I have four of these IMUs laying around I grabbed from eBaY. It turns out the pinout is the same for their other modules, so might as well make a couple of these as I tinker with them
So far, the largest sink of time in making these is 1) the 3d print and 2) the diodes... I can wipe out the second portion by having the assembly house do those for me. Since they're common diodes, I can probably get away with doing that cheaply.
This time I 3D printed the plate as well, and had great results. I think if you do not have a laser cutter, this works just as well. The whole keyboard feels much more rigid and square. This one turned out well.
Alongside my own vehicle, I have been helping a friend out with his liquid fueled rocket project. It uses an LR101 lox-RP1 engine running at 1000lbf roughly. It is pressure fed via some helium tanks. So far the vehicle is passive, but he was interested in adding some electronics to control attitude closer to the ground, at lower velocities, and when the ultimate trajectory has the highest sensitivity to winds and perturbing forces/torques. This will be done with four cold gas jet thrusters at the top of the vehicle. They're purely for making the vehicle upright, so any lateral translational motion is fine with us.
So I designed a little STM32 flight computer board that includes a GPS unit, radio, two IMUs and an interface to control some valves. The valve control board is a relay board. I know someone out there is going to whine about this aw but never put relays into rockets, they're going to short under vibration. Well I'm not sure what to tell you. I have smashed the shit out of this thing with hammers, tables, etc and cannot get them to click over (verifying with actual solenoid valves and their power supply). Am I doing something wrong? I can easily swap with MOSFETs, but I need a an argument that beats my hammer.
I designed this board in April 2020 (almost a year ago now) and have not gotten a chance to fix up the code for the STM32. I was launching satellites in my day job. But now that they're in space, I guess I can have some free time, right?
This is what the board looked like in Eagle after all the placement and routing. The posts on the corner are for standoffs that will mount to the relay board. The large component on the right is an Aceinna IMU. The board can control up to eight valves / relay driven devices and is powered off of one supply, being bucked down to the right logic levels for chip and accessories. The 8 molex locking power connectors on the top go out to the valves / load them selves. The vias you see below go down to the relay switch board.
I had JLCPCB fab them, because they are super affordable and fast. So far they've been of excellent quality.
Checking out the new chinese microscope..
First power up. I flashed a simple LED blink script to make sure the STM got flashed. I am using the breakoff portion of a Nucleo board to program the chip vis SWD interface. I should write a blog about that one too.
I had a slight issue with a 5V trace to the relay board.. can't win them all. So far looks good.
Here is me forcing a valve on. I have since written the IO expander (MCP23017) code to do all this for me.