I thought I'd start a new topic to discuss what we learned yesterday, feel free to add and comment.
We might have worried about two much about temperature - the second flight used minimal insulation, no heat source (we also left the hand warmer out of the first vehicle for weight. Cold didn't seem to be a problem
It's not clear why we didn't get aprs pings, maybe the radio power was set too low, and maybe something (the antenna?) got disconnected.
I think in general we need to listen more carefully when someone suggests Murphy might be in the room. We should provide redundancy for things like ensuring that things stay connected, whether they are electrical connections or recovery pole components!
I'd say there is general agreement that flying a spot is a GOOD idea. Aprs is useful for providing real-time telemetry, but shouldn't be a single point of failure for mission critical aspects like recovery.
At the same time, as recommended by the Parallax near space ballooning book it would be good to log telemetry on-board the vehicle. If we proceed, I'd suggest that we develop a 'flight computer based on something like an Arduino, or the new TI Launchboard which handles the sensors, and produces a log to a flash card and the aprs stream.
One of my frustrations with the spot tracking site, valuable as it is, is that the fancy javascript heavy UI on their web site doesn't work very well on touch screen phones, whether they be iPhones or Android devices. There are so many places which do something when you touch them that it's hard to pinch zoom without suffering unintended consequences.
According to the FAQ, they do produce an RSS feed for a track, so I'm tempted to write a small web server app which tracks the feed and produces a simpler, iPhone/Android friendly tracking UI more suited to a balloon chase.
I think there would have been
I think there would have been a whole lot of things done very differently if we were not constrained by the competition/rules package.
Redundancy costs, and when 'cheap as possible' is the goal concessions against best practices are to be expected.
We made a good shot at it with the rules we were provided and the very short time frame we had.
Full-length preflight simulation
In reading about other flights, it appears a fairly common practice is to run the packed-up capsule in a freezer with the aprs and camera and everything else running until power is exhausted or some other problem occurs. You monitor telemetry the whole time, checking signal strength and so on. If you have something in the freezer for the camera to shoot, you also check the pics to ensure they come out okay.
If we intend to do this a number of times, would it be worth the effort of building a small vacuum chamber we can shove in a freezer?
One flight I read about lost telemetry because as the batteries ran down the camera power supply adjusted and began emitting RF that screwed with the telemetry electronics. That would be uncovered in a battery-exhaustion preflight checkout. I think in that same flight the camera got goofed on reentry because the membrane-covered buttons were all depressed by increasing air pressure. The vacuum of space had removed the air under the membrane on the way up, and on the way down the atmosphere literally pressed every button at once and held them all down for more than a day before air seeped back under the membrane.
Just wondering.
I've been noodling around
I've been noodling around with the Spot site and found that it's possible to auto-tweet and SMS location updates in tracker mode. That would be far easier than tracking through Spot's website which, as you said, is very graphical and not smart phone friendly.
I LOVE the idea of recording telemetry on-board! I know it's expensive and we could probably homebrew something better but my friend is an avid cyclist and he showed me the new Garmin gadget he bought (https://buy.garmin.com/shop/shop.do?cID=160&pID=331). The thing records and automatically maps stuff for you.
Edge 305 automatically measures your speed, distance, time, calories burned, altitude, climb and descent, and records this data for your review. For extra-precise climb and descent data, Edge 305 also incorporates a barometric altimeter to pinpoint changes in elevation.
Cool, eh?
SPOT, and wireless internet:
If you want to use the SPOT on an iPhone, you can get the SPOT iPhone app for $2.99. That will display the lat/longs, plus the speed the SPOT is traveling at, and the distance it is from your current iPhone GPS location.
Completely aside from the particular mobile platform (iPhone, Blackberry, Android) I argue that you can't rely on mobile internet in rural areas; in nearly any chase and recovery, we traverse rural areas. I know how many of us have smart phones; I have a HTC Evo, wow works great in central Raleigh, 4G and all. But, I argue we could build a much stronger chase and recovery team if we elect to have one dedicated team member at a stationary, grounded, perfect internet connection, with the phone number of the team members, etc.
iPhone SPOT app:
http://itunes.apple.com/us/app/shared-page-mobile/id346524744?mt=8
Jonathan
SPOT sending SMS in 'Tracking' Mode: Do Tell!
