I know that most everyone would rather just fly rockets, have altimeters that tell us how high the rockets went, without having to talk about or know about how altimeters work or whether they're accurate or not.
Me too.
But the subject keeps coming up when someone flies 2 altimeters and they're 10% different, or 10% different from their GPS, and especially at a time like now, when Tripoli is about to change the rules about altimeter use in record attempts.
The good news is that baro-based altimeters are getting more accurate, and GPS are getting cheaper and easier to use, so it's easier than ever to measure your rocket's altitude in a way that you can believe the results. The bad news is that GPS still isn't a complete answer because of its size, cost, and the tendency of some (most?) units to lose satellite lock until well after apogee. The other bad news is that no matter how good baro units get at measuring pressure accurately, converting pressure to altitude with better than 10% accuracy will never be a no-brainer, despite what some altimeter manufacturers will tell you.
I won't go into GPS units much here because I don't know all that much about them. But I do know about baro-based altimeters. To get an accurate altitude reading from a barometric altimeter, 2 things have to happen:
1. The altimeter has to measure the ambient air pressure accurately
2. The air pressure reading has to be converted into altitude accurately
#1 is straightforward to verify using a ground test in a vacuum chamber. There are companies that do nothing but calibrate instruments, and they follow lots of standard procedures to objectively measure the accuracy of the instruments. One thing that people should know is that pressure sensors, like almost any kind of measurement device (even a ruler) are affected by temperature. Different altimeter manufacturers have different strategies for dealing with the effect of temperature on pressure, and some are more effective than others. But unless a calibration is conducted over the same temperature seen in flight, you won't really know what the accuracy of the unit is. A good calibration will be performed over multiple temperatures in order to measure this effect. I have been working with Jim Wilkerson from Tripoli, who is arranging an altimeter test program, to use a shop that calibrates over temperature, and has offered to do a free set of calibrations for Tripoli.
#2 is impossible without more information than is available to the board you fly in your rocket. That's the bad news. The fact is that the atmosphere is always changing, and so unless that little board you fly in your rocket has some information about what the atmosphere is doing, the best it can know is how to convert pressure to altitude if the pressure is changing according to a typical profile. Unfortunately, the typical profile that all altimeters assume (the standard atmosphere model) can be off by 10% or more, even if you're just measuring relative altitude above the ground. The effect is most noticeable on hot days, in the middle of the summer, when all altimeters will read lower than the actual value. There are corrections to the standard model that can cut the error down from 10% or so to 5%-6% or so, based on the air temperature at the ground. But 5% error still isn't very good. The good news is that NOAA directly measures the relationship between pressure and altitude, at about 100 sites across North America, twice each day, rain or shine. They launch weather balloons that use either radio sounding or GPS to measure their altitude and speed, and an accurate pressure sensor to measure the pressure. Even better, the data from all these balloons is easily available at the following website:
http://weather.uwyo.edu/upperair/sounding.html.
If you click on one of the stations, the most recent sounding data will pop up on your web browser. The first column is pressure, and the second column is altitude. I have made a little Excel tool that anyone can use, with any barometric altimeter that assumes the standard model (I think they all do), where you can cut and paste that table in, you can enter