Arduino-based altimeter, GPS, telemetry, and possibly guidance

[awseiger]

A micro SD clicks in, I don't see why that would disconnect in flight. They are usually very secure.


Glad to see this coming to! I'll happily assist in any way.

For the press-fit connections on the Arduino, have some method of screwing the board down, like a screw hole that lines up with one on the Arduino.

 

[Derek]

I wouldn't worry about the shield coming loose in flight. Have you ever tried to pry a mega shield off the main board? There are a lot of pins and they do not come off easily. I'm always afraid I'm going to break something when I try to pull them apart.

I think think the micro sd card can be secured as well. you just need to find the right socket. I've used the flip cover type and I wouldn't recommend those. the cover is way to easily opened. the click in ones seem better.

1MB every 10 minutes is not a lot of data. you could always put a spi flash on the board. a 256mb part is only a few dollars.

https://www.digikey.com/product-detail/en/S25FL256SAGMFI011/S25FL256SAGMFI011-ND/3862752

not as convenient as microsd, but it will never fall out in flight.

 

[zortness]

Another change I'm looking at right now, switching to N-Channel MOSFETs to control the deployments, rather than the solid state relays employed currently. Seems like a cheaper, more standard solution (my Raven 3 appears to have 4 MOSFETs in 2 different varieties on board).

Anyone have experience using MOSFETs for separated circuits? PM me, thanks.

 

[thobin]

I would be interested in a kit too.

TA

 

[zortness]

Update:

Revision 2 of the board is nearing the final phases of design. Test fab and debugging soon.

I created a mailing list for people who are interested and want to participate in the design discussion:
https://groups.google.com/forum/#!forum/osaerospace

Changes in Rev 2:

• reduced number of parts
• all sensors now 3.3V (reduced complexity)
• 4 pyro channels instead of 2, mosfet controlled
• continuity detection
• input voltage detection
• updated to BMP180 altimeter (backward compatible with BMP085)
• 200G 3-axis accel (ADXL377) replaces 2-axis 55G accel
• added a buzzer (CCV084)
• friendlier pin usage for additional Arduino shield stacking
• Xbee retains ability to switch between UART0 and 1, GPS now permanently tied to UART2
• better LED and switch accessibility
• better accessibility to external GPS antenna port (between pin headers)
• all ceramic capacitors
• still about $111 parts cost

Bill of Materials breakdown/costs
Manifest

 

[zortness]

Just an update: ordered some boards to assemble and start testing.

 

[therocketfarmer]

interested in your boards, thanks!!! get a hold of me if you have anything available.

 

[zortness]

I've got one assembled and functioning, pics here:
https://plus.google.com/115738820947903142267/posts/37AggdWrywg

I've been debating on selling these. It would probably cost me a few thousand $ to be able to do that, so we'll see. I've got a few "life-decisions" in flux at the moment that will basically determine the feasibility of that.

Right now, working on the new software.

 

[Phil125]

If you wanted to make it thinner you could probably use an Arduino Nano as a daughter board on a Thinner board with the rest of the add-ons. Might be able to get it down to an inch wide that way.

 

[cerving]

I think it's likely that the software is too big for the ATMega328P MCU on the Nano/Uno, which is why he designed it as an Arduino Mega (ATMega2560 MCU) shield.

If you wanted to make it thinner you could probably use an Arduino Nano as a daughter board on a Thinner board with the rest of the add-ons. Might be able to get it down to an inch wide that way.

 

[zortness]

Yea, reading this many sensors requires too much FLASH and RAM for the smaller chips. It's actually approaching the limit of the Mega.

The other thing is that I'm using 3 hardware UART ports, which the smaller-form-factor arduinos can't handle.

Next project will probably be with something like the STM32F4 as the core.


Status of the actual project: I'm playing with the software now (using ChibiOS for threading). The R2 board is going to go through a revision: I realized the Due's pinout doesn't actually match the Mega, so I have to move some analog pins around. I'm also having an issue with the ADXL377 being super-noisy. Going to try and resolder it tonight, maybe I used too much solder paste or something. Will probably order the R2.1 boards shortly, as soon as I get that analog noise issue figured out.

Again, not sure of a timeline on sales... California taxes are definitely a concern at the moment.

