**Pictures to be posted once they have been taken. Pictures not taken during development.**
This should be a quick project, relatively, to modify a foam glider with servos for control. The Styrofoam glider has a 4 foot wing span and glides a good 100 ft or so. The rear wings are flipable to switch from a fairly level flight to doing loops. I’ve modified the rear to be permanent, then cut the back wings off. I installed a servo on each side with a support structure to attach to the foam wing. I might do well to also install support structure for the servos.
While I had the modified the fuselage on the desk, I played with the servo controls so that they move in synch. I wired up a small perfboard to connect the servos into it and deal with power and control signals. I need to add a wireless receiver for it to be radio controlled. The idea is to make it able to turn. In the process of adding the servos and perfboard, while still needing to add the micro-controller and batteries (with holder), the plane has had a significant amount of weight added to it. I think I will weigh the differences and then add some weight to a spare plane to see what will happen in a crash landing which is highly likely.
The Styrofoam was fairly weak around the servos with less than an eighth of an inch between the outter shell and the place the servo sits. I added duct tape around where the servos mount and the wings to try and help with stresses.
If this works as intended, I may see about modifying it further to allow me to incorporate a motor. If it doesn’t work, I may look into the methods of building one from scratch using insulation foam.
Aug 10 2015 – Worked a lot on this over the weekend and some evenings. I had everything operational, but the electronics not yet installed. The servos worked well, though they seemed jittery. I added code for the servos to change no more then 30 points between updates. As it was before sometime an errant reading would come in and the servos would go from min to max then back to min or where ever it was. In flight that would be disastrous. With the new modification it still reads it in and still has a slight flutter, but it is now slight instead of major and will be workable.
I also updated the code for the wireless so it is more responsive on the lighting indicator for links. I also changed the flow of the wireless code to be in line with the IMU code I created for talking to a device. Now instead of large code blocks iterating every step, there are subroutines to do the work.
Two nights ago the microcontroller I was using for this project stopped working. The 5 volts it was receiving was not being converted to the 3.3V that the microcontroller needs. 5V from the main input was yielding 0 out, while 5v from the USB was yielding 2.6V out, much too low for the microcontroller. While I initially suspected the voltage regulator, I’m still not 100% sure that is the failed part. I need to learn more about troubleshooting to determine what truly failed. That said, in both cases the 5V is reaching the Vin pin for the voltage regulator. Reading up on the datasheet of the voltage regulator, an RT8250, it says to avoid rapid on off as it can damage the chip. The 5V it is being fed is from another switching voltage regulator to bring the 8V+ from the batteries down to the 5V that the LED and servos run off of. Technically the servos can run off of 6V, but the RGB LEDs are listed as 5V max. My theory is that the first switching regulator is producing power that is too erratic and damaged the voltage regulator on the microcontroller board with rapid on/off. The microcontroller board uses an inductor to smooth it’s output, so I will need to look into doing the same thing. I had capacitors to help absorb the highs and fill in the lows, but it was clearly not enough. I’ve never used an inductor in my circuits, so this is something new for me. As a side note, through testing, I used a different board to provide the 3.3V necessary to the board on the 3.3v line. When doing that the damaged microcontroller responds normally and everything appears to work. With this testing I have ordered 5 new voltage regulators. Now I need to figure out how to remove an 8 pin SMT chip with body ground mount. I have an idea, and it involves a lot of hot air J
Aug 17 – I still do not have an inductor to use the switching regulator. I had purchased a linear regulator to replace the switching one as there is a minimum of noise in linear regulators. The down side is that it will be less efficient, but it’s simplicity at the moment has its own value. I replaced the RT8250 voltage controller on the board which had it’s own share of challenges. This chip not only had the 8 small fingers, but also a surface mount pad underneath it. Typical way to solder it is to put it on a hot plate. Due to all the other components I had soldered to this board, that wasn’t a good option. I tried using hot air to remove the old one, letting the weight of the board drop down when the solder was molten enough to give way. A side effect of this method is 2 finger pads on the boards lifted up. They were not disconnected, but did need to be pushed back down and treated with care as the traces are very small and lack any real strength. I was able to line up the new chip and heat up the pad solder with my hot air gun enough to make contact. The back 4 pins also connected to their solder, but it was lopsided and the front pins were raised up. Too much solder on the underside not getting molten enough to drop the chip enough. I was able to bend the pins down to the pads and finish their solder connections by hand. Powered on the board and everything worked… Whew! I was glad my instincts were true and it was indeed the old voltage regulator. To help prevent what happened as best I could, I bought the linear voltage regulators to feed it the 5V power instead of using the switching regulator. It is my guess that the noise on the switching regulator is what caused the other one to fail. I suppose we will see as time goes on. I bought 5 of the voltage regulators to have spares in case this happens again.
