g monkey's blog View Details
Posted by g monkey | Apr 02, 2021 @ 11:49 AM | 3,819 Views
M2/Heliflight3d/TX16s set up

Here are the files that I set up on a Radiomaster TX16s transmitter for my M2 with the Heliflight3d FC, along with a brief description on how they are used.

Transmitter model file, M2 HF3D.otx
timers:
Timer 1 is a count up flight timer (telemetry will warn of low bat), reset on FC disarmed
Timer 2 is used for total model flight time

flight modes:
TX switch SE changes FM
flight modes set separate cyclic and rudder rate/expo values,
and IU1 and IU2 set a flat throttle value which can be varied by S1
FM1, normal, throttle uses curve CV1, collective curve CV4
FM2, idle up 1, throttle uses GVAR5, collective curve CV5
FM3 idle up 2, throttle uses GVAR5, collective curve CV6

throttle hold and arming FC:
Throttle hold is set by TX switch SF, goes active in combination with the throttle position.
Throttle will only go active if switch is set and throttle is low (must be below -50 or 25%).
The flight controller is armed by switch SG

Other switches:
TX switch SH (momentary switch) will activate rescue mode
Rotary knob S1 will vary the throttle setting in IU1 and IU2 by +/-30 (15%)

Global variables
GVAR1 - cyclic rate
GVAR2 - rudder rate
GVAR3 - cyclic expo
GVAR4 - rudder expo
GVAR5 - throttle value for flat curves in IU1 and IU2


lua script instructions

The lua script 'main.lua' should be put under 'WIDGETS\HeliAll'
It is set up for a large screen as on the TX16s.

The script relies on the following to be set up in the model file:
Global variables
GVAR1 - cyclic rate
GVAR2 - rudder rate
GVAR3 - cyclic expo
GVAR4 - rudder expo
Logical switches
LS02 - throttle hold 0, active 1
LS03 - flight controller armed 0, disarmed 1
Posted by g monkey | Jan 15, 2021 @ 03:00 PM | 11,824 Views
Flying with HF3D on the OMP M2

I thought that as a novice pilot I wouldn't really be able to tell the difference between the stock flight controller and the Heliflight3D controller. I was pleasantly surprised to be proved wrong.

Although the OMP M2 has received a lot of praise for its build quality and flying characteristics, there has been some criticism regarding a 'robotic' or 'notchy' feel to the cyclic response. This quirk is thrown into sharp relief by the HF3D controller, which has no dead-band at all. I found that this makes a big difference to the kind of flying that I am doing - hovering practice in different orientations, slow controlled flight and accurate spot landings - where many small corrections are required. Perhaps better pilots can easily work around this (or with appropriate curves in the transmitter), but I now realise that I was battling against this dead-band on top of trying to learn basic flying skills.

The M2 is simply easier to fly with the hf3d controller. I've dialled down the cyclic rate on the transmitter to 40%, tail to 70% and with a tame pitch curve, I have a fluid, responsive and smooth flying machine that any novice could use.

Of course, the hard core 3d pilots will appreciate the tail holding ability of this controller too, but I am some way from stressing this aspect of the flight envelope.


Telemetry

I now have some very useful information coming back to my transmitter. I have set alarms based on the power...Continue Reading
Posted by g monkey | Jan 13, 2021 @ 06:41 PM | 10,657 Views
Looking for a cool lockdown/winter project? Do this!

This is assembling an open source based FBL controller and installing it into an OMP M2.
Why do it? Well, you end up with a fully configurable and innovative flight controller with top notch flight performance and some very useful telemetry features.
Like what? Like motor current telemetry, so you can monitor the actual battery power used. Also, you have access to all of the FC parameters so you can configure it directly from the transmitter.

If I can do this, anyone can. I got into r/c helis nine months ago during the first lockdown, starting with a Nano S2 and a DX6. I then moved on to an OMP M2, and when I started on this project I knew nothing about FBLs, ESCs, OpenTx, etc.
Was it a lot to learn? Yes, but if you have the time to read, why not?

What I have tried to do here is highlight the internet sources which I found to be very useful and add a few tips based on my experience.

The thread that started it all, by eqtrian, introduces the project and documents its progress (page 1 and the bottom of page 18 are perhaps the most important):
https://www.rcgroups.com/forums/show...ller-%28FBL%29

P.I. Engineer has put together a handy zip package that is available/pinned on the Heliflight 3D discord channel
http://discord.heliflight3d.com/, then under the opm-m2 section.

