[islandlabs] Data acquisition system for the high-flyer.

Burns, William burns at cshl.edu
Thu Jan 14 22:22:43 EST 2010


John: 

Part of our goal is to keep costs down. (The MIT project flew for under
$150)
A datalogger would provide function above and beyond the MIT payload.
So.. As long as someone's willing to pay for it, I have no problems w/
it costing more.

> -----Original Message-----
> [mailto:list-bounces at islandlabs.org] On Behalf Of John Kuczewski

> .... Bill, do you know the specs of the batteries you 
> mention? The PIC has a very wide operation voltage (2.0V - 
> 5.0 V) and has some instructions for putting it to sleep.

There are 2 types of power supplies, both doing (I believe) buck/boost,
one of them operating from a single 'AA' battery, and the other from 2
'AA' batteries. Both types supply 5V.
These power supplies either have USB style plugs on them, or some other
cell-phone power plug which could be cut off, and hard-wired to your
microcontroller.

The 2 cell power supply should go to the phone 'cause it's more critical
re: being able to find the payload landing site, and because it's the
exact type used in the MIT "reference design".
If we can get the microcontroller to work on the single-cell supply
that'd be cool.
(We'll have separate power for each device so they can't take each-other
out)

We already have the 'AA' lithium batteries.
I don't have power specs on these batteries.
They were chosen as part of the MIT reference design because of their
ability to operate in extreme cold environments where other battery
chemistries will not.
This combined w/ the hand-warmers should keep our batteries (and most
electronics) in a functional temperature range.

> o  Software: The PIC would sleep for most of the time, and 
> would only wake up upon a interrupt, service the interrupt 
> and fall back to sleep again. I would hope that this system 
> would be interrupt driven only, to conserve power that's 
> unless if we want to add GPS or some long rage wireless radio 
> to communicate with the system. Do we?

GPS would be very cool (temp/weight budget constraints aside)

There was some talk of doing/using APRS.
APRS would be a winner given it's ability to transmit data to
practically any point on the globe. (and be easily logged + viewed on
the internet)
We never found an up-to-date APRS equipment list, didn't find/verify the
weight of older hardware, or determine power requirements.
So... I still don't know if this is possible. (it might be too heavy, or
draw too much power)

We *do* have a much higher lift capacity than MIT, and it would be
*very* cool if we can get this to fit.

> Doing some homework, I found that SD can be put into 
> communicating via SPI [4,5] and not the proprietary way of 
> talking. And even better, Microchip has some code for talking 
> to an SD card via SPI [6] -- this would be a great starting 
> place for us. Any objections or other ideas?

Being non-volatile is a huge advantage. I wonder if there are any
near-space (radiation) problems.
If it works on the ground, and fits in the power/weight budget, I'd fly
it.
It'd be cool to fly w/ all (radios, SD, and every sensor-category)
options.

> Sensors:
> 
> There is also LM74 [9] that talks via SPI. It is up to you guys.

I'd lean towards the SPI chip. (I assume there's more overhead involved
in running the A2D, and that the A2D port might be more valuable for use
w/ some other sensor that's not available in SPI form)

> o Barometric Pressure: We might want to use the MPX2100AP 
> [10] from freescale that I found in this paper [11]. Any thoughts?

Very cool. Should provide interesting data. Very useful/complete-looking
app-note.

> o Accelerometers: I know that sparkfun has a whole lot, even gyros!
> (yum! :). Might want to look there?

Interesting sensors, but would provide the least valuable data. (for
this launch/application)
If we have to cut sensors to make budget, this one would be the first to
go.

> o Humidity, wind: If you guys have an recommendations on 
> cheap sensors, let me know! I haven't look into these yet 
> (almost the end of lunch).

Don't know but:
Here's a PIC based temperature application:
http://www.microchipc.com/sourcecode/index.php#PIC18LF4550_LCD_temp_sens
or

Here are some 1-wire temperature loggers. (DS1922L and DS1923)
http://www.maxim-ic.com/app-notes/index.mvp/id/1796/

This would involve the most "integration" w/ the rest of the payload.
(putting a hole in the box)
Some 1-wire sensors combine temperature, humidity and airflow in the
same device. (I don't have part numbers)
I don't know what range of temperatures they're accurate in.

Any "wind" sensor would only be able to measure relative to the speed of
the (spinning?) payload.

> Alright, I think that is enough to start with. Any thoughts 
> on this system? I haven't gotten into the software design 
> just yet, maybe this evening I will post again with some ways 
> of putting it together on the software side.

Sounds awesome so far.

-Bill


More information about the List mailing list