Combining a Featherweight Tracker and Raven into a 38mm av-bay

[Adrian A]

I'm back to doing rocketry, and working on another attempt to fly a 3-stage 38mm rocket for Balls this year. The sustainer and the 2nd stage will each have a Featherweight Tracker and Raven in a 2" long av-bay. This is a tight squeeze and it requires a number of modifications involving the use of a soldering iron. So this thread isn't describing a turn-key product. Instead, if anyone is wondering how I pack my electronics when I'm going for altitude records, this is what I do. The end result is going to look a lot like this av-bay shown below, which I probably made as a spare for the 2019 flight:

This has a Raven, 2 batteries, 1 magnetic switch, a USB Raven battery charger, and the tracker positioned with its GPS antenna field of view preserved, as seen in the next photo:


I'm going to walk through step-by-step how to get there from stock Featherweight products, but first some background that might help:

I made a sustainer av-bay like this most recently in 2019, for a 2-stage rocket at LDRS/Airfest. That flight had a dud nosecone ejection charge, so it came in ballistically from 33,000 feet until the main fired at 700 feet and tore apart the front end of the rocket, including the av-bay. Surprisingly, the tracker fell free, still connected to its battery and powered on, and led me right to the crash site. The sustainer body was undamaged, and the Raven survived well enough that I was able to pull data off of it. The 38mm active bulkhead (the end of the 38mm av-bay with the magnetic switch, and battery and raven connectors) took damage in the form of a bent Raven header connector, 4 bent/broken threaded rods, and a cracked glue joint, but I decided to repair it so that I could re-use the screw switch installation that lines up with the sustainer. This is my starting point for this installation after I soldered on a new Raven header and replaced the threaded rods, and reglued the bulkhead to the white coupler tube:



The white coupler tube glued onto the bottom has 3 functions: 1. Mount the screw switch for the airstart arming 2. Provide a forward stop for the sustainer motor. 3. Provide a cavity for the head end ignition wires to go without shorting to the end of the motor case. The screw switch is there so that I have 3 inhibits to accidental sustainer ignition (Raven, mag switch, screw switch) while I'm transporting the rocket with the head-end igniter installed, and I have a safe method to arm the system once it's pointy-end up at the pad:

1. Start with everything off and disarmed (with the exception that the rocket is assembled with the airstart igniter in place) The arming screw for the airstart is not just loosened but removed so that it can't rattle closed. The rocket is 2-fault tolerant in this configuration.
2. Install the rocket into the tower.
3. Turn on the Raven. Listen for expected beeps, which only happens when it has completed a successful self-check. The beeps indicate that just the apogee and main deployment charges are connected. At this point, if the rocket launched, it would deploy everything safely but be unable to fire the airstart.
4. Install the screw of the screw switch and tighten it securely. Now the beeps for the airstart indicate that the airstart is ready to go.

O.k, back to the build thread part:

I started with an off-the-shelf tracker, and then removed the SMA antenna connector. I used a hot-air rework tool to do this. It's probably also possible with a hot iron, lots of flux, and patience.

I'm replacing the antenna with a simple length of wire. Solid core or stranded will do. I'm opting for sold core so I can keep a nice cross-section of copper while minimizing the holes I will need to drill later. The length of the antenna is important, and it's not the total length of the wire that you need to pay attention to, but rather the length that it sticks out past the ground plane that goes up to edge of the board. Online calculators will tell you a length around 3.1-3.25" for the 902-928 MHz band that the Featherweight Tracker uses. I decided to start a little long and then use measured reception strength to fine-tune. I started by putting the tracker with the stock antenna out on my deck, the GS at my desk, and used the FIP to measure a baseline signal strength of -53 to -51 dB. Then I removed the antenna connector, soldered on a wire that stuck out 3.25", and re-measured. The signal strength was up to about -46, so I'll take the win with no trimming needed.


A new feature of this av-bay build is a prototype of a new passive bulkhead that I might produce and sell. This is a bulkhead that goes on the other end of the electronics from the active bulkhead. In a stock av-bay the jobs of the passive bulkhead are: seal off the electronics from the deployment charges, provide a clamping force through the threaded rods to hold the av-bay in place, and provide some labeled terminals to connect the charges to. In this case, rather than making the passive bulkhead the diameter of the airframe ID, so that it clamps onto the end of a coupler, I'm going to use my chute cannon to seal against the end of the coupler, and this passive bulkhead will pull it down with a single bolt going through a T-nut, from a position recessed from the end of the coupler. Below is this special passive bulkhead that I had manufactured with a hole for the T-nut. I manually drilled holes for the T-nut prongs. Shown is the 10-32 stainless shoulder bolt that will go through the bottom of the chute cannon in the final assembly, and a bunch of grease to make sure no epoxy would get into the threads while I glued the T-nut.




More to come...

 

[cerving]

Interesting AV bay. Are you concerned with the proximity of the two allthreads next to the GPS antenna detuning it, realizing that they're somewhat in the dead zone...?

 

[Adrian A]

Interesting AV bay. Are you concerned with the proximity of the two allthreads next to the GPS antenna detuning it, realizing that they're somewhat in the dead zone...?

I was concerned until I tested it. The GPS likes its spot there, as it turns out.

