I could use just a little guidance

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I must say your patience with this project is admirable. I've seen folks get snippy because they didn't have time to launch their 7th rocket for the day....

Well, on the high altitude flights, there isn't much choice. I can go to Balls or to Balls. But I would like to fly the stabilization system more often. Maybe I'll work on that. Let me see, I have a spare set of servos, several spare UDB5's, need a spin can and a new rocket.... OK, let's do it.

Jim
 
I might be interested in on of your spare UDB5s if your looking to get rid of one. Also I would be happy to build you one of my stabilization bays for you to fly if your interested. The servos I'm using are aluminum cased with double ball bearings, my rocket fell flat on the canards onto gravel from 3500 feet bounced 5 feet up and the servos and bay had no damage at all. One canard had some cracks in the fillets. I'd like to get more high velocity testing of my mechanics, and plan to make a unit for Gary Kawabata, but the lack of availability of the UDB5 put that on hold. I still haven't found another option that has all the features I'm hoping for, I want to add a barometer or high G single axis accelerometer to calculate rough velocity and allow for adjusting the servo travel accordingly.
 
Jim, did you do an autopsy on the dead servo? Curious to know why it died if you did.

Yeah, I think I know. When I took the servos out of their mounting, I discovered that three of the cases were broken. It was not possible to see this without taking the system apart and inspecting the cases very closely. Two of the servos just had cracks along the top of the case. However, on the servo that caused the problem, it looks like the failure would have allowed the spline to be pulled upward, perhaps disengaging the gears. In the first pic below, I have broken off part of the case, but you can see where the case was broken. In the video, it appears to me that the canard rises up a bit and flutters just before the failure. In the second pic, the canards are at similar angles, but the one on the right has more of a gap under it.

When I tried the servo after the flight, I could move it roughly 10 degrees one way and then the other. Not good. Then, I tried it again a few days ago, and it appeared to work just fine, allowing the usual small amount of motion. I think it had just re-engaged at that point. That's when I looked more closely and found the failure. I suppose this could have happened during the flight, but I suspect it was from a hard landing on a previous flight. So, my advice on this is to inspect the servos closely before each flight, even if you have to take things apart, and don't assume that they are OK just because they appear to be working. Lesson learned.

Jim

IMG_1341.jpg

Canard 2.jpg
 
Jim,
First of all, let me applaud your tenacity for pursuing such a complex problem. I know you have a pretty tight design space to work in, but here's a few things that may spark some ideas for next time around:

Those Hitec servos are great for R/C but they simply aren't made for the kind of loads you're imposing on them with a direct drive control surface. I can't quite tell what your new servo is but it looks like the 7955(?). Great servo, but again I think the lateral loads on the case pose some risk.

The drag of your control fins puts a bending moment into the shaft of your servo. The bearings in the servo are just pressed into plastic and are not designed to take out of plane bending moments. The plastic case is not designed for out of plane bending moments. So with increasing velocity and increasing drag on the fin, and things start to flex. The fin turns on servo command, more drag, more flex. I think you may have some assembly tolerance, case flexure, and internal tolerances that are building up to the gear train possibly binding, which will further exacerbate the problem as velocity goes up. I haven't put numbers to this, but I think it's a plausible hypothesis that is worth exploring. I believe an all metal servo mounted to a metal platform would greatly increase your chances of success.

Here's an example of some metal cased servos in a similar sized package that may handle lateral loads a little better than a plastic case, in addition to being a much stiffer platform: https://www.mksservosusa.com/home.php?cat=24. I've never used MKS servos before but they sure look the business. Bonus - high voltage servos can run on a 2S lipo with no regulator.

Another option is to mount the servo in a block and decouple the external forces from the servo drivetrain completely. https://www.servocity.com/servos/servoblocks

It may be that your airframe bearing does some of what the above servo block would do, but I haven't seen enough detail of your design to make that out. I'd really love to see this thing in person next time you come to Kansas.

Servo City has quite a lot of hobby robotic stuff that might be useful in your endeavors. https://www.servocity.com/motion-components/rotary-motion

I'm not affiliated with Servo City in any way, but I am a very happy repeat customer and they're located in Winfield KS not far from the Kloudbusters Rocket Pasture.
 
Thanks for your suggestions guys. I appreciate it.

They are the 7954 servos. They are from ServoCity (I like them too). I will probably operate these on 2S.