His package contains the following files:
- compiled firmware for the FC board (Heliflight 4.1)
- a dump of the commands for configuring the...Continue Reading
Posted by g monkey | Dec 08, 2020 @ 05:42 PM | 11,309 Views
She is finally finished! Pictures below, plus the real one that inspired the paint scheme. I had no access to decals - they would need to be white for the registration and have a white background for the 'danger' arrow, so not so easy to make yourself. It is a shame because some more detail would look great and distract you from the paint job. I'm not sure where my talents lie but they do not include painting scale models. This took me forever to do and I'm still not completely happy with the results. I was using Tamiya spray cans but even if I had a proper airbrush I'm not sure that I could do better.

She came in at 315g without battery, so about 61h heaver than the stock M2 and only 8g heavier than my PLA prototype. I used Prusament PETG filament for everything except the frame, where I used Prusament PC Blend. The windows didn't come out quite as transparent as I would like. If I sprayed more clear coat over them they might improve, but at the risk of making the paint run in other places (don't ask). The next one is the EC135, which will be much easier to fit everything into as it will be larger scale. I might look to do separate window sections with that one - it would certainly make it easier to paint!

STL files can be found here if anyone wants to have a go:
https://www.prusaprinters.org/prints/46068-bell-206b3
Posted by g monkey | Sep 22, 2020 @ 05:52 PM | 9,496 Views
Colorfabb XT-CF20 Prints
I now have the frame, tail boom, vertical and horizontal stabilisers all printed in the CF20. It does feel much stiffer and stronger than the XT, so I think that the airframe will be fairly robust. I have started the post processing of the parts, giving everything but the frame a coating of XTC-3D, as recommended by several people on the web. This is just a thin epoxy coating that helps smooth the layer lines and hide small artefacts, but I have done a layer, sanded, and then done another layer and sanded . . . and I might have to do some more sanding. I decided to try another print of the rear of the fuselage as it was taking longer to sand it than to try for a better print. The XTC does make the XT more transparent though (see pic), as it smooths the surface and reduces diffraction.
After the sanding it will be time to mask the windows and get a layer of primer on.
Posted by g monkey | Sep 02, 2020 @ 04:51 PM | 8,957 Views
Colorfabb XT Prints
This seems to be tough stuff compared to PLA, more flexible and less prone to shatter. After printing out the front fuselage half I decided to do the nozzle swap from the 0.4 to the 0.6mm to see how it would look, as it has one less perimeter with the larger nozzle. It is certainly more transparent but also slightly thicker and heavier, and it feels extremely robust. I'm going to try a resin coating on it as I have read that that also improves the transparency. Next up is the Colorfabb XT-CF20, where I'll try it on the frame and tail boom.
Posted by g monkey | Sep 01, 2020 @ 09:15 AM | 10,551 Views
Proof of Concept
After getting all the bits printed, tapped and bolted together, the all up weight excluding battery came to 307g, an increase of 53g over the stock M2. A quick test hover in the garage followed by a few flights outdoors confirms that it flies well despite the extra weight, and she looks great in the air. I did increase my throttle curve slightly from 70 to 75% as the ESC was getting a bit warm, and the tail motor was working harder as it is mounted on the wrong side compared to stock, but overall I was quite pleased. She is still a lively machine - at least for me, as I binned it on the third flight! The frame shattered but the fuse held up pretty well with just a couple of cracks along layer lines - not what I had expected at all. Really just the landing gear and frame to reprint and she is ready to go again.

Next Steps
The plan was to get a working prototype/design and then investigate other materials. The next one is transparent Colorfabb XT, a PETG based filament. This is unlikely to produce completely clear windows for a fuselage, more like frosted glass, but still a better look than painted windows. I have printed the landing gear and it is clearly (hah) more flexible/tougher than PLA, so perfect for that application.
This should also be ok for the front fuse shell, although it may not be stiff enough for the rear half, we shall see. After this I will try their carbon fibre infused XT-CF20 to print the frame, but that will require a nozzle swap. Once I am happy with the materials and prints I'll move on to a proper finish and paint job.
Posted by g monkey | Aug 18, 2020 @ 07:27 AM | 9,860 Views
Fusion Confusion
Edison said something about “successfully finding 10,000 ways that won’t work” while trying to invent the light bulb. I lost count of the number of redesigns that I made while trying to get a good fuselage surface and then turning it into a solid. I hit a brick wall in the design process where I could create surfaces and stitch them together, but then not be able to thicken that into a shell. I finally asked for help on the Fusion 360 forum and got some excellent help and advice. For those interested in the detail, the advised workflow in this case is to use the ‘surface’ menu to perform the ‘loft’ command. There are several similar ‘loft’ entry points with the same (to me at least) user interface. This can be accessed through the ‘form’, ‘surface’ or ‘solid’ menus. I had been using ‘form’, which results in a t-spline surface (useful if you then want to free form sculpt the surface latter), but it is more complex to use and it is not parametric (a pain if you are making design changes). Fusion also doesn’t like to deal with surfaces that are nearly tangent, or to thicken surfaces that narrow to a slender point – these last two things can cause a lot of frustration, as it will simply fail and not tell you why. Once I got past this point I made good progress.