It would be outstanding if SMS messages were sent while the SPOT was in tracker mode! @Danforth: Do tell how this is done! I see how you can feed SPOT 'Tracker' mode pings to Facebook (via Yahoo FireEagle to FootPrint History, or FireEagle to FriendsOnFire), but I haven't seen how to push out SMS from SPOT 'Tracker' mode.
I see how you can auto-tweet the SPOT location in Tracker mode: SPOT to FireEagle to EagleTweet. But, I'm not finding a path from SPOT Tracker mode to SMS.
In native SPOT Tracking mode, they say it won't push SMS messages out (http://faq.findmespot.com/index.php?action=showEntry&data=566) but I'm all about learning a way to get that working!
Jonathan
Recording Environmental Conditions: Data logger
I have a Kestral 4000 which will log barometric pressure, altitude, temperature, humidity, and a bunch of calculated values (density altitude, windchill, dew point, wet bulb, heat index).
It will also record wind speed via a little impeller.....but wind speed is zero for a balloon-- we move with the wind! There is no relative wind.
You can log as frequently as once per second, or 'zoom' way out. For our use, I would think once per minute would be cool.
There are some limitations. It "only" goes down to -49 fahrenheit, and 10mb. Ok, 10milibar is wicked high, maybe 100,000 feet. So, that is pretty cool. http://www.nkhome.com/kestrel/kestrel-4000/features.php
It runs on AAA, so we could use lithium for their low-temperature performance.
I have been hesitant to fly it because of the cost ($350), and I was afraid of losing it.
For GPS data, we could put a handheld GPS in the capsule (e.g. an eTrex Vista, or another handheld) and pull the data off at recovery. That will give us GPS altitude and a track; some (like the Vista) have a barometric altitude as well.
Just ideas!
Thank you~
Jonathan
The Kestral sounds really
The Kestral sounds really cool, Jonathan. The cost is pretty high so I would suggest we have a redundant recovery plan (multiple trackers).
Regarding the Spot's SMS in tracker mode, I must have misread the site. It's still not hard to take a Twitter stream and rebroadcast over SMS though. In fact, I think it's a native Twitter feature.
Thanks for the tip on the Spot app! That sounds far more useful. I've emailed Spot's tech support / customer service requesting that they add altitude data to their reporting. I'm not sure if it's technically possible or even will be very accurate but I figure that there's no harm in asking.
My biggest concern for future flights is in obtaining higher and higher quality photo/video. I am really amazed at the images you produced with your little Nikon but we have two serious photographers on the team (I forgot the other guy's name!). With my connections in the art world it might be possible to FUND future flights with print sales... just an idea.
I am planning to take the photographs from the next launch (with my own camera equipment) and produce daguerreotypes from them.
I have 2 GoPros and if we had
I have 2 GoPros and if we had redundant trackers a Canon G9 we could potentially fly.. Have a underwater case for it, wonder if that would help or hurt?
iPhone app and some other thoughts on tracking
If you want to use the SPOT on an iPhone, you can get the SPOT iPhone app for $2.99. That will display the lat/longs, plus the speed the SPOT is traveling at, and the distance it is from your current iPhone GPS location.
Completely aside from the particular mobile platform (iPhone, Blackberry, Android) I argue that you can't rely on mobile internet in rural areas; in nearly any chase and recovery, we traverse rural areas. I know how many of us have smart phones; I have a HTC Evo, wow works great in central Raleigh, 4G and all. But, I argue we could build a much stronger chase and recovery team if we elect to have one dedicated team member at a stationary, grounded, perfect internet connection, with the phone number of the team members, etc.
iPhone SPOT app:
http://itunes.apple.com/us/app/shared-page-mobile/id346524744?mt=8
Jonathan
Jonathan,
Thanks very much for the link to the app. I looked for this on Sunday on my iPhone but there were just too many things with "Spot" in the name that I was unable to find it.
