS for putting it right.
|extra slim||12/03/2018 15:52:30|
|406 forum posts|
Nice job BEB, I'm sorry to tell you that the assembly point you have reached is what I find the easy part!!.
The fun comes with obtaining a good setup. In my early days, I got a well known heli shop to set up my first one, only to find out later on, they had done a "bare minimum" job with a -2 - +8 range, so hovered around mid stick... but mechanically it would never reach the full extent of pitch ranges!!.. grrr...
I then had to re do the whole head etc to establish a baseline 0, and equal travel to +- 10 (latterly 12) etc. Since the first I have always taken my time to establish as near to a perfect setup as I could get, with everything mechanically set square at zero degrees, used the swash leveller to make sure the swash remained in contact with the tool all the way through the range, and then from there, establishing the curves etc was just sat with the tx and was dead easy. Initial curve was like setting EPA on a fixed wing, cutting travel down to -2 - 8, with 4-5 around middle stick, and then play with head speed to establish a decent hover.. flight modes 1 and 2, were then again easy to get V curves etc for when I started with aero's and inverted..
I actually put off learning to fly "whisks" until I was in my early 30's as peoples tales had put my off as if it was a dark art... I was looping and rolling etc within about 2 months.. if you are a decent "rudder" flying aero pilot, it really isnt too tricky.
|Peter Christy||12/03/2018 16:08:34|
|955 forum posts|
One thing you will find is that learning to fly a heli makes you much more adept in the use of the rudder on fixed wing!
102 forum posts
BEB, Sorry to be the (potential) bearer of bad news but if you have a conventional swash leveller tool the head is the last thing you fit, as the leveller slides down the main shaft and rests on the swash plate.
My normal method of assembly is put together everything (the frames, tail, install the motor fit and skids etc), so you have an almost complete helicopter but without the head, then fit the servo's, ESC and G Pro but do not connect the servo's to the G Pro yet. Connect the three swash servos to your servo tester and set them to their (theoretical) mid travel point. Fit the servo arms and then the servo to swash plate links, initially adjusted to the lengths shown in the manual. An important point here is the plastic link ends have an 'A' moulded into one side and this must always be on the side away from the swash arm as the ball links have a biased and if fitted the wrong way round they can fail or come of in flight, which makes things very interesting and some time messy!
Then connect the swash servo's to the G Pro (remove the leads to or uncouple the motor first) power up the G Pro and start to program it to set up the manual "0" degree swash position and this is when you (ideally) make any minor adjustments to the servo to swash links to get the swash plate level. You may need to use a very small amount of servo sub trim in the transmitter to get it absolutely level, remove the leveller and then fit the head and main blades with the motor still disconnected. You then programme the G Pro to give you the collective pitch and cyclic deflections suggested in the manual.
Have fun but it is a very steep learning curve which will tax your concentration and flying skills but should give you back that rush that comes from achieving a new skill.
One final comment is "a happy helicopter is one that does NOT wag its tail"
|Biggles' Elder Brother - Moderator||16/03/2018 23:37:08|
15339 forum posts
Sorry I haven't posted for few days - I've been away from home and only just come back.
The good news is I haven't fitted the head yet Barry, any pictures showing it in place its just resting there for show!
To do the next bits I do indeed intend to use the servo tester. But before that I have to set the link lengths and connect them all - they're tomorrow's jobs! See you all then!
102 forum posts
This seems like a good weekend to return to the build BEB - with the mini Beast from the East upon us and glad you have not had to remove that Jesus bolt!!
|Biggles' Elder Brother - Moderator||17/03/2018 19:31:15|
15339 forum posts
Yes, the Beast from the East - Part 2 seems to be here. It's gone very cold and we have been having snow flurries all day. so it's definitely back to the heli!
Just to be clear. The bolt I had all the trouble with was a sort of "lower Jesus bolt" - down at the bottom of the shaft next to the main gear. I haven't even taken the "upper Jesus bolt", that will hold the rotor head in place, out of the bag yet! Yes, I know, you can't really have two Jesus bolts, by definition, and I suppose its the top one that is the "real thing" - but hey, you know what I mean!
Well an interesting and challenging day's work. I started by assembling the links between the three cyclic servos and the lower swash plate. (Hey, get me, I'm using all this helicopter talk now!):
As a staring point these have to be all the same length - as accurately as possible. The manual, somewhat unhelpfully really, says this length is to be 32.5mm between centres. The reason I describe that as unhelpful is that the centre isn't a physical point, you can't put one end of your vernier gauge up against it - it's in fresh air! Ah well, we do the best we can.
