Yet another Single Channel Encoder

Button pushers
Post Reply
MaxZ
Posts: 194
Joined: 31 Jan 2019, 11:48
Location: Boskoop, Netherlands

Re: Yet another Single Channel Encoder

Post by MaxZ » 08 May 2019, 09:21

VC10 RON wrote:
08 May 2019, 08:27
How do you manage to make those holes so neat especially covering them with solorfilm?
With time, Ron :D :D

Max.

MaxZ
Posts: 194
Joined: 31 Jan 2019, 11:48
Location: Boskoop, Netherlands

Re: Yet another Single Channel Encoder

Post by MaxZ » 11 May 2019, 12:34

Hi Mike,

Yesterday I have taken the Tomboy, my "victim" to learn s/c control, out for a spin to assess the travel settings for bang-bang operation. I adjusted the dual rate settings on my propo Tx until I was able to throw the sticks from stop to stop until I was satisfied with the plane's reactions.
I am now setting up my s/c Tx initially to be connected to the propo as a student. That did not take much, because all the travel and trim settings are determined by the propo. I only had to adjust some midpoints (don't ask me why), and reverse the elevator travel.

Now for my question; I remembered that I mixed 32% throttle into the elevator down travel to compensate for a lack of downthrust in the Tomboy setup. It's ok in the trainer-student setup, since that will be taken care of by the propo settings. But when I am going solo on the s/c, besides having to reset all of the travel and trim settings, I would like to have some down elevator compensation as a percentage of the throttle output.
Is there an easy way of adding a few lines to the code to cater for that?

Cheers,
Max.

Martin
Posts: 347
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Re: Yet another Single Channel Encoder

Post by Martin » 11 May 2019, 13:16

My single channel encoder allows different elevator (and rudder) centre positions for power on / power off - and you can also specify different throws for power on / power off. Basically there are two control 'end' positions plus a centre position for each power setting and you can edit them all independently.

You probably don't want to switch to a different encoder at this late stage - and my system requires the transmitter to be equipped with a display and encoder to make adjustments. You might wish to try it for your next transmitter build. The main advantage, compared to most other encoders is that you only need one transmitter for all your single channel models as each model has its own memory with adjustable settings. Some people see that as a disadvantage as they enjoy building a different transmitter for each model! Also the whole display and lots of adjustable settings doesn't fit well with the 'retro' aspect of single channel in many modellers' minds.

The authentic vintage way to do it is to add some down thrust to your motor - you could test fly with your proportional transmitter and keep adjusting the down thrust until no throttle-to-elevator mixing is necessary. With most motor installations you just have to slack off the motor fixing screws and insert some packing behind one or more of the motor lugs: you can use washers around the screws which is a bit more fiddly as you have to take the screw(s) right out, but it is more permanent once you've determined how much packing is necessary.

MaxZ
Posts: 194
Joined: 31 Jan 2019, 11:48
Location: Boskoop, Netherlands

Re: Yet another Single Channel Encoder

Post by MaxZ » 11 May 2019, 20:42

Hi Martin,

Yes, I have read about your encoder, and I will certainly consider it if and when I build another transmitter. but right now I still have to get to grips with the one I have built.
As far as downthrust is concerned, I know how increase it by adding some form of packing behind the motor lugs, but the way I constructed the Tomboy leaves me very little room to do it that way, hence my "electronic" solution of mixing in some elevator. So I am hoping that it can be incorporated with a few extra lines of code in the S/C transmitter as well.

Cheers,
Max.

Mike_K
Posts: 346
Joined: 16 Feb 2018, 06:35

Re: Yet another Single Channel Encoder

Post by Mike_K » 11 May 2019, 22:08

Hi Max

It's good to hear you are at the flying stage with the S/C transmitter.

Yes, a throttle to elevator mix would be a fairly straightforward addition. At the moment the encoder uses 16 different configuration settings out of a possible 32, so there are 16 spare. I'm part way through adding a proportional rudder (for 1+1 transmitters) to the encoder, so once that is completed and tested I'll add the throttle to elevator mix. I probably won't get time until after the Ponty meet next weekend as I still have a couple of models to complete - why do things always take twice as long to finish as you hope they would?

The throttle to elevator mix could be done in two different ways, either an elevator value could be set for full throttle and then the encoder would calculate the required elevator for mid-throttle (as a percentage) or copy what Martin has done and have two values, one for mid and one for full throttle. Any preferences? Also, I'll make it so it can be either down elevator or up elevator to cover both situations.

Mike

MaxZ
Posts: 194
Joined: 31 Jan 2019, 11:48
Location: Boskoop, Netherlands

Re: Yet another Single Channel Encoder

Post by MaxZ » 11 May 2019, 22:46

Hi Mike,

I am an electric only flyer these days, so for me it woud make sense to just add a value for full throttle and calculate the mid throttle value since the power curve is pretty linear I guess. The current mix I have on my propo works out as a linear increase of down elevator with the throttle setting, which is fine.

I could imagine that an I/C motor with its non-linear power curve would be better off with two different values though.

Cheers,
Max.

