Rocket Avionics Concept your thoughts?

[tmassey]

Hello all,
I am in the process of developing a well I think pretty cool rocket computer. I know I know there are tons of them out there already. However my idea falls around the fact that when I want to add on to my rocket new sensors and systems they are all disconnected systems that do not integrate well. For instance say I want to build a small rocket that will house an altimeter. But my tax return comes in and I decide now I want to add an accelerometer to see exactly how the rocket fly's. They are separate systems, I will have to get data from one place then another and I am duplicating chips unnecessarily, this adds weight.

This thought came to me why not take a computer style approach. A Computer has USB, PCI, AGP etc.. Or if your old school they might even have ISA or MCA. All available for expansion. So I first started making my design with my microcontroller designed so that it made available most of it's IO pins for connecting other devices. the only on board items were a Micro SD slot, USB Connection and a Barometric Pressure sensor. I exposed 8 Connectors with 5 Pins each for power and 3 PIN SPI interfacing. but in the end It was ugly. It looked like a spider was emerging out of my rocket. So I then decided to change it up a bit and get rid of most of the wires. Lets stack the boards ,but my small rocket isn't going to expand much before I am too big probably 2 boards max. So then I decided what if I make the boards Round and position them perpendicular so they are not sideways. and it worked.

So for the question. I have some ideas on where I want the product to go but honestly I don't want to spend countless hours and dollars developing a system if I cant sell it. What I have so far is a Windows software application that is drag and drop simplicity at it's core. As easy as using one of the big name Smart phones of today. A plug and play stackable computer with firmware close but still testing to get out all the bug's well thats not possible but 99.999%. Board designs for some 9 different sensors and add-ons. Oh and Wireless up link to a computer for live telemetry and programing through the Windows app. I have spent the past 3 months working out the details of this design. Most of the details are worked out on the main computer but I still question the concept. I am very proud of my results. but my concern is I am one person. what do others think about an expandable yet tightly integrated system. and what kind of ad-ons would others like?

I Started putting together a daily video blog on my website www.synapsesg.com if you want to watch the progress or see the problems I am running into. I know this is kinda hype and marketing but I am more at a breaking point in my project. Spend the money to continue or move on to something else.

The boards are designed to fit in the 56mm dia BT-70 size tube I tried going smaller but it simply didn't work well and would have drasticaly increased the number of layers I woudl need on my boards. and in turn increasing the cost.

 

[r2kordmaa]

right now i'm designing an avionics computer with:

• telemetry
• gps
• accelerometer, barometers, etc analog inputs
• communications logging

it will fit on 33mmX100mm pbc

in future systems im thinking of adding live video downlink and 3axis gyro system with perhaps added active stabilisation with fins
read my thread here for schematics and more

 

[dostergaard]

. . . All available for expansion. . . . So then I decided what if I make the boards Round and position them perpendicular so they are not sideways. and it worked.

. . . What I have so far is a Windows software application that is drag and drop simplicity at it's core.

I really like the idea of being able to add features without having to replace what I have in order to get new capabilities. That said, I would also expect expansion cost to be reasonably incremental.

Also I would encourage you to create an open API for your product so that others can create utilities and other modules. An example of that can be seen at Altus Metrum.

Just my :2:

 

[Handeman]

I like the idea of expanding by stacking disks. It is a neat solution to get lots of electronics in a small space. My question is how do you mount the base unit and the expanded disks and how well do they stand up to heavy G forces and hard landings?

 

[troj]

A stacking approach, like Arduino shields, has its merits, but it has one significant downfall -- size.

Most people fly rockets that won't handle exceedingly large electronics, and a stacking approach just seems to me like it would add to size/bulk.

In my opinion, if you can't fit two of them, with appropriate power sources and switches, inside a standard 3" coupler, then you're severely limiting your market. Edit: Okay, re-reading, sounds like you've got the size down to something very usable. How tall is the stack? How is it powered?

