Mars Lander Rocket Quadcopter Project

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I’m not familiar with non-toy quadcopters, but is there a way to invert the problem? To deliberately windmill the blades so they are all doing the same thing? Like a sub-G throttle level pre-launch? I’m guessing the throw mode nixes that.
I’m not familiar with non-toy quadcopters, but is there a way to invert the problem? To deliberately windmill the blades so they are all doing the same thing? Like a sub-G throttle level pre-launch? I’m guessing the throw mode nixes that.

Well, for this project I want for it to take off on 100% rocket power. The unpowered electric motors may have slightly different friction at which a prop finally begins to windmill, so one will tend to windmill before the others.

I am considering again folding props. There are no commercial ones in CW and CCW matching pairs. But I did have an offer for 3D printed props, just I've been leery of reliability. But with 2 of 4 boosts crashing, and the "simple reliable" rubber band system not reliable, I need to try something else.

Option 3 is motor brake via the ESC's. But I have not been able to get the USB connection between model and my computer to work with the BL Heli's Configurator software to even try prop brake. USB works for connecting to Mission Planner.
Last edited:
I checked the model over. More damage, an arm pried off from the 1/8" plywood base, but it can be reglued. This photo gives a better idea of some of the damage to the 3D printed descent stage, and there is still dirt in the end of the arm that broke where a motor was.


I turned the model on and did a servo test. With the hassles of using the bigger digital servo in place of a smaller analog, I had to reorient the servo in a different way. It seems that opened up a failure mode for the "V" music wire to behave differently on one side than the other and not retract as much as the other side. So all that tedious time trying to solve the problem of why analog servos were not working, forcing a bigger digital servo to fit, took so much time that I thought I'd solved it but had not done the degree of bench testing that I should have.

At least with the need for repairs, I also have time to ask on an Arducopter forum how to solve this analog servo problem so I can go back to the servo and mechanical set-up I had before. But I am also going to explore 3D printed folding props.

A test I did by adding a NASA logo under the ML-5 number, on the crashed ascent stage. I think I'm going to do this on the next one. Because it also kinda looks like "50", representing 50 years (I chose "5" to represent 5 decades). I won't put any other NASA logos on it, other than under the two ML-5 locations.

BTW - the version 3 Ascent stage has eight holes near the bottom for a practical reason. Air vents. In real rockets, any unpressurized area has to be vented or air pressure differential could burst the structure on ascent (or collapse on descent). Even the shuttle orbiter had air vents (rectangular vent doors on the fuselage sides). But for this model, it really needs some airflow to help keep the 4 in 1 ESC from overheating on a hot sunny day, hopefully this is enough (The ESC board is about at that level). So neat to ask Mech-G "hey, would you mind doing XYZ on the next 3D print version?" and boom it's done. As well as some neat ideas of his own. :)

Update: Maybe I subliminally remembered this, or seeing something like it long ago:

Last edited:
Repaired the model well enough to fly as a Quadcopter late today. Had to fix a broken arm, where the motor snapped off, and reattach broken wires:


Found that rather than remove the broken arm, I was able to cut it square, use a telescoping piece of graphite tubing as a coupler to join the shortened arm to a replacement extension (fortunately the 3 wires fit thru).

Had to do a few other things, check out motor rotation direction and fix as needed and other odds and ends. I did not attempt to replace the servo or add the onboard camera, which would have taken too long to do before dark (and I still have not gotten around to the servo problem). It'll be awhile before I do rocket boosts again.

Also for expediency, I used the old Version 2 Descent stage, which has a bit of damage but is flyable as a Quad. I put the damaged decaled Ascent Stage shroud on it, the damage is not very noticeable from the front.

Powered it up, and it flew fine! Had a few good minutes flying it in very low wind. Then I took it up higher and tried "Loiter" mode. And Loiter worked. Hmm, Loiter was not working Sunday when it had the camera/transmitter, digital servo, and bluetooth altimeter onboard, so one or more of those must affect the compass. After the battery got low and I landed, I not only swapped batteries, but got my digital camera out so I could take some close photos of it flying while loitering in a hover. Got some nice pics. I'll have to do that again later when I replace the damaged parts and have all the decals on it (maybe with the moon). I shot some video too, but won't post that till Thursday. Also, I activated RTL/RTH, and it automatically flew back over where it had been powered up, and landed by itself vertically about a meter or so where it had taken off from.