Reference Photos
I also discovered that there is a Bell 206 based at my flying club, which I had never noticed before (ok, I am a fixed wing pilot). That gave me the opportunity to get lots...Continue Reading
Posted by g monkey | Aug 04, 2020 @ 08:43 AM | 9,144 Views
Slicer
For anyone who is not familiar with the 3d printing process, the slicer is the software that converts the CAD output (typically an .obj or .stl file) into a printer input file (.gcode). The slicer controls lots of printer parameters, such as nozzle and bed temperatures, fan speeds, layer heights, extrusion rates and head speeds - all of which can vary during the print. I've gone with Prusa's slicer (a version of the open source Slic3r) because they have profiles already set up for various filaments and nozzle types, which means many less things to worry about for initial prints. Everything is tweakable later, if you know what to tweak and what the consequence will be (I don't, so I'm going with the stock settings for now).

Filaments and Nozzles
I'm starting with PLA filament, as it is widely recommended to do so. It is regarded as the easiest material to print with and has fairly good mechanical properties. The downside is that it can degrade outdoors, it is somewhat brittle, and it has a low glass transition temperature. The plan is to get some experience with this and then move on to ColorFabb's XT (tougher than PLA) and XT-CF20 (stiffer than PLA). The later has 20% carbon fibre in the mix and requires a harder nozzle due to its abrasive nature, so I have a tungsten carbide nozzle waiting in the wings. That nozzle is 0.6mm compared to the standard 0.4mm brass nozzle that comes with the printer. The larger nozzle should also result in faster print times...Continue Reading
Posted by g monkey | Aug 03, 2020 @ 06:18 AM | 8,151 Views
Using Fusion 360
Fusion 360 is quite a powerful and flexible package, but there is also a fair amount to learn and it can be a bit confusing at times. There are often several different ways of achieving the same result, even various ways of accessing the same tools/menus through multiple entry points in the user interface. After many iterations and trying various approaches, I found the following workflow and tools to be useful:
Canvases - this allows you to import images to be used as a reference while you work. You can move/rotate/scale them and assign them to a plane, and also adjust their opacity. I used them for the imported 3 view and cross section drawings but also for photographs (eg. the airfoil of the horizontal stab). This is a good first step to get these set up.
Construction sketches/planes - the canvases above are meaningful to you but not part of the Fusion world, so you need to create points, lines and planes that you can refer to later on. In any sketch (2d drawing) you can designate things to be deemed 'construction' to be used as a reference, rather than 'normal' (as part of whatever you are building). I recommend building lots of construction items upfront, even if you don't end up using them. You can do this later and move them around in the timeline, but will find it easier if this is done early on.
Sketches - a good next step is creating outline sketches (top, bottom, sides, waterlines). When you create surfaces later on, you will probably use...Continue Reading
Posted by g monkey | Aug 01, 2020 @ 07:11 PM | 9,933 Views
I thought that I would share my attempt to scratch build a Bell 206 JR using OMP M2 mechanics, a CAD package and 3d printing, as perhaps others might find some of this useful or have some tips to share with me.

Background
Quick bit about me. I only started flying RC helis during the lockdown, starting with a Blade Nano S2. I did fly fixed wing RC many decades ago (think AM radios) but I'm still very much a novice pilot. I'm unlikely to progress into 3d flying, but I decided to buy an OPM M2 after reading the rave reviews. The mechanical simplicity and build quality impressed me, but I really hankered for a scale heli, especially a Bell JR - I just needed a scale kit for the M2. Not finding one of a suitable size, I decided to build one. The thing is, the material of choice is not the balsa that I used long ago. The M2 is all carbon fibre and aluminium. What to do?

Inspired by several things that I have read on the web about 3d printing, different filament materials, and the appearance of downloadable 3d printed kits (both helis and fixed wing), I decided to give 3d printing a go. In the words of a famous UK motoring journo, how hard can it be? Knock it up on a PC, click a few buttons, send it to the printer and Robert is your mother's brother. What could possibly go wrong?

Printer choice
The first decision was selecting a 3d printer. I plumped for a Prusa i3 Mk3s. Widely praised in the 3d printing community and press, it seemed a sensible choice...Continue Reading