 

[OverTheTop]

If you want to save some space in tight AV bays consider using titanium bike spokes. They can be made to custom lengths at good bike shops for a reasonable price and have a breaking load of around 280kg each. The 2mm diameter is much thinner than regular allthread, saving lots of space. They are a standard 2-56 thread.

The conductivity of Ti is about 1/4 that of steel (10.1x10^6 vs 2.4x10^6 S/m) so the effect on the fringing fields around the GNSS patch antenna (or any other antennas) would be reduced.

 

[Adrian A]

If you want to save some space in tight AV bays consider using titanium bike spokes. They can be made to custom lengths at good bike shops for a reasonable price and have a breaking load of around 280kg each. The 2mm diameter is much thinner than regular allthread, saving lots of space. They are a standard 2-56 thread.

The conductivity of Ti is about 1/4 that of steel (10.1x10^6 vs 2.4x10^6 S/m) so the effect on the fringing fields around the GNSS patch antenna (or any other antennas) would be reduced.

Interesting idea. I didn't know that about bike shops. I have only bought spokes pre-threaded with specific lengths. Do you know what kind of machine it takes to do that? In the 24mm diameter version of this av-bay I use steel 2-56 threaded rod in order to get it to fit. Since the GPS works well either way, I'd prefer lower resistance for the deployment charges, but Ti probably wouldn't be a deal-breaker.

 

[OverTheTop]

Do you know what kind of machine it takes to do that?

It is a thread rolling machine. The one I saw takes up about a square feet of bench space and is hand operated. The threads are rolled rather than die cut for superior strength due to work hardening.

Remember Ti is much lighter than steel too, a bit over half the mass.

Something like this:
https://www.aliexpress.com/item/4001276967820.html
Here is an example of how thread rolling works, for those who don't know:
https://www.manufacturingguide.com/en/thread-rolling

 

[Adrian A]

I'm using the stock-length (2.5") threaded rods, and I added some spacers and heat shrink tubing to prevent shorts:


The red one is the +Arm out that gets turned on by the magnetic switch. The green ones are apo and main, and the clear one is the 3rd output. The 4th output is only on the active bulkhead, and it's connected to the airstart arm screw switch on the underside.

I'm doing a fit check using the Raven4 I modified in 2019. The white plastic spacer keeps the tracker from pinching the Raven power wires:



Here's a closeup shot through my electronics microscope (which I use for everything, like reading glasses but better):

 

[Adrian A]

I was finishing up the sustainer ebay for StratoSpear, so I have some updated photos.

Blue Raven prototype with the supercap installed off to the side, and the new rounded corners.



Extension cord for the Blue Raven battery. The male and female connectors came in a pack from Amazon or eBay, I don't remember which. I just joined a male to a female and shrink wrapped it. This makes me think that maybe I should make a 38mm active bulkhead that already has the Blue Raven battery positioned for this.



With the Blue Raven capacitor moved to the side, and its battery moved to the opposite side, that makes room for the tracker and its battery:


It wouldn't be rocketry without some blue tape. The tape is just there to keep the wires from getting snagged when I put it through the sustainer body tube. Before I fly it I'm going to add some double-sided foam tape in key spots to give a little more protection from rubbing and hold things more securely. But even as it is, it's packed tightly enough inside the coupler that nothing can move far enough to cause real problems. Note the heat shrink around each of the threaded rods, which is there to prevent a variety of potential shorts. The one threaded rod that looks out of position was put that way temporarily so I could make a test connection for deployment charge testing last month.




At the bottom you can see the arm screw switch that connects the 3rd output threaded rod to the wires I have that go to the head-end igniter. When I need it for safety I like to remove the screw entirely so I know it can't close by accident (for example vibration during transport)

 

[COrocket]

Would it ever make sense to consider developing the raven/tracker functionality onto a single board to save even more space?

 

[Adrian A]

Would it ever make sense to consider developing the raven/tracker functionality onto a single board to save even more space?

Potentially. The chip I'm using for the Blue Raven has enough interfaces to also support a LoRa radio and GPS module needed for a combination unit. GPS tracking, 9-DOF inertial sensing, high speed data recording, low power states, and an iridium modem interface are included in a combined data return capsule that I designed for the NASA LOFTID program, which will hopefully fly this November.

A 38mm av-bay for a combined unit would actually need to be a little larger (longer) than the <2" length I'm showing here, but it would be simpler to set up, and it would be less expensive to produce than 2 separate units. I think a lot of customers prefer the functions to be in separate units for a variety of reasons if the cost is comparable. One reason is that trackers need to be in RF- transparent part of the airframe (nosecones are almost always FG), while altimeters work best in the middle of the rocket to deploy in both directions, where performance-oriented rocketeers often opt for carbon or aluminum for structural efficiency. Others like the flexibility of having a tracking module that can move from rocket to rocket even if the altimeter is dedicated to one rocket, or vice versa. The optimal batteries for the two units are different too, with the altimeter needing moderately high current for deployment charges that can preclude the use of batteries with built-in cell protection, while tracker's current never gets high enough to activate cell protection circuits, and the higher capacity the better. So I plan to continue with the Blue Raven and a GPS tracker as separate products indefinitely.

But because a combined unit is attractive in other ways (simpler interfaces, simpler mounting, potentially lower cost) a combined unit is on my longer-term to-do list, after the next generation of tracker is done (next year).