"I see a silhouetto of a Mars Lander
Scaramouch, Scaramouch will you do the Fandango"

Here's video. Includes an automated RTH/RTL landing.
(note, this is uploading as I post it, won't be available till around 1:20-1:30 AM EDT)
Last edited:
Well, since Loiter worked well Wednesday with the Mars Lander, and again Thursday, I did a one-man-band thing (or Plate-Spinner-thing) before dark.
With, the LANDERS.


The LM drifted very slowly with the very light wind. So, I had to go back and forth between taking photos, then flying the LM back, and taking a few more photos. One time, I flew the LM right over the ML by about 5-6 feet, and the downwash of the LM caused the ML to descend a bit and be a bit unsettled, till I backed the LM away. From then on, I made sure the LM flew towards one side or another.


Hat-mounted GoPro camera quit on me early, poorly aimed and somehow failed to save the MP4. It did save a low-res LRV file. So I converted that. So, not the best, but so far the only video of both models flying together. This was mainly a photo shoot anyway. And added to the end a clip of the LM landing next to the ML.

At some point in the future, when I have an undamaged 3D printed body on the Mars Lander with all the decals, I’ll try to do another photo session like this with both, and try to get some good video.
This is a really terrific project!
I'm going to reconsider other options for solving the prop windmilling.
Is one of the options to just let them?

Do you have a way to short the leads on each motor? That won't stop the rotors completely but it should slow them down.

For extra credit, after the launch-to-loiter system is worked out. you should add an 18 mm motor mount to the ascent module with a blast protector over the delicate stuff inside. Launch-loiter the whole, then launch the ascender from the loitering platform, to be recovered on a parachute.
Last edited:
This is a really terrific project!Is one of the options to just let them?

Do you have a way to short the leads on each motor? That won't stop the rotors completely but it should slow them down.

For extra credit, after the launch-to-loiter system is worked out. you should add an 18 mm motor mount to the ascent module with a blast protector over the delicate stuff inside. Launch-loiter the whole, then launch the ascender from the loitering platform, to be recovered on a parachute.

I'll let someone else with a Mars Lander Quadcopter tackle a rocket boosted Ascent stage during hovering flight. Also if you read from the early messages, and look at the Quadcopter part inside, you'd see there's no room for anything inside like that. Not even room for an emergency chute eject. I ended up moving some things down into the Descent Stage due lack of space in the Ascent stage or GPS/compass interference, like the Receiver and Low Voltage alarm.

The boost issue with windmilling (spinning on their own due to airflow) is when one prop starts to windmill before the others, causing a lot of drag to pull the model to one side. The three possible fixes are mechanical stops (the servo-rubber band release), folding props, or possibly "Prop brake" by the ESC's (Electronioc Speed Control). I have not been able to get the ESC programming software to connect via USB, between my laptop and the Flight Controller.

Folding props, no commercial ones that are the right size and also rotate CW and CCW. But Jamie Claye has taken a 3D printed folding prop design, and mirrored it, so there can be CW and CCW folding props. I'm going to try them out. But small model props are quite sensitive to things like balance and surface finish (3D prints are not smooth unless you spend a LOT of money on special print technology), the folding props might be too inefficient or vibrate too much. But I'm checking into them.
Last edited:
First, some fun stuff. Last Sunday, after flying some contest rockets, I flew the Mars Lander as a Quadcopter, as Ray King used my camera to take photos and video. And he got this great shot:

And some video. Ray followed it, so there is some better video of the model flying at far distances than my other video have shown. Also a good idea of what it is like when I pitch the nose over to do a high-speed horizontal pass.

And now the not so good news. Thursday, the Mars Lander had a Cato. Catastrophic failure. But, no, it did not get burned, or blowed up. It did not even have a rocket motor in it. Jamie Claye brought up the possibility of using some 3D printed props that rotated CW and CCW. He took an existing folding prop design and modified it, for a mirror-image. Ray King (who lives about an hour from me) brought to that Sunday launch, some sets of folding props he made on his 3D printer, using Jamie's file. But there were some issues, the tips did not print well enough. A complication of the process. But I got an idea of how they worked out physically, particularly when mounted to a motor.