I also agree with you that we shouldn't use this as the only method of tracking. You're absolutely correct that rural wireless internet is spotty at best. I am, however, very interested in the possibility of having Spot send us SMS messages. While those can, at times, be very delayed, I think that it might be possible to receive them in locations even where wireless internet is spotty. Again, however, this would be a good secondary tracking method to use along with someone tracking at a location with good network service.
"Redundancy" doesn't have to
"Redundancy" doesn't have to be expensive. Some of the things I'm thinking of are:
making sure that all connectors two means of making sure that they say connected,
- screws on Dx- type connectors
- zip ties
- Yes I'll say it Duct Tape!
Lat/Long via SMS:
I'm a huge fan of getting the Lat/Long via SMS. With a voice connection you need a real-time, duplex connection to be effective. Spotty service means you can’t really talk, or use internet effectively, on a handset. However, if you don’t have service, SMS messages persist on the network. If you are driving around in rural areas, yes you will have spotty service-- but as soon as you do get signal for a minute, your phone checks in with a tower and gets the message that has been queued.
You don’t even need a good connection, or have it very long. If your device can check in to a tower even for a minute, even several minutes after the message was sent, we get a Lat/Long balloon fix.
Perfect.
We just don't have a solution where the SPOT will do that, when in 'tracking' mode.
One way: learn more about hacking the SPOT. It *will* push out a message if the ‘OK’ or ‘Custom’ buttons are manually pressed. Those buttons not only update the website, but also push a message out to 10 people-- by e-mail or SMS. Great! We just need to figure out how to make the SPOT keep sending an OK message, again and again.
Jonathan
one way...
One way: learn more about hacking the SPOT. It *will* push out a message if the ‘OK’ or ‘Custom’ buttons are manually pressed. Those buttons not only update the website, but also push a message out to 10 people-- by e-mail or SMS. Great! We just need to figure out how to make the SPOT keep sending an OK message, again and again.
Jonathan
Take it apart, find the contacts, remove the switch and solder in a relay or similar and have a micro-controller programed to 'push' the button every interval. Probably methods to hack the spot to do the same thing, but.....
Question if this would violate some Terms&Conditions somewhere though..
Another way to get position reports to the field
I'm a huge fan of getting the Lat/Long via SMS. With a voice connection you need a real-time, duplex connection to be effective. Spotty service means you can’t really talk, or use internet effectively, on a handset. However, if you don’t have service, SMS messages persist on the network. If you are driving around in rural areas, yes you will have spotty service-- but as soon as you do get signal for a minute, your phone checks in with a tower and gets the message that has been queued.
You don’t even need a good connection, or have it very long. If your device can check in to a tower even for a minute, even several minutes after the message was sent, we get a Lat/Long balloon fix.
Perfect.
We just don't have a solution where the SPOT will do that, when in 'tracking' mode.
One way: learn more about hacking the SPOT. It *will* push out a message if the ‘OK’ or ‘Custom’ buttons are manually pressed. Those buttons not only update the website, but also push a message out to 10 people-- by e-mail or SMS. Great! We just need to figure out how to make the SPOT keep sending an OK message, again and again.
Jonathan
Rather than hacking the spot. I'd be in favor of just having a fixed 'mission control' with reliable internet access which could send either SMS or email notifications out to the recovery teams phones.
This should also be able to be automated by having a server which reads the RSS feed from the spot service and forwards the reports. The software for that should be easy to write.
Scripting SPOT:
I'm a fan having both: a fixed person at a reliable internet connection, and having the spot send out SMS.
I'm working on a script to pull down data from the SPOT shared page and pull out the lat/long. There is a 'timeInGMTSecond' elemet as part of each ping. I will isolate that, and if it is newer than the last one I pulled down, then I'll parse out the lat/long and send it out by e-mail / SMS to whatever team members want to get updates. I'll cron the script to pull once-per-minute during launches, so within about 60 seconds of a new ping reaching the SPOT website, the lat/long will go out by SMS to the chase phones.
But, a person at home still helps a lot in coordinating, and also looking at the lat/long in google maps-- "the capsule is behind the Lowes!"