 

[keithp]

Next project will probably be with something like the STM32F4 as the core.

Just so you know, the Altus Metrum code, AltOS, runs on the STM32L already. It's GPLv2, so you can take it and use it in your own projects/products.

 

[zortness]

Yup, I had actually seen that, Keith. Good stuff!

 

[Binder Design]

Hey that's neat! I was wondering when someone would do this. I love DIY projects.

On another note, I'm looking for someone to help design a board for me, non rocket related. A guitar pedal. Would you be interested or know someone who is good at that type of thing? I only have a real basic understanding of it, so I need some help to get to the next stage of production.

Drop me a PM if you can help me out. I can pay in rocket merchandise.

 

[zortness]

Back from some travel. The analog accelerometer was busted, replacing it fixed the analog noise issue. Not sure what caused it to break, probably my soldering. Still going to do a revision of the board to at least make it physically compatible with the Arduino Due. The only things that are going to change are the output pins and a bit of routing. Everything else is dandy at this point. Trying to work out some time to do some test launches (I live in the middle of LA, so it requires driving far away).

Got some personal stuff worked out as well, now I can work on being able to actually sell these for those of you who want them.

 

[samuraisoldier]

Hi there, I realise this is 2 and a half years later, but I am part of a team at Imperial College London and would like to talk to you about purchasing the avionics board you talked about here? Hopefully you still check this website!

 

[Former Squirrel]

I have a decent amount of interest (thanks guys!), but my lawyer says I'd have to incorporate an LLC or something similar to reduce my liability in case something goes wrong on a high power launch. There are a bunch of fees associated with that in Cali, unfortunately, but I'm figuring out all of the details.

For Brazil: I'd have to check export laws, etc. on where I can send these things. I imagine it won't be an issue, but I don't know yet.

The clear winners so far for preference:
- Unit as an Arduino Shield
-- With a healthy bunch requesting a standalone system, and even some processor upgrades
- A "V2" board seems to be the general preference
- Mixed feelings on the Kickstarter idea, mostly positive
- Quite a few people ready and willing to help on hardware and software

My plan for now is to make some improvements to the board (v2), and then start a group buy or small funding project.

Improvements listed (so far) for my V2 plan:
- tie VIN and EXTBATT together (allowing the JST plug to power the main Arduino board)
- replace the TI Power MUX with either a P-FET setup or a diode
- replace the two large electrolytic capacitors with ceramic capacitors
- replace the 5V 55G 2-axis accelerometer with a 3.3V 200G 3-axis accelerometer (unless I can find one with a lower scale at the same voltage)
- remove 5V regulator
- add a 1A fuse
- move the XBee uart channel select switch and LED down for easier view/access (the XBee currently covers them)

Overall, this should remove some complexity from the board. I started mapping out these changes the other night, and am currently trying to teach myself about P-FETs to see if it's worth using one over a diode.

If anyone wants to help with the hardware design for V2, PM or email me! That might help me get it done more quickly.

Several people brought up their discomfort with this as a shield, mainly because of the press-fit pin connections. Unfortunately, for the price range people requested, I won't be able to do a run of completely integrated boards for the first time around. My plan would be to spin that off as a separate project (same software), eventually offering both the shield and the completely integrated board (maybe a couple of months from now), and possibly even a modular setup.

I also had some people express concerns about using the microSD card losing continuity due to vibration of the pins. I'm trying to find a better solution with lots of storage space, but don't have one yet. Testing with my RC car, I can generate data files of about a 1MB in less than 10 minutes of goofing off, so I'm logging a substantial amount of data. In the meantime, there are several successful examples of very large rockets launching with microSD cards in slots (Qu8k is one of them). Hopefully it's not an issue, but we can try to design the software accounting for potential loss of the card mid-flight. Definitely a risk to be noted.

I'm working on an avionics bay for one of my 2.6" rockets (even as I type this), so hopefully I can get some test flights in shortly. I've had some "life changes" recently that have made this a bit more difficult than I had anticipated.

Again, everyone, thank you for the support! This has been an awesome experience and experiment for me so far.

Have you thought of Adafruits Feather boards? The LoRa with an M0 Express and a gps wing to start. Im working a feather wing pyro board.