Thursday, I got a set of props that Jamie Claye made. The tips did print better. So I decided to try to do a Quadcopter flight to test them out.

Throttled up, building up RPM and thrust for takeoff, and a very brief weird noise occurred….. and I realized it had lost a prop blade, or more (and never left the ground). Stopped it, and saw no blades, it broke all 8 of them.

Pretty much at the 45 degree pivot part of the blade root, the airfoil cross-section is visible on what’s left. I did not find any of the broken blades, they went out into the grass.

But two of them hit the Descent Stage and added more damage. But that’s OK, it’s the version-2 stage that crashed a few weeks ago, I wasn’t going to risk any new parts.

I had a GoPro going but it quit on me, no video.

Even before flying, I had wondered a couple of things. The blade area is not as wide overall as the props I have used, less prop area.

Even with a bit more span, the reduced area would mean the props would have to spin faster to try to produce equal thrust to the fixed props. Also, the irregularities in the surface would cause various aerodynamic issues, partly drag, and also loss of lift, so needing even more RPM to equal the fixed props (therefore even more stress on the props).

I’m not sure where to go from here. Using stronger material itself, won’t solve enough of the problems. Needs a totally revised prop with more area, like the ones I have used, or “Bull-nosed” props. Ironically, that whole 45 degree thing, is not needed for this model. That 45 degree angle portion allows the folded prop to be “wider” when folded, to allow for the width of a plane’s fuselage up front. And that is the weakest link, that angled section. If it was straight across, that would help. But the bullet nut for the prop won't let it fold back 90 degrees. However, making the hub a bit wider (spacing the holes out farther) to account for the bullet nut could do it.

But given everything, I do not know if 3D folding props can end up being something viable that I can trust.

So, it may be back to using rubber bands to hold the props till a servo releases them, the original set-up, not the different set-up with that bigger digital servo.
Last edited:
If you should continue to have difficulty with the servo actuators releasing the props, how about using igniters to "cut" them?
One way or another, this is going to be INTERESTING !

With the failure of that first set of folding propellers, I had to try my hand at making it work. The breakage occurred at the ~45 degree bend in the blades. That bend was apparently needed for the original application which required the blades to fold around a spinner cone.

George's motors have a small enough cone/prop nut that the bend is not needed. Using Fusion360, I designed a 5x3 blade and a 4x3 blade that would fit onto a hub similar to Jamie Claye's original hub. I added a stop that would keep the blades from going past horizontal when unfolded. I made the root of the blade of the right thickness that it forms a vertical stop against the prop nut to keep the blade from going past vertical when folded.

I printed a prop in white PLA to see if it was possible to get a smooth enough surface finish using a 0.3mm nozzle and 0.1mm layer height on my Artemis printer.

upload_2019-6-18_10-50-51.png upload_2019-6-18_10-51-15.png

The blade is printed with the leading edge down on the print bed, so there is no need for support material. This makes for a very clean print with only a little cleanup needed.

The prop looked good, but I was concerned about the PLA being tough enough, and also withstanding the summer heat without warping, so I printed a test propeller inblack nGen Amphora filament. nGen is more heat resistant, and somewhat tougher than PLA, yet stiffer than PETG.

To make sure the props would not do a RUD on us, I built a "spin cage" test setup, shown here.


The motor is bolted to a simple tower that I 3D-printed in PLA. Four printed clips hold a cylinder, made of sheet steel, up about an inch above the surface to allow air flow in from from underneath.

A servo tester and 20-amp ESC allowed me to test the props at the full range of the motor speed. The steel cylinder would hopefully protect me from flying debris if the prop flew apart.

I tested the props at 10.5 volts, full throttle for 5 minutes. I also gave them 25 deployment cycles, starting with blades folded then going to half throttle, to simulate what will happen during transition from rocket boost to quad hover. I have a Turnigy Tachometer, but I have not yet been able to get the lighting just right so that it will tell me the RPM of this particular setup.