I began by testing everything, but again both servos were not turning and the power LED on the board was flashing ever so slightly when a servo tried to turn. After testing and re-soldering all my connections looking for shorts or some reason (including removing the power switch and hard wiring), I tried my ESC with 11.1V battery and had no problems turning the servos and keeping the board powered. It would appear these super cheap 5,000 mah batteries are not all they seem. Looked up and sure enough plenty of other people indicating trouble with them and that they lack the output for other projects. I might be able to run 4, 2 series batteries in parallel, to make sure it gets the power needed, but that also weighs quite a bit. I bought 2 more batteries from turnigy, same company as my 11.1 battery. All made for RC applications and good quality. The new batteries are also rated the same as the 11.1 in that they are 30C batteries and can put out the amperage needed. So far, I only need a little over 1 amp of power to turn both servos and power the board w/ accessories. Certainly no more than 1.5 amps. That is less juice than the RC tankbot I put together. That said, the tankbot is near the edge for what it can put out power wise as it had similar issues. In the mean time while I wait for the new batteries, I can try to put the rest together. Getting closer to a launch date. Need to decide how I want to launch the plane and ensure plane balance.
Launching with a catapult would be fun, but I don’t know if it is doable with this foam plane.
Aug 25 2015 – While waiting for the batteries, I did indeed try 4 batteries, 2 serial and in parallel. It made no difference. I tried having them go to different voltage regulators, but initially this did not help. I added a larger 470uf capacitor on the power rails going to the microcontroller and everything worked. I will try to see if I can filter the servo’s more with multiple capacitors, but I think my general topology has been to blame. I initially had the battery inputs go to the voltage regulators, to the servos, and off those rails to the microcontroller. In hind sight, after the voltage regulator, I need to run separate lines for the microcontroller and the servos. In addition, the large capacitor to the microcontroller power line as well as coupling that line to ground with a tiny capacitor may help with the power ripples. All of this appears to stem from the fact that servos, and motors in general are very noisy on the power lines. I didn’t think it would be this bad, but we learn something new everyday. In hind sight it makes sense. When I was testing power connections, I tried the ESC and bigger battery to test and everything ran fine. At the time, it didn’t dawn on me that with that setup, its entirely different topologically, with power going into the board and then going on to the servo board or with the board being powered separately entirely.
I also did a glide test without the microcontroller. The servos, without power, were able to maintain where they were without issue and the glider still stayed in the air. Not as much as it now weighs at least double what it did before, but with three test throws and one a little on the hard side, the components and foam seemed to hold up. Next time will have the microcontroller on it for real flight. I may see about wrapping the electronics in plastic wrap to protect from water in the grass and make sure nothing pops out.
The new batteries showed up, but at the moment I am unsure how I want to connect them. They weigh less then 1 of the 18650 batteries I’m using, so there is a lot of savings there all on it’s own. They also have much less capacity, but that is ok as they are rated to put out the power needed by a respectable company. I need to order more Dean connectors so I can connect them easily to the ESC I have.
Nov 7 2015 – A lot of things have happened without report. Installed one of the new batteries forward about ¼ back from the nose of the plane. Power lines run back to the perfboard which has lines to the microcontroller. After installation and testing, it was found that the new battery performed the same as the other 18650 cells. Something else must be the issue. Looked at the circuit for my power board/servo control and tried to isolate the microcontroller more. Looked up similar issues online and indeed more filtering of the power source is needed. Might still get some of those ferrite beads on order to help the power flow, but I have a large enough capacitor now to keep the microcontroller running.
Ran a flight test and it crashed horribly snapping the microcontroller off from the plane requiring the screws to be pulled out of the plane and new ones put in with a method of replacing the microcontroller board without tearing into the plane.