Now I thought about this and came to the conclusion that functionally what I really wanted was them to be 32.5mm as close as I could reasonably get them, but more importantly, I believe, I want them the same length to greater degree of accuracy. In other words; all three at 32.7mm is probably better than the three being 32.4, 32.5, 32.6. So, how to achieve this?
After nominally setting all three links to 32.5mm by Mk1 eyeball and a pair of calipers set off a vernier, I found a tool with a parallel shank the diameter of which was a sliding fit for the ball links and then slid all three on it:
Hey presto! Its obvious the one on the right is too long by some way, whilst the middle one is "a bit" too long. A little fettling later I had:
Pretty close to equal length. I then rechecked the centre distances, they were still nominally 32.5mm, so I was happy with that as a starting point. Next up we install them on the helicopter:
We can't just stick these links on anyway round. The links are embossed with a letter 'A' on one side and, as Barry said, this must face outward, away from the swash, for all the links.
|Biggles' Elder Brother - Moderator||17/03/2018 19:32:02|
15339 forum posts
OK it looks good, but is it level? Time to find out! To do this I'll use a swash plate levelling tool:
I watched a number of videos to see how to use this - as is typical in our hobby it comes in a bag with no instructions whatsoever! The idea is that this slides down the main shaft and sits on the three promontories of the swash plate which are the ball joint mounts. The swash plate is level when the servos are at centre and all three legs of the tool are in contact. So, first we need to set and hold the servos at centre and as we said yesterday we can use the servo tester and the low power 2s lash-up to do that:
(I have often wondered why so many servo testers have three outputs - now I know why!)
So, let's see what the swash plate "levelness" is like then:
Well, not too bad, but a definite gap on the right there. Now we just adjust the links to get it sit level - obviously we are going to start by shortening that link on the right a tad. (You might wonder "if all the links are same length why isn't it level?" The answer is of course that while the servo arms are "more or less" all at 90 degrees - as near as I can get them - they are not all exactly the same! So we have to adjust the link lengths to put in the small degree of compensation required.)
This is a bit of an itereative process; remove the link, tweak, reinstall, check, remove another link,....etc. I was very happy I had invested in a pair of ball link pliers - I'd not had any before and on the odd occasions I used ball links I just dealt with them "by hand". But here the number of times you have to remove and replace, combined with the limited access to some joints, makes the specialist tool not a luxury but an essential. If you decide to have a go at a heli - buy a pair of ball link pliers, its easy with them, it would be a nightmare without them!
So, after much tinkering, I arrived at this:
It's a bit difficult to photograph this arrangement and obviously I can't show all three legs at once, so you'll just have take my word for it they all appear to have no gap! Phew.
So, what's next? Well that's a sort of "static" levelling - what we need now is to check what you could call the "dynamic" levelling - i.e. does it stay in contact at all three points as it moves through its full vertical range. To do that I need to connect the GPro - to get the cyclic mixing - and I need to program the Taranis with at least a rudimentary set up so I can control the pitch. I have been developing a very basic Taranis program in the background while we have been building, so I'll show you that later.
|extra slim||17/03/2018 21:23:32|
|406 forum posts|
Good show Mr BEB, your next task is the one that I found was done in 5mins, or takes hours of messing with EPA and adjusting eyes...but is well worth getting just right.. the good news is after that is cracked the rest is a relative breeze albeit a methodical and time consuming one...good luck buddy.
|Biggles' Elder Brother - Moderator||18/03/2018 18:40:34|
15339 forum posts
Wow, busy day, lots to cover. I may not have got to where I hoped to today, but it was a interesting and very satisfying session, so all in all I am happy with where I actually am! Let's start with the Tx programming.
Basically there are two main approaches to this:
1. You can do all the mixing between the 3 cyclic servos in the Tx - in which case, in OpenTx, you invoke the heli options and pick up the predefined cyc1, cyc2, cyc3 parameters.
2. You use an external "box" to do all that for you and as far as the Tx is concerned this is basically a plane!
We will use option 2! The GPro flybarless controller fitted to the heli will do all the clever stuff for us.
I thought I'd show some of the "milestone" points in the OpenTx program I have developed. It is important to note that this is not intended as a fully finished flyable set up. None of the parameters are properly fixed or established. It's really just the shell of a program - enough to do a Rx bind, test the connectivity between the Rx and GPro, calibrate the end points, check the directions everything is moving in and finally set up the GPro. Then we can set the "throws", ranges and breakpoints etc!