A check with micrometer showed the blades had not suffered any deformation from the tests, so hopefully George won't have any more surprises when he tries these on the Mars Lander.

I printed a set of the 5x3's with two-blade hubs and a set of 4x3's with three-blade hubs for George to try out.


Thanks so much for all the work on the folding props! Quite a lot of testing involved. Jamie Claye has also been working up improved folding props, without that troublesome 45 degree bend, and his improved ones have survived full throttle tests too.

I look forward to testing out the props you're sending, and if the Quad-only tests go well, i'll probably do more rocket boosts this weekend (MASA has a launch scheduled for Saturday, sorta windy, worrisome chance of rain).

So, here's a post I've written about some recent flying I've done with totally different props, but they don't fold.

Went to a great local hobby shop to get some supplies Saturday (well, it’s a half hour away but worth the drive). I wanted to get some spare 2-blade props to replace ones I had broken. And saw some 3 bladed “Race and Freestyle” props, in funky translucent colors like neon orange and purple, and….. clear. Hmmm. The props were 5.1 x 5.0, so 5” pitch (I have been using 2-bladed HQ Props at 5" dia x 3" pitch). I was curious to see what clear props might look like in flight, and also curious to try such a radical pitch. So, I got a set (pack of 4, 2 each direction). They have a funky shape, sort of boomerang or dogleg blades, a real mess to try to 3D print. They are also somewhat flexible.

I flew the ML with them Saturday afternoon, late Monday, and late Tuesday. Could not tell much difference. if I had a way of determining current draw, I would expect the current draw would be increased due to the greater pitch and third blade for the same span. May be a bit quieter. I do think that on descent there may be a vortex ring state issue occurring, I heard a sound I did not recall hearing before, only happens when I go to maximum descent rate. But it did not behave badly, just hints there may be a vortex ring issue lurking. I had suspected the higher pitch (67% increase) might make the model susceptible to that.

GoPro camera not working right, so I used my Canon digital camera, with neckstrap, and let the camera hang off my neck facing forward for some of the video. Once I found that Loiter mode was working, I used the camera handheld more. But also used the transmitter some. I was able at times to control throttle and roll (left stick) and still point the camera. But pitch and yaw piloting (right stick) required letting go of the camera. Except when I was laying on the ground.....weird way to fly and trusting Loiter (to a safe extent)
The model is still beat-up, I'll add a new 3D printed body set with decals later. Anyway, the video I shot has a lot of close-ups.

And a unique view from directly underneath.

The landing was...... beyond the margin.
An example of a blade printed in the nGen material.

First Layer height = .2mm
Layer height = .1mm
Nozzle size = .3mm
Temp = 230C
Print surface Temp = 75C
External Perimeter Extrusion width = .45
Internal Perimeter Extrusion width = .4
Speed = 20 mm/s


The curved bump is where an internal perimeter ends as the blade gets thinner toward the tip.

Rectangular pads help with adhesion to print surface. Glue stick also helps.


  • upload_2019-6-19_9-31-56.png
    133.3 KB · Views: 24
Have you considered using carbon blended PLA? It looks like there might be a little room to increase the fillet size on the prop too. This is a phenomenally cool project BTW!
Carbon blends are definitely an option. Also, there is glass fiber-filled nylon, which some commercial drone props are made from. And ABS, of course.

I don't use ABS a lot, because of the odor it generates when printing. There does appear to be low-odor ABS available now, though.

If there are any issues with the nGen material, I will probably try the glassed nylon.

It is an interesting engineering tradeoff, since there is an advantage to using well-behaved materials that result in a good surface finish, but those are not necessarily the strongest or toughest materials.

As for the fillets, they are already pretty aggressive, though you can't tell that from the photo. But you are right, there is room to go bigger.
Last edited:
I recently got some folding prop blades from Mech-G. The blades tested out nicely. A test session (one battery) with three takeoffs, the blades deployed each time. Last test, one blade did not deploy at LOW start-up RPM's, and the model shifted on the pavement. When I went to a bit more throttle before takeoff, it deployed (centrifugal force).
The sound with the folding props a bit unusual!