Jonathan
The Spot shared page has an RSS feed
I'm working on a script to pull down data from the SPOT shared page and pull out the lat/long. There is a 'timeInGMTSecond' elemet as part of each ping. I will isolate that, and if it is newer than the last one I pulled down, then I'll parse out the lat/long and send it out by e-mail / SMS to whatever team members want to get updates. I'll cron the script to pull once-per-minute during launches, so within about 60 seconds of a new ping reaching the SPOT website, the lat/long will go out by SMS to the chase phones.
Jonathan
Rather than 'screen scraping' the web page, probably better to use the RSS feed they provide
It's documented in the FAQ. There's a relationship between the shared page link and the rss feed, so
for the link for the take2 flight:
http://share.findmespot.com/shared/faces/viewspots.jsp?glId=024yYUFoJv7H...
the RSS link is:
http://share.findmespot.com/messageService/guestlinkservlet?glId=024yYUF...
and you get data in easily parseable XML form.
First, I think temperature
First, I think temperature testing is an excellent idea. However, your average freezer is not cold enough to provide accurate simulation. I recommend an insulated box and dry ice instead. Also, the lowered pressure makes almost zero difference to thermal behavior until you get way past 100 kft altitudes. (Basically it doesn't matter until you're talking about thermos-bottle quality vacuums.)
I understand the recovery from the tree was a bit of an adventure. Some things I've learned recovering (a modest number of) rockets:
Long shock cords. If there's 100 feet of cord between chute and payload, you're likely to have one or the other on the ground.
Put a loud sonic beeper in the payload. It can be really easy to be nearly under the payload and not be able to see it. Having a loud beeper helps.
My recovery kit includes a number of things you might not think about. FRS or similar radios: you won't always have cell coverage, and being able to coordinate anyway is helpful. Pruning snips: if you have to get through 20 feet of brambles to get to the payload, you'll be glad you brought them. Gardening gloves and bug repellent also make it in.
I have a slingshot, fishing weights, and line to snag those long cords and haul things out of trees. I use a braided line (I think called "spider wire" or something), sold as fishing line, attached to a 1 oz fishing weight. It doesn't curl up as much or tangle as much as monofilament. I then use the lightweight line to drag a piece of heavier cord, and pull with that. Note that you'll want plenty of excess fishing line, and more than double the amount of cord for the highest point you're recovering from. For an 80' tree, I'd want something like 200' each of line and cord.
I have a long pole, with a piece of wire pointed backwards (a barbed hook), and a retractable razor knife in the corner at the top. (Think tree pruning saw, but sharper angle, and a razor blade in the corner.) It's perfect for cutting shock cords to release things tangled in trees. (Of course, once you start cutting cords, you're probably giving up on the parachute, but those are fairly cheap.)
Hack Me Spot: Script
I've gotten a script running that will pull down the SPOT data from their website, parse out the most recent TRACK message, and send out the location by e-mail & SMS to any number of team members. So, the lat/long can be 'pushed' to our phones while the balloon is in flight; you don't need a smart phone or internet connection; it is just a text message, with the lat/long as the first thing to come across.
So, with that, everyone on the larger team can more readily be a track/recovery person. I like that. No special equipment needed, just a cell phone and a map. (Or, more likely, a car GPS...as long as it takes lat/long inputs.)
If you DO have a smart phone, and internet, the message includes a link to the position on Google maps.
So, that's cool. EXCEPT: The xml feed from the SPOT website does not update with the same frequency that the SPOT device itself beacons. From the FAQ: "Also note that the XML data is cached by our servers. The message service should be called no more frequently than every 15 minutes, since the data is cached and data will not be made available any faster."
So, the script will push out the most recent XML data available....which might not be the most recent ping sent by the SPOT.
Imperfect.
Advantage: Pushes messages out to cell phones; no special equipment; no additional cost (except SMS charges on your phone, if you are charged for text messages by your provider.)
Disadvantage: Lags behind real-time.
Jonathan
Hack Me Spot: Chip mod (Pic12C629)
I have a friend that has a physically hacked SPOT. I talked to him, and was able to go upstream to the source of the hack, to the electrical engineer that made the mod.