The air frame bearings are a critical part of handling the lateral loads. The tolerances to the servo hub are very tight (and giving me fits fitting the replacements). But, assuming I can do that, I don't think the lateral flight stresses will be a problem at all. Landing forces are something else though. The servo blocks would be much better there, but there is not enough room to use them.

A few pics are attached showing how the servo, servo hub and air frame bearings are configured.

Jim

IMG_1344.jpg

IMG_1348.jpg

IMG_1340.jpg
 
The air frame bearings are a critical part of handling the lateral loads.
Agreed!


You're on the right track with the 7954s. Should be a good servo if the case holds up. I'd still look to stiffen up the mounting platform as much as possible though. If budget allows, consider drawing one up and have it 3D printed - for fit check only! - then have one machined out of aluminum once you have the parts arranged.
 
Agreed!


You're on the right track with the 7954s. Should be a good servo if the case holds up. I'd still look to stiffen up the mounting platform as much as possible though. If budget allows, consider drawing one up and have it 3D printed - for fit check only! - then have one machined out of aluminum once you have the parts arranged.

Thinking back on this, I think it is likely that the servo damage occurred before the air frame bearings were installed. Hard to believe I wouldn't have noticed it, but it is possible.

Actually, I think the air frame bearings do provide protection during landings because the base of the canards is almost flush with the top of the air frame bearing. In the current design (see previous pic) I have the cylinder sticking out about 1/8" from the bearing. I need to change that so that they are flush.

Jim

DSCF1046.jpg
 
You planning to bring that sucker to BNE or are you gonna make me come all the way to Melbourne to see it fly?

Sorry. Flying at Serpentine next year (April?). Come on down and visit us Mexicans :) (for anyone in the USA, that's a Queensland joke :wink:)


Yeah, I think I know. When I took the servos out of their mounting, I discovered that three of the cases were broken.

Thanks for the info Jim. I am limiting my angular excursion so should not have as much tension on the servo shaft. I am not expecting earth-shattering performance.
 
So, I have two main "stabilization" activities at present. First, I need to rebuild the servo holder to accomodate the larger, stronger servos. The servos I have been using have proven not to be sufficient for the application. Second, I'm building a new test rocket with forward canards so that I can fly the stabilization system more often than once or twice a year. That frequency is just not generating enough experience and data (or fun).

The servo holder is completed, although I need to purchase two additional servos and then permanently mount them. The assembly of this turned out to be very difficult. The problem was getting the servos lined up exactly with the existing bearings and using the existing mounting holes in the air frame. And with the larger servos, there was almost no space to work with. But, it's done and it should work fine.

The new test rocket will use forward canards (instead of the system being located at the bottom of the upper stage). When I started the stabilization project, I had a stabilization spool that I used for locating the stabilization system at the top of a rocket. I replaced that with a different spool for use with the staged flights, but I still have the upper spool. I'm in the process of installing the bearings in this upper spool (it didn't have them before), which just involves enlarging the air frame openings and then attaching the bearings. Once I get all four servos mounted in the servo holder, I will install it in the upper spool and then the bearings can be glued in exactly the right spot. Then, I can move the servo holder between the two stabilization spools.

The new rocket will also have a spin can, since I don't think that forward stabilization can work properly with fixed fins (due to the control reversal issue). The spin can is easy to make, but requires rolling an oversize tube for the fin can itself. The lower air frame for the test rocket will be a 4" coupler tube (rather than an air frame tube). Air frame tubing will serve as the bearing surface, and then the rolled tube goes over that air frame tubing. The trick is to roll the tube at just the right ID so that the ball bearings fit between the 4" coupler tube OD and the rolled tube ID. I calculate that five wraps of 0.005" mylar around an existing mandrel will be the correct size. The mylar will be formed around the mandrel as its own tube, with the mylar wraps glued together with 3M 72, so that it can be removed from the mandrel in one piece after the tube is rolled. I'll probably do this over the weekend. Hope it works....

Jim

IMG_1353.jpg

IMG_1354.jpg

IMG_1355.jpg

IMG_1352.jpg

IMG_1356.jpg
 
Have you considered 3D printing that part of the assembly?

The servo holder would be the perfect part to 3D print. In this particular case, though, I am retrofitting the part to my existing tube with the bearings, and the alignment of everything isn't 100% perfect. I did give a little thought to how I might "map" the exact positions of the existing servo splines. I know how to do that for plant equipment and pipes, but not for small things. Maybe it can be done.