So with that in mind here is my very simple Inputs tab in OpenTx Companion:
Basically just the 4 primary inputs, as if it was a fixed wing aircraft, the weights (rates) are all fixed at the single value of 100% and a fixed 20% expo is applied across the board.
Next here is my mixes screen:
I'm setting up 6 channels, rudder (I label as Yaw), elevator (I label as Pitch), Throttle (I label as Thr), aileron (I label as Roll), Gyro and finally a channel for the collective pitch (I label as C-Pit).
The throtlle channel is switchable via switch SA. This can select normal mode and two idle-up modes. In addition switch SF gives us a throttle hold mode. (a panic button!)
Channel 6 offers 4 switchable pitch curves to match the 4 throttle modes.
Finally, channel 5 allows switching between the rate mode and heading mode of the gyro.
Curves are provided for the four throttle modes and their matching pitch profiles. In normal mode the throttle curve below is set:
While in both the idle up modes the curve below is used at present:
For throttle hold the curve is:
Considering the matching pitch curves. The pitch curve used in normal mode is:
While in all other cases (both idle-up modes and throttle hold) it is a simple 45 degree straight line as below:
So that's basically it. There are voice announcements for the throttle modes and some logic to protect against accidental start ups etc. All the outputs are set at +-100 at present, this will be altered at end point calibration for the GPro. All the switches used are under my left hand - this seems logical to me as I really can't see me taking my right hand off the stick!
A question please: at present no failsafe is programmed, what failsafes are folks using on helis - if any?
I'll post again in a little while to report on the hardware tests done today. I told you it was a busy day!
|Biggles' Elder Brother - Moderator||18/03/2018 20:34:25|
15339 forum posts
Right, to the hardware.. The first task was to bind a Rx to that Tx programming. I would have liked to use a X4R. Although it only has four physical output channels for direct servo connections, it has Sbus and, in common with all the X-series FrSky receivers, it is a full 16 channel Rx. But I don't have one without robbing one out of another model! So it will have to be a X8R, but if that works OK I'll swop it over for a "4" later. The bind was done out of the model using my 2s power lash-up system:
The solid green telling us we have a good bind.
Next while the programming does offer us some protection against the motor starting up while we tinker "under power" and the fact that the main blades aren't fitted is a another line of defence, I still believe in "belt-and-braces" so to make triply sure that this thing won't bite me I have disengaged the motor pinion from the main gear, so even if in the very unlikely event that the motor did start, it has nothing to drive:
Ok, to work. I fixed the X8R onto the heli:
I didn't want to stick a full 6s battery on this at this stage - so I'll use my "lash up". Instead of connecting the lead from the heli's ESC to the GPro, I connected an extension lead to the little test ESC. I then connected the 3 cyclic servos to the GPro and the rudder servo following te instructions. Finally I ran one single servo lead from the Sbus output on the Rx to the Sbus input on the GPro. That all looks like this:
The orange,red and brown lead is from the ESC in my 2s lash-up and poviding power, the black,red and yellow wire is the single Sbus link to the Rx. All the others are from the servos on the heli, 3 cyclic and rudder.
OK, time to test this! Connect the USB on the GPro to the PC (for initialisation and set up), turn on the Tx and connect the 2s like this:
And,..... it works!
Here is proof that the Rx is getting power through the Sbus and is happy with its RF status:
The all important green light. And here is the indication that the GPro is working fine:
Now of course its time to waggle the sticks. Did so,...the swash plate and rudder all move! Hurrah!!! We have a radio link! But hang on,....er the right things aren't moving! At first I can't figure it - the CCPM mix seems to be completely messed up! Then I start working my way systematically through each control and looking at what happens - still no insight, until,....I move the aileron stick and the rudder servo moves! Ah - I know what it is!
To be continued,..... (an aeromodelling cliff-hanger!)
|Biggles' Elder Brother - Moderator||18/03/2018 20:57:38|
15339 forum posts
The explanation (I think!)
Among the many things I like about OpenTx one of the best is the fact that the channel order is up to you.
Spektrum/JR have their TAER (Throttle, Aileron, Elevator, Rudder). Whilst Futaba have AETR (Aileron, Elevator, Throttle, Rudder). These are fixed by the manufacturer - you are stuck with them!
But in Open Tx you can have any channel assignment you like. Now obviously it wouldn't be smart to vary this too much or you'll get confused. So people tend to develop their own standard according to their preferances. I happen to standardise on RETA (Rudder, Elevator, Throttle, Aileron). YOu can see this in the mixes tab in the Taranis programming post above. The reason is that pretty well all of my aircraft have separate aileron channels, by putting the aileron at the end I can an extra one easily and keep them neatly together - RETA1A2.