So, Saturday at the MASA launch I tried another rocket boost, first boost trying the folding props. Took off.... and immediately pitched hard and gyrated under thrust till it hit the ground. Damage worse than last time, wiped out the Descent stage.

The cause was stupid. I was using a Firestar ignitor that I usually secure with scotch tape around the extended nozzle of the motor. I could not find the scotch tape, so used some masking tape. A previous attempt to launch lost continuity, the clips were not grabbing tight enough, so I made sure they were holding tight. At liftoff, the clips stayed attached. The ignitor normally rips thru scotch tape, but did not rip thru the masking tape. So, the model took the clips with it. As it got higher, the launch wires yanked the tail of the model toward horizontal, then let go once it was rotating in pitch. The model sort of tumbled and also rolled and hit the ground.
Video sequences include handheld by Art Gibbons, ground looking-up camera, and onboard camera looking out the side to show a folding prop that never got a chance to deploy. In the screengrab below, you can see the ignition wires in mid-air at lower right.

Video link:
In the video above, the beginning has some of Friday's testing of the folding props. And after the crash some footage flying my 1/16 Lunar Module Quad. Some scenes during launching of models (it was not too close). Wind was also gusting strong for awhile. Most of that LM video was with the camera hanging from a strap around my neck, camera on my torso. I landed it next to the Moon Target for the "Land the Eagle" contest but somehow didn't count as it didn't use parachute or streamer recovery. And....wasn't rocket boosted. Pesky details!

Flew some other models for fun (not on video). My Star Spangled D Bird (two D12's, built in 1984. NRC kitted it for G power). Also an A3 powered R/C Rocket Glider, and a "Spider Bowl" converted into a flying saucer

Back to the Mars Lander. It will take awhile to fix. Also I need to get some more spare parts. I have to shift more towards preparing for NARAM. So, most likely after I fix it, and test fly enough to be sure it is flying well as a quad, I will paint and decal it to look nice. And not try a rocket boost again until NARAM next month. Because if it crashes AGAIN on a rocket boost attempt, I won't have time to fix it one more time before NARAM.

The folding props made by Mech-G worked well. And Jamie Claye has also been working on folding props. Jamie was the one who convinced me to finally try 3D printed props (no commercially made sets of matching CW and CCW props available). I DO think the next time I try a rocket boost, it is going to go very well. Of course I also felt that on Saturday morning too. :(
Last edited:
I finally began the repair to the Mars Lander. After unbolting the "Tower" assembly and disconnecting wiring, here's what the Quad main structure looked like, with the 4-in-1 ESC board attached.

Oh.... crap. Notice the DARK discoloration on the ESC board, the corner on the left side, where I'd put some masking tape over it?

Yep, it's fried. So, gotta order another one. In the crash, the "tower" assembly came loose and may have crunched into this board. Or else wiring from a motor got shorted, but I could not find signs of that.


Otherwise, the repairs to the structure were not too hard. Gluing one arm back on. The other arm broke the graphite tube, near the motor. I used the same method I used before, cut the arm for a clean 90 degree edge, and glued a short piece of the same tubing to it using smaller diameter graphite tubing as a coupler, and epoxied a new motor mount in place at the end. Still need to replace three bullet connectors I had to cut off to get the wiring to tun thru the smaller diameter coupler. After getting a new ESC, it'll take at least a night to reinstall everything and go thru the Quad set-up and make sure the motors turn the right directions.

And I'm not counting replacing busted 3D parts for the Mars Lander itself, painting, decals, and installing the Quad into the Descent stage.

Once it is back in flying shape, I'll test fly it as a quad to be sure it's OK. And take it to NARAM. I won't try another rocket boost until NARAM. A bit bummed not to try it at the July 20th MASA launch, but just no time to repair it again for NARAM if it crashed. But I've got something more suited to fly on the 20th (Lunar Module), and can rocket boost the ML Quad at the August MASA launch if it makes it thru NARAM.
Thinking outside the box. Maybe cone headed.
Could you put paper cones on the rotors for boost phase that would be held on by gravity and then inertia and airflow on boost, the just fall off and flutter down at apogee? Might not work on a windy day especially at a big event where it has to sit on the pad for a while
Finally got the Mars Lander fixed and flying again after the June 22 boost crash.