Jim
 
The servo holder would be the perfect part to 3D print. In this particular case, though, I am retrofitting the part to my existing tube with the bearings, and the alignment of everything isn't 100% perfect. I did give a little thought to how I might "map" the exact positions of the existing servo splines. I know how to do that for plant equipment and pipes, but not for small things. Maybe it can be done.

Jim

If you need some 3D printing, let me know I am sure we can work something out if you want to go that route.
 
If you need some 3D printing, let me know I am sure we can work something out if you want to go that route.

Thanks Kyle. I can't use it for what I am doing right now because my existing air frame bearings aren't perfect, but if I was building it new, this is what I would do.

I did manage to get my spin can fin tube rolled this weekend. As usual, I hung a little weight on the carbon to tighten up the wrap. It will take a few days to see if I got the diameter right.

Jim

IMG_1358.jpg

IMG_1359.jpg

IMG_1360.jpg

IMG_1361.jpg
 
My spin can tube seems to have turned out OK, although the ID is a bit larger than planned. I used 5 wraps of 0.005 mylar on my 4.03" mandrel, which should have given an ID of 4.08". I got 4.09", which I guess would be predictable. That gives a gap of 0.095 to fill with bearings. The 5/64" bearings (0.078") might be a bit small, but 3/32 (0.093") bearings might be fine. I'll try the 5/64" first, and if things are too sloppy, I'll go with the 3/32 and sand to fit.

Jim

IMG_1363.jpg
 
I've made quite a bit of progress on my new spin can. Recall that this spin can will be for a new rocket to allow me to fly (more often) the vertical stabilization system.

I rolled an oversized tube that will be the actual fin can. This turned out pretty well. Then, I cut all of the tube sections that are required for the spin can, and squared them up using my laser approach (which I think works pretty well).

I ended up going with the larger 3/32" bearings. I may need to sand down the "coupler tube" air frame to get it to turn a little more freely, but it actually would work just fine as is.

https://youtu.be/o68ZcMVe_o8

This is just a test rocket for the vertical stabilization system, so I'm keeping things as simple as possible. I went with 3/16" fins, and I'm planning to just use a fillet for attaching the fins, and no tip to tip. I might change my mind on this, but at present, I don't anticipate flying this rocket above Mach 1.

Jim

IMG_1366.jpg

1369a.jpg

IMG_1375.jpg

IMG_1377.jpg

IMG_1381.jpg

IMG_1383.jpg
 
Last edited:
Well, on the high altitude flights, there isn't much choice. I can go to Balls or to Balls. But I would like to fly the stabilization system more often. Maybe I'll work on that. Let me see, I have a spare set of servos, several spare UDB5's, need a spin can and a new rocket.... OK, let's do it.

Jim

So, it looks like it was three months ago that I decided, a little spur-of-the-moment, to build a new stabilized rocket that I could fly more often than just at Balls. It was more difficult to do than I thought, but the rocket is done. A few pics of the rocket are attached. As previously discussed in this thread, the rocket has a revised stabilization section with much stronger servos. The last step here was to mount the support bearings for the servo axles. The rocket also has a spin can to avoid control reversal from the canards ahead of the fins.

The next step is a test flight, hopefully, over the next few weeks. I have a couple of L motors for flights between 6,000 and 8,500 feet (maximum velocities of 700 to 860 ft/s), so I need to choose one of those. I want to set the launch angle at somewhere between 5 and 10 degrees and then have the rocket go vertical. I need to engage the roll control right from the start of the flight, as roll does not help the gyros at all. But, I need to decide whether to activate the vertical stabilization from the start, or alternatively, a second or so after launch. Any thoughts on that?

Actually, I'm not quite done - I need to mount a camera. As much as I'd like to have the canards in the view, I'm not sure mounting a (key chain) camera around the joint between the cone and the stabilization module would be a good choice. Thoughts on that?

Can't wait to fly this again!

Jim

IMG_1404.jpg

IMG_1405.jpg

IMG_1406.jpg

IMG_1407.jpg

IMG_1410.jpg
 
So, it looks like it was three months ago that I decided, a little spur-of-the-moment, to build a new stabilized rocket that I could fly more often than just at Balls. It was more difficult to do than I thought, but the rocket is done. A few pics of the rocket are attached. As previously discussed in this thread, the rocket has a revised stabilization section with much stronger servos. The last step here was to mount the support bearings for the servo axles. The rocket also has a spin can to avoid control reversal from the canards ahead of the fins.

The next step is a test flight, hopefully, over the next few weeks. I have a couple of L motors for flights between 6,000 and 8,500 feet (maximum velocities of 700 to 860 ft/s), so I need to choose one of those. I want to set the launch angle at somewhere between 5 and 10 degrees and then have the rocket go vertical. I need to engage the roll control right from the start of the flight, as roll does not help the gyros at all. But, I need to decide whether to activate the vertical stabilization from the start, or alternatively, a second or so after launch. Any thoughts on that?

Actually, I'm not quite done - I need to mount a camera. As much as I'd like to have the canards in the view, I'm not sure mounting a (key chain) camera around the joint between the cone and the stabilization module would be a good choice. Thoughts on that?

Can't wait to fly this again!

Jim

Very nice. I noticed no T2T on these fins is that because it is sole purposed for relatively low velocity flights. On the debate as to when to engage the roll control, what do you see as the downside to having it engaged out of the gate?
 
Very nice. I noticed no T2T on these fins is that because it is sole purposed for relatively low velocity flights. On the debate as to when to engage the roll control, what do you see as the downside to having it engaged out of the gate?

I would fly this mainly at our local field, which has a 10K waiver. So, it would fly below Mach. I got thicker G10 for the fins to start with, so I decided for no tip to tip. I can always add it. The largest motor that can fit would be a small M (12.5K, 1,160 ft/s). I would probably add some reinforcement if I tried that flight.

Roll control would be engaged from the start, but the issue is when to activate vertical stabilization. If it is on the pad at an angle, the canards will be at an angle at launch. I don't think this is a problem - just seems a little odd.

Jim
 
Has your stabilization section always had the servo/canard pairs offset like that? I guess I never noticed before.

Will anything prevent you from mounting a camera looking forward? Perhaps balanced on the other side by one looking aft at Jim's-spinning-fin-thing?


Very nice. I noticed no T2T on these fins is that because it is sole purposed for relatively low velocity flights. On the debate as to when to engage the roll control, what do you see as the downside to having it engaged out of the gate?

Added control complexity of having to negate roll And alter trajectory?
 
Jim - I want to sincerely thank you for sharing your journey! I've pointed the IREC team I advise to this thread and the information contained within has been invaluable.

Edward
 
Has your stabilization section always had the servo/canard pairs offset like that? I guess I never noticed before.

Will anything prevent you from mounting a camera looking forward? Perhaps balanced on the other side by one looking aft at Jim's-spinning-fin-thing?

Yes, they have always been offset like that. I think the servos could be oriented vertically and then the canards would be at the same level. Just a little more difficult to fabricate, but possible.

Great idea on the camera. I don't think I can get a second #16 in time for the first test, but that's what I need to do.

Jim

PS - Camera ordered
 
Jim - I want to sincerely thank you for sharing your journey! I've pointed the IREC team I advise to this thread and the information contained within has been invaluable.

Edward

No problem! I get plenty of ideas from you guys in the process.

I think we're going to try and port the firmware to another platform since the UDB5 is no longer available. Need to get started on that.

Jim
 
I need to engage the roll control right from the start of the flight, as roll does not help the gyros at all. But, I need to decide whether to activate the vertical stabilization from the start, or alternatively, a second or so after launch. Any thoughts on that?
Personally I would start active stabiliisation after boost. That makes sure there is some altitude distance between the rocket and anything that matters, before kicking in the steering. That's how I will be playing mine a little later in the year. Straight for boost, then enable. YMMV, but you need to think off-nominal operation for the failure modes.

As for the keychain cameras, I am looking at mounting two of them to provide logging of what the fins actually do. I will 3D print some fairings to screw onto the outside of the rocket. Keep the drag and aerodynamics symmetrical and there will be almost no change in performance.

Good luck!
 
Personally I would start active stabiliisation after boost. That makes sure there is some altitude distance between the rocket and anything that matters, before kicking in the steering.

I think that's a good idea (to delay the start of steering). I think I would prefer to do this, though, before the speed gets too high.

Jim
 
Jim - I want to sincerely thank you for sharing your journey! I've pointed the IREC team I advise to this thread and the information contained within has been invaluable.

Edward

+1 The high schoolers I'm working with are building a hand-laid carbon airframe, and your handbook is their guide. I heard "Have you read Jim Jarvis' book on this?" an awful lot before and during tube layup.
 
Back
Top