Now if the servos connect directly to the Rx no problem - you know that channel one is the rudder so you physically connect the right servo to the right channel on the Rx. But here we have the GPro between the Rx and the servos and its getting the data via Sbus and it has definite ideas on what channel controls what. In short, because its Sbus, it assumes the channels are in Futaba order - ie AETR!
What is my evidence for this? Good question! The clue comes when I move the aileron stick and the rudder servo responds. You see, under Futaba rules aileron is channel 1 - but for me channel 1 is rudder! Aha gottcha!
How do we fix it? Well I just have to bow to the demands of the GPro, I will have to reassign the channels in my Taranis programming into the Futaba order - then I believe it will basically work. I may still have some issues because of reversals etc. but that's different, basically the right servo mixes should be linked to the right sticks. It's not a big job, I will do it tonight and test and report back tomorrow.
Interesting stuff eh?
102 forum posts
Hi BEB, I have not programmed a Taranis tx but it is the general rule not to put any expo on the rudder when using a FBL unit, as it needs 'solid' linear inputs or you may find the tail constantly drifts or feels very loose.
Other than that your initial settings look OK but if you ask ten helicopter pilots for their set up, the chances are every one will be different.
|Biggles' Elder Brother - Moderator||18/03/2018 23:23:59|
15339 forum posts
Thanks for that tip Barry - makes sense, I guess the last thing the Gyro needs is a non-linear response from the control system. So I've removed the expo from the rudder on the inputs page:
I' ve also sorted the order of channel allocation in the mixes tab - we should now have "Futaba Order" of AETR:
Just have to upload this to the Tx and test it tomorrow.
|Tom Thomas||19/03/2018 03:45:48|
315 forum posts
Great looking helicopter BEB! Watching with interest.
Is it me or do the plastic skids let a great looking heli down?
102 forum posts
One other thing BEB is it might be useful to increase your initial expo on the cyclic functions (Elevator and Aileron) to around 40% initially (unless you have the reaction times of a Ninja that is!). This will help to soften the area around hovering which is a skill you need to develop quickly as it is used primarily during landing.
As an example one of our clubs 3D heli Gods uses 60% cyclic expo on all his helicopters and most of his flying time is spent performing tricks and trimming the grass, whereas, I only do loops and rolls and general thrashing around and have 35% on mine.
Hope this helps.
|Biggles' Elder Brother - Moderator||19/03/2018 20:32:14|
15339 forum posts
Another good day with strong progress. I started by downloading the edited Taranis code and testing that out. Worked fine! So my idea that the GPro assumes that the data on Sbus is in Futaba channel order would seem to be right!
Right with that settled it was time to start setting up the GPro with the help of the supplied software tool. This software basically consists of a series of 6 screens. We can review them here:
On this screen we select the type of Rx set up we have - I'm using Sbus. Then for each channel we set the centre point at zero and the range as +-100. This is done by tweaking the sub-trim and the end points on the Tx. Very straight forward.
Here we select the location and orientation of the GPro - mine is down low and inverted. Then we set the direction of rotation of the main blades - for us clockwise.
Here we set a few directions and types. For this heli the swash moves up for positive pitch, so we select that. Our swash type is HR3. Finally, inspecting how things were moving it was obvious that the ch3 servo (cyclic servo on the right) needed reversing so everything worked together correct.
Here we level the swash plate again, but this time using the on-screen trim adjustment. To measure the pitch at zero we use the pitch gauge bought for the project and mentioned early on:
The full range positive and negative pitch are set here from the on-screen sliders and using the pitch gauge. The recommended range is +-12 degrees. Frankly the sounds scary!! So I've set it as +10 and -5, I think that will suit a beginner like me better!
Last of all te rudder is set up. We set the centre point and then the left and right throw are set to be the largest possible without binding.
All done - a working helicopter!!!
It looks quite easy written out like that - and indeed it was not too difficult, but it does take time. There is a good bit of reiterating and shuffling about - particularly with regard to checking that the swash is level at centre, top and bottom. Also, for some unknown reason, I had enormous trouble getting the rudder pushrod the right length!! Just tired I think!
Well pleased with where we are up to. I sat for while just twiddling the sticks and watching the swash plate rock about and move up and down - brilliant! Next "flight parameters"!
Please login to post a reply.
Want the latest issue of RCM&E? Use our magazine locator link to find your nearest stockist!