By soldering in a Pic12C629 (via 4 solder points) that is running some software he has written it will make a SPOT1 turn itself on, send a message, and then turn itself off. (Well, there is still some minor power drain for the PIC, but not much: 3uA draw in sleep mode, between PowerUp/Ping/PowerDown cycles.) This means it will run for a long time: estimated 9 days on a pair of AA lithium. The SPOT 'Tracking' mode, by contrast, ends after 24-hours.
Once modded, the device becomes a ping-monster, with no other functions. Literally, if the batteries are installed, it will just power cycle, ping, sleep--- until the power fails. There is *no* user interaction-- you start the thing by putting in batteries. You turn it off by taking the batteries back out.
He has offered to solder in his PIC for us, if we wanted to mod a SPOT1.
Advantage: turns a SPOT1 into an auto-pinging machine that will auto-push out SMS messages to our phones.
Disadvantages: We need a SPOT1 to mod; we're limited to 10 e-mail/phones to send the messages to (by the SPOT website.)
Anybody interested in modding a SPOT1? If so, want our group to solder the PIC, or send it off for soldering/testing by our EE-Hack-Me-Spot friend?
Thanks!
Jonathan
Spot 1 tracking
I have a friend that has a physically hacked SPOT. I talked to him, and was able to go upstream to the source of the hack, to the electrical engineer that made the mod.
...
Advantage: turns a SPOT1 into an auto-pinging machine that will auto-push out SMS messages to our phones.
Disadvantages: We need a SPOT1 to mod; we're limited to 10 e-mail/phones to send the messages to (by the SPOT website.)
Jonathan
It seems to me that the 10 person limit could be dealt with by having some 'ground control' software which read the RSS feed, and sent out e-mails and text messages via an sms gateway.
I'd be willing to tackle writing such software.
RSS? You mean the XML page we already parse?
Rick, you've refered to the 'RSS' feed; can you point me to exactly what you're talking about?
If you mean the XML page, as I wrote above, we already have a script that pulls that down, parses it, and sends out notification to as many devices (by SMS and by e-mail) as we want. So, put the SPOT in track mode, and we're now pushing lat/long out to as many phones as we want, by SMS, no internet required, no problem.
But, the XML page is not updated as frequently as the Tracking website. You've said RSS-- but as far as I see, the functionality is just taking that same XML data source, and feeding it into some other tool, like Yahoo Pipes, to create an RSS feed. The source is still the same: the XML page, which we are already pulling down, parsing, and sending out. No sense in pushing that into an RSS feed; we already have a script to pull in that data; we can do whatever we want with that data-- push it out, log it, etc.
On a separate note, the unmodified SPOT only runs in TRACK mode for 24-hours. For long duration, one may want the device to keep going until the batteries are depleted. That's another thing the physically modded SPOT does: runs for about 9 days 'pushing' messages out, all on one set of lithium batteries. That modded SPOT is essentially pushing the notification button, and sleeping; the notification button pushes near real-time alerts out, but it is limited to 10 e-mal addresses or phones. (We could, of course, parse out the XML--- but then you’re back to a data source that isn’t as fresh, and isn't near real-time.) For the physically modded SPOT1, an e-mail distribution list would, of course, solve the 10 e-mail / SMS limit.
Jonathan
But do we really need more than 24 hours?
On a separate note, the unmodified SPOT only runs in TRACK mode for 24-hours. For long duration, one may want the device to keep going until the batteries are depleted. That's another thing the physically modded SPOT does: runs for about 9 days 'pushing' messages out, all on one set of lithium batteries.
Just curious, why would we need more than 24 hours of tracking on a balloon that should be on the ground in 2-3 hours? Once we know where it is, we can take longer than 24 hours, if necessary, to retrieve it, but it certainly will not be moving for more than a couple of hours.
Unless you are assuming someone on the ground found it, threw it in the back of their car and is driving off with it. And if they are doing that, there is not much we can do about it, even if we know the location - are we really going to follow a moving vehicle into a house and ask about it? If they were trying to 'take it' and not notify us, I don't think we really want the confrontation about it.