Mike_K
Posts: 346
Joined: 16 Feb 2018, 06:35

Re: Yet another Single Channel Encoder

Post by Mike_K » 13 May 2019, 09:03

Hi Max

I'll do a compromise. I'll program two values for the throttle to elevator mix (full and mid throttle values), but if only the full throttle value is set, it will calculate the mid throttle setting automatically. So you'd only need to set the full throttle value, but you could set a mid throttle value if needed.

I doubt most people would use the throttle to elevator feature with I/C engines as it is usually so easy to put some down thrust in by putting some washers between the engine and bearers.

While I'm at it, would there be any need for a throttle to rudder compensation mix? I've never used one, but then I've never used a throttle to elevator mix on an S/C model either. In fact, all my S/C models don't have elevators :)

Finally, I'm feeling too despondent to do a separate post on the progress on the 3D printed OS Pixie MK2 case, so I'll do it here.

Printing the lid is getting beyond my ability (or at least the ability of my modified Velleman K8200). Or maybe it is the slicing program I'm using, as the detail on the printed lid is not replicating what I've drawn and seeing in 3D viewing programs. And worse, last night I decided to upload the STL file to get it printed commercially and three different sites say there are manifold errors in the STL file. I didn't know what a manifold error was until I looked it up on Google. And I found nothing on how to avoid manifold errors when using OpenSCAD. So is it the manifold errors causing my printing problems? Or is the detail too fine to 3D print?

Maybe I should have taken Neils advice and used a different 3D cad program. Hopefully, I'll get enthusiasm back before the weekend and have another go at printing it using different settings and take it to Ponty.

Mike

OS_Pixie.jpg

MaxZ
Posts: 194
Joined: 31 Jan 2019, 11:48
Location: Boskoop, Netherlands

Re: Yet another Single Channel Encoder

Post by MaxZ » 13 May 2019, 10:26

Hi Mike,

Your "compromise" sounds good. Throttle to elevator on I/C engines; there is not much difference between I/C and electric, if I'm honest I could do the same for increasing down thrust on the Tomboy, it is just fiddly because of the confined space the motor is in.
Re. throttle to rudder; I have not yet encountered a need for that. I usually set the firewall 2 or 3 degrees to the right which seems to work out ok. Any tendency to turn under power can be countered easily with a blip of rudder, or just leave it be. On pitch it is different, you cannot blip the elevator downwards so you will have to build in sufficient down thrust or compensate by mixing elevator with throttle.

I don't have a 3D printer, but I do have some experience with STL creation. I send those off to get printed commercially (Shapeways), or occasionally to a friend who has a printer. I am also not familiar with OpenSCAD, I am using McNeel Rhinoceros for 3D cad. I know that manifold errors are caused by either "naked edges", i.e. only one face connected to that particular edge while there should always be two to make the STL watertight, or on the other hand too many faces connected to that edge.
I have to guess here, but could it be that you are trying to create the surface detail and have it just touch the surface of the main body? In that case it can happen that the STL generator does not see it as one object and skips the detail. Or maybe your slicer does not recognise it. I always make sure the detail intersects with the main body before generating a STL and my cad system seems to make a proper watertight and single outer surface file out of it, but this may differ from system to system.

Cheers,
Max.

Martin
Posts: 347
Joined: 16 Feb 2018, 14:11
Location: Warwickshire

Re: Yet another Single Channel Encoder

Post by Martin » 13 May 2019, 11:21

Yes - make sure that any intersects (unions or differences) overlap. For example when drilling a hole down through a cube, make sure that the cylinder starts, say, one millimetre below the cube, and extends, say, one millimetre above.

That said, I have found manifold errors in some of the more complex items I've created in OpenSCAD, even though I've been very careful to implement overlaps wherever possible.

There are some on-line STL viewers that will attempt to fix manifold errors in your STL file and allow you to save the modified file. They tend to be the ones that are trying to get you to purchase a finished printed item, but you don't have to actually do that.

I now use Fusion 360 for more complex designs - it doesn't seem to create manifold errors in STL files - in fact I think it may be able to import, fix, and save existing STL files, though I've not tried that. When complex features and assemblies slow OpenSCAD down to a snail's pace, Fusion 360, running on the same powerful PC, seems to be able to cope easily.

Downside of using Fusion 360 is that it has to run on Windows or Mac, it seems to need a fairly powerful PC to run it, and although it is free at the moment for hobby use, there is the nagging doubt that you may have to pay in the future to continue using it.

I also spent a lot of time learning FreeCAD. In my view, this is promising for future use, but not really ready for everyday use yet - you have to employ lots of work-arounds to allow for missing features or bugs. That said, it's been a few months since I last tried it, and it is improving all the time.

User avatar
tiptipflyer
Posts: 239
Joined: 16 Feb 2018, 22:49
Location: Germany

Re: Yet another Single Channel Encoder

Post by tiptipflyer » 13 May 2019, 13:20

Hi Mike,
regarding your OS Pixie MK2 case, I would like to give it a try. The budy, who is doing the 3D-printing for me, is quite experienced and we got very good results when printing my Futaba FT-3 case. Maybe we could improve the printing quality on your case.

Cheers
Frank

Post Reply