Can it be made price-competitive? Take a look at what Adrian has done with the Raven 2 at $150 for an idea of where you have to compete, from a price perspective. That's a small, full-featured package, at only $150.

-Kevin

 

[gpoehlein]

I love this concept as long as it isn't too expensive (within reason of most other avionics out there.) To my way of thinking, the "plug and play" capability should also include "unplug and still play" - in other words, if I don't need a certain instrumentation for a given flight, can I remove components to control weight and still use the system? To me, this would increase its usefulness over something that you can add to but not take away from. Kind of like, if I have an external USB drive for backup storage, there is no need to run it all the time - only when I need it.

 

[BEC]

You might take a look at Eagletree Systems RC data logger for another example (and price model) for an extensible system, just to get some idea of what other things are out there. It can be used with just its onboard sensors or with one to many added ones. Configuring which ones are used at any one time is easily done via the same software that allows download and plotting of the logged data.

 

[degreaser]

... An example of that can be seen at Altus Metrum.

Just my :2:

Thanks for that link. I hadn't heard of this project. The fact that the firmware is GPL'ed, uses open source development tools, has an open hardware architecture and interfaces via standard Unix/Linux tools is a huge plus in my mind.

 

[tmassey]

I love this concept as long as it isn't too expensive (within reason of most other avionics out there.) To my way of thinking, the "plug and play" capability should also include "unplug and still play" - in other words, if I don't need a certain instrumentation for a given flight, can I remove components to control weight and still use the system? To me, this would increase its usefulness over something that you can add to but not take away from. Kind of like, if I have an external USB drive for backup storage, there is no need to run it all the time - only when I need it.

That is the whole Idea behind this system I have a survey off of my website and I am providing a discount to all who fill it out. Once the product is released. I have conducted several polls on forums across the model rocket community and all signs point to yes on this product.

 

[troj]

Interesting... I just looked at your website, and some of the modules you mention are along the lines of things I've seen teams do as SLI projects!

The gas sampling is the one that really caught my eye, in that regard.

-Kevin

 

[gagreen]

as a former avionics journeyman i would love to see expandable avionics, but as an avionics maintainer i also hated them. The idea is solid and integrating more and more into a smaller package makes good sense, at the same time redundancy is also important, for example losing power or a short could kill the rest of the system. depending on price could be extremely costly to replace as opposed to a compartmentalized system with different unconnected components where you only replace a single instrument. I guess it all comes down to craftsmanship, and a solid ofp.

newer fighter jets have a pretty good system going, one main bus controller and reporting component that all the systems talk to. In the case of a bus controller fail the systems can also operate in a downgraded but functional way. Downsizing this to a 3" rocket would be a challenge but a set of magnifying glasses and going all surface mount circuitry could make it extremely small but tedious and probably improbable for mass producing out of your house lol. good luck

 

[Louis]

To make an extension system working in a small scale you will need a bus system. Most sensors come with SPI, but you need a chip select for that, thats ugly for a bus.
Easier is I2C (or TWI), which only needs 2 lines + current = total of 4 lines.
You can hook many devices on TWI, with 400 Hz it is normally fast enough for these purposes.

One idea is:

Design a system board with controller, flash, comm, pyro outputs and barometric sensor, a small altimeter in fact.
Then put a connector on it for the bus, on which you hook your sensors. The bus has SCL, SDA, +5V and GND. 5V to have the posisbility to get 3V, 3.3V or 5V for your sensors.

For these you just make a small pcb, your preferred sensor, voltage regulator if needed and a small atmega for sendig data via TWI bus. Your controller asks these sensor boards for data and they answer by sending it.
You could set specific sensor adresses to hook up different sensors and the controller recognize them automatically.

Your main board handles the flight and chutes, saves the data to the flash chip and sends it out to a computer. So you can use it without sensors, just to fly, and with hooked up sensor boards to measure.


You can also use SPI and add 2 lines for the chip selects, that would allow you to select up to 4 different sensors, since most modern sensors have built-in SPI with 3V you could hook them up directly with a small CS-select-circuit.

Louis