So we either assume it is on the ground in 2-3 hours and we know where it is, or someone took it and they will hopefully call us - if they took it, but have no desire to call us, basically it is a lost (or stolen) capsule...
Distance
Just curious, why would we need more than 24 hours of tracking on a balloon that should be on the ground in 2-3 hours? Once we know where it is, we can take longer than 24 hours, if necessary, to retrieve it, but it certainly will not be moving for more than a couple of hours.
That depends on whether we're doing a simple up and down flight, or if we're going for distance. For example: https://secure.wikimedia.org/wikipedia/en/wiki/Spirit_of_Knoxville
Another example
Just curious, why would we need more than 24 hours of tracking on a balloon that should be on the ground in 2-3 hours? Once we know where it is, we can take longer than 24 hours, if necessary, to retrieve it, but it certainly will not be moving for more than a couple of hours.
That depends on whether we're doing a simple up and down flight, or if we're going for distance. For example: https://secure.wikimedia.org/wikipedia/en/wiki/Spirit_of_Knoxville
Also here: http://www.projectbluehorizon.com
Re: Distance
Just curious, why would we need more than 24 hours of tracking ...
That depends on whether we're doing a simple up and down flight, or if we're going for distance. For example: https://secure.wikimedia.org/wikipedia/en/wiki/Spirit_of_Knoxville
Okay, that sounds like a lot of fun, and something I had not even considered. So ignoring the cost of the electronics, what would be a reasonable cost for the balloon + gas for a trip like that?
For our normal up and down flights, I still don't think we need more than 24 hours, but I can certainly see the need for something like this on long duration flights. Of course, for that flight, we will need a small computer on board for telemetry, dumping balast, etc in addition to the tracking.
I have the balloon for a Cross-Atlantic attempt....
Well.....I was gonna wait until after our next altitude shot to propose it.....but, after our September 26th space shot, if our fledgling ballooning group is still intact, excited, and interested.....I was going to propose a long-distance, zero pressure gas balloon flight. This flight would have a completely different set of requirements, different materials, and a very different challenge. The balloon flies in a completely different fashion; it does not reach a burst altitude; rather, once the balloon is packed, expanding gas is vented through an open appendix. The balloon reaches a pressure altitude where it levels out, and it floats. And Floats. And Floats.
The problem becomes how to survive nighttime cooling. The sun goes down, the balloon cools, the gas contracts, the balloon sinks, and you end up in the ocean. You need a system to ballast to offset nighttime cooling, either via remote command, or automatically.
We have some ideas on that.
One of the links above is for the SNOX team. I have the balloon that would have been SNOX6, if that team had launched it. It was built with that trans-atlantic mission in mind. But, it isn't going to be SNOX6; we have it now; it can be what this team designates.
That's one of the reasons I was so keen to go with Tanner to WRAL, to chat with the weatherman about the jetstream. We have September to launch our high-altitude attempt with a reasonable chance of AVOIDING the jetstream and getting our capsule back without it going into the ocean. And then....if the team is intact.....once we progress into the winter months and the jetstream starts swinging far enough south where we could catch it....... we have the balloon.....the gas....the team.....and problems to solve.
It would be an outstanding endeavor; it is awesome what we could do, if our team is up for it!
Jonathan
P.S.: For a snapshot of what a zero-pressure scientific balloon can do, take a look at this:
http://www.youtube.com/watch?v=SC3yQuotbeQ
Re: Cross Atlantic...
Count me in as being very interested in that attempt.
I had run across the youtube video earlier and was amazed at just how big the payload was - much larger than our very small payload as it had very different goals when constructed.
But you mention you already have the balloon - awesome! About how heavy would the payload (including the necessary ballast) can we lift with the balloon you have - and how much helium would it take?
Distance attempt
Awesome! I didn't realize you already had the balloon.
I've already started looking at the avionics used by the Knoxville group (which is difficult, since they don't describe it very thoroughly anywhere). To do something like this will be a major undertaking, even with the balloon already in hand, but it will no doubt be awesome.