Did a test flight with 3-blade fixed props to check out that it flew OK. Will put the folding props back on it for rocket boosting.

Won't fly it again until I get to NARAM. Probably fly it as a Quadcopter Sunday, possibly a rocket boost Sunday too (I'll hang out at the US Team flyoffs/World Cup area Sunday thru Tuesday, then the NARAM contest range the rest of the week). Plan to do several rocket boosts thru the week, if it doesn't crash again.

Last edited:
An overdue update about the Mars Lander Quad at NARAM. I got it fixed from the late June boost crash. Put decals on at the field Sunday, wasn’t going to fly anything else that day anyway. First time I added decals to the Descent Stage.

I flew it at the Contest Range, during a break in flying at the US team selection flyoffs and North Coast Cup. It boosted great on an E15, best flight ever (This was boost #6). Automatically went into hover mode near apogee. Flew it around a bit and landed safely. Only later did I realize there was no video footage taken of the flight, first boost without any. Monday was massively windy so I didn't fly it.

Tuesday, flew it from the Sport Range, much more appreciated over there. It boosted nicely again (Boost #7), and “Balsa Bill” of Balsa Machining Service used my camera to get video. I did a pass by the HPR pads and spent a bit of time at low altitude before landing. Went back over to the Contest Range, doing the last day of the flyoffs/Cup, and flew it again (Boost #8). But it pitched over during liftoff and ended up crashing.

This time I do not have an obvious cause, micro clips were secured, has folding props, and so on. I suspect two things. One, that it is marginally stable, as the real model is. I had expected for the 1300 mAh battery to provide plenty of noseweight but on rechecking the CG is not as far forward as I had planned for a good margin. Also, there may be a bit of lateral mass imbalance. I need to come up with a good way to set the model up vertically in a manner where it can tilt towards the “heavy” side, and add a bit of clay inside the nose of an electric motor mount to balance it out laterally, like adding balance weights to a car tire.

I am also going to literally add noseweight. The hollow nose cap is easily removed, so I will fill it with clay. I expect the clay will not affect GPS reception, or the compass magnetometer, on the GPS module directly below the nose cap.

Finally, I have always theorized that the real ML sometimes goes unstable due to the landing feet being angled the way they are, rather than being flat. If the model happens to pitch or yaw a bit into the airflow, one foot is flatter to cause more drag and the other foot is steeper to cause less drag, then it can build onto itself from there.

So, I am going to modify the feet to be FLAT, like the Lunar Module feet, no longer angled. Makes t a bit less accurate, but worth it to improve the stability.

Here is a link to video of boosts 7 and 8.

A really crappy thing about NARAM-61 was one person’s choice to BAN any electric or I.C. motor powered R/C model from NARAM, from both of the two launch ranges. This was NOT a safety issue, not an RSO issue, not an NAR issue, and not an AMA issue - he admitted to me it wasn’t any of those. He invented it himself. Because he simply had the power to do so. So, I was not allowed to fly it as a multicopter only. As I had at NARAM-59 with my Lunar Module quadcopter. And as several people including Chris Taylor have done at NARAMs in recent years. NAR has allowed multicopters to be flown at NARAMs, NSL, and TARC, and plenty of NAR launches. It was shocking, frustrating, and just plain “ticked off” to run into that garbage by one person’s personal choice to be like that. So, I was only allowed to fly it when it was rocket boosted.
This is totally a bookmark post so I can track progress on one of the most awesome projects I've seen in rocketry for quite some time! A wonderful "small field" solution!
For those who may be interested, I'm posting the PDF file for the custom decals I made for the Mars Lander Quad.


It is sized to work with the original kit, and Mech-G's 3D printed model. As-printed, the outer size reference rectangle edges should be 8" wide and 5.125" tall. If not, tweak your printer settings (I always test print in normal paper before using decal paper).

I designed it to fit on Testor's 8.5 x 5.5 decal paper for inkjet printing.

Of course, the big red stripe with USA in white, is not part of the original kit. I did that as a roll pattern to show the "front" when flown as a Quadcopter. Instead, the second red USA text decal (lowest right) would go there.
Last edited: