Stretch Mustang Painting

I’ve started painting and finishing the stretch Mustang.  This picture is after sanding after the first coat of paint.



BeagleBone USB

I’ve been working on the flight electronics for the 6″ rocket.  One issue is that I’ve got a few peripherals that I want to connect through USB, but the BeagleBone Black only has one USB port.  I’ve been working on connecting them through a USB hub, and have been having some problems.

What I’m trying to connect is an IMU and a GPS.  I can connect each one individually directly to the USB port on the BeagleBone and they basically work.  There is a minor odd issue where I occasionally get spurious characters in the data stream from them.  This would be usable by checking for and throwing out bad data records, but it does make me wonder whether there is a problem with the USB port on the BeagleBone.

Regardless of that, when I connect them through a USB hub I can’t even get that functionality.  They show up in /dev/serial, but I can’t get any data from them.  Looking around on the internet turned up some possible ideas, but nothing conclusive.  It said that the BeagleBone is finicky about which USB hubs it is able to work with.  Maybe if I just get a different USB hub everything will work, but no guarantees.  Also, there was mentioned a problem where a powered hub, if it doesn’t have a protection diode, can feed +5V backwards, which the BeagleBone can’t handle.  It was suggested to use such hubs unpowered, or cut the power wire in the cable between the BeagleBone and the hub.  I tried running with the hub unpowered, but it did’t make a difference.

Here are my possible options for going forward:

  1. For now, just hook up one device directly to the BeagleBone’s USB port.  I’m going to start with them in ride-along mode where I’m not depending on them for anything until I see they can put out good data in flight.  This will remove the problem from the critical path while still getting data from one device at least.
  2. Try out a different USB hub.  I will probably do this eventually.
  3. Communicate with the boards some other way like a direct tty serial line.  I’d rather not bother with this.  USB is so convenient.
  4. Other even more out of the way options like going with a different flight computer.

Pyro-Free Parachute Ejection

I didn’t work on rockets at all during the holiday season, but I’m getting back at it now.  I made some more progress on my pyro-free parachute ejection system for the 6″ rocket.  Before, I had a CO2 cartridge and NOS solenoid that will blow the drogue chute out, and batteries that can supply enough current to open the NOS solenoid.  What was missing was a high power switch that could be controlled by the on-board computer and handle the 10 amps that the NOS solenoid takes.

I ordered a high power solenoid driver from Sparkfun.  I made an oversight in that the solenoid driver takes 5v logic in, but my Beagle Bone can only output 3.3v on its I/O pins.  So I had to make a little low power switch out of a transistor that can be driven by 3.3v and output 5v, which becomes the input to the solenoid driver.

I got everything hooked up and was able to drive the solenoid from batteries with the driver controlled by the Beagle Bone also powered by the same batteries.  I’m using an Any Volt to supply regulated 5v to the Beagle Bone, which also probably provides some brown out protection.

img_0232 img_2424

I did a bunch of puzzles over the holidays, when not doing rockets, which you can see on my work surface.

Stretch Mustang Parachute Ejection Test

I did a ground test on a parachute ejection charge for the stretch mustang.  I started with a quarter gram of black powder and it only pushed the nosecone about halfway out.  I’ll go up to a half gram and test it again.  After that about all that’s left is decoration and launch day pre-flight prep.

[Update] I did another test with a half gram and it worked fine.

More Progress on Stretch Mustang

I epoxied the e-bay bulkhead into place in the Stretch Mustang.  So then I could test fit all of the recovery components inside the rocket, and I’m happy to report they all fit.  There are a few minor odds and ends to finish up.  The only major thing left is painting.

It’s fairly close to being ready to fly.  However, winter is here in the rockies so I don’t know when the next day will be when I feel like braving the weather to go fly.  In theory Northern Colorado Rocketry has launches once a month even through the winter, but they are often cancelled.


More Progress on Moment of Inertia

I epoxied the ring fins on Moment of Inertia.  For a motor retainer, I mounted the ring fins back off the end of the body tube far enough to have a fiberglass ring behind the motor.  The ring is attached to the ring fins with screws.  It’s probably easier to understand with a picture.

img_0035 img_0038

I also got a Semroc elliptical balsa nose cone.

All that’s left is the recovery system.  I talked to a friend who is an electrical engineer and he said circuit boards should be good to a lot more than 100 Gs acceleration even with components like big capacitors soldered on.  Maybe there’s a particular component on that timer that limits the G rating.  I think I’m going to pick an off-the-shelf altimeter and hope for the best.


Further developments about the Moment of Inertia rocket:

There have been problems with ejection charges on Cesaroni Vmax motors not firing.  Apparently, the rapid change in chamber pressure at the end of the main burn snuffs them out.  NAR and TRA have enacted a safety restriction requiring electronic recovery deployment for rockets with Vmax motors.

No problem, I thought, I’ll just add an altimeter or timer to my design.  I was looking at the PerfectFlite miniTimer4 and I saw this:

  • acceleration limit: (best accuracy) 15G
  • acceleration limit: (accelerometer) 23G
  • acceleration limit: (timer function) 100G

Dammit!  Moment of Inertia is going to experience well over 100Gs.  Other boards I’ve looked at don’t list an acceleration limit, but I think I should be suspicious about any COTS electronics being reliable at those acceleration levels.

So what are my options?  I could make my own fully potted electronics and test them (somehow) at high G forces.  That seems like lots of work.  I could wait until Cesaroni fixes the ejection charge issue.  Cesaroni has a lot on their plate right now.  That fix might not be coming any time soon.