Jan 13 2016 – Holiday time takes up a lot of time and didn’t have much energy for physical work. The new mounts on the back of the plane are in. Now I just need to finish fashioning a board for the microcontroller to sit on. Will see about getting this done within the next few days as things should be back to normal a bit.
July 19 2016 – Going to start working on this project again. Was debating making a new controller, but I think I will stick with what I originally made. I will eventually want a new controller with a display, but this one will do for now. Need to shield the RF module and solder on the antenna extension with both the transmitter and receiver.
July 25 2016 – Got the new deck for the Parallax Propeller mounted as well as the Propeller itself. Remounted to power and servo control breadboard and plugged everything in. Reloaded the program to the microcontroller and everything was working nicely in the bedroom. Took it outside to try it and it appears between release and landing it is losing connection. Oy-vie. While the plane is still it shows good connection, but as soon as it is moving it losses connection. I will need to follow through on the modifications to shield the wireless modules and increase its antenna size. Outside there are more signals competing with each other in the same 2.4Ghz band so I will need to make its signal stronger. Maybe once the antenna is lengthened, it can wrap around the body of the plane to increase reception coverage. If it can’t latch on while the plane is moderately moving, that will be a serious damper. On the plus side, it works well enough with a slow moving tankbot so this shouldn’t be hard to fix.
July 29 2016 – Got everything reconnected and went for a test flight. Ran into some issues where the controller would be unresponsive, or the servos were not handling the load. These are micro servos and not the first time I’ve seen them be unresponsive. Plane weighs in at about 3-4 times what it used to weigh which will put added stress on the servos. I may need to look into pulley system if I want to continue using these micro servos.
The other issue with unresponsiveness is that the controller seemed to be momentarily losing connection while it was in flight. I did make the modifications on the receiving side (plane) but it had minimal results that I could see. Talking with one of my coworkers, he said the modification would be more important on the transmitter than the receiver among other warnings about standing waves and issues that can fry the transceiver. I will do what I can, but I need more power Scotty!
During the last test flight the plane had stopped responding while about 6 ft high and nose dived to the left. It crumpled into 3 pieces. Breaks were right before the main wings and right after. I am in the process of putting pins in and gluing it back together … of course making it heavier. Using bigger servos would also make it heavier. Still thoughts on adding a propeller, but I really want it to get the glide to work first before adding propulsion. I was playing with the fragments before gluing them back together and the middle section with the large wings would glide well on it’s own, but having such a shallow depth would make it more susceptible to flipping. I can see where the idea for a flying wing could have come from. Would be a neat project at another time. Wondering if I should add the IMU so it can take control when the signal is lost to level out the plane in an effort to avoid these sorts of crashes, but if the problem is with the servos, then that would not help. May add lights so I can fly it at night and be able to properly see what the wings are doing and if the connection is being maintained. I’m also still considering a whole new controller than the one based on the BS2 so I can get faster timings.
Aug 1 2016 – Over the weekend it came to me that another source of the wings not moving in flight could be friction. There is quite a bit of play in the wings forward and back movement. When moving forward the wings would flex back and rub foam on foam. With the weaker nature of the servo, it is likely that it is binding some and causing issues. I have coated all surfaces with tape to see if that caused the issue. Another possible issue, or combining issue is that the tape that was on the wings would dip some and stick to the servo providing resistance while trying to turn. I’ve added extra tape in locations where that was an issue. Between these two fixes, it may be enough. In the process of repairing the plane from it’s breaks, I’ve added reinforcement to the back wing that was partially broken. This will add to the weight, but increase the strength as well.
Aug 14 2016 – Plane flew again a few nights again and actually flew. Being much heavier, it does not float like it used to. I have another unmodified plane and the modified one weighs in at 3-5 times heavier. I should put it on a scale to know for sure. Various test flights started out with left flight. Then it went to banking left and right. This was done at night so the connection light could be seen. There was some loss of connection, but not anywhere near constant. Even got bold an launched from on top of a hill where it glided down more or less gracefully. Too much loss of airspeed caused control issues. Still considering putting the propeller on the front. I want to fix the communication issues first and the work on a new controller has been started.
While posting this, I see I said in the beginning this should be a relatively short project. After going on for a year, I’m not sure of that now hehe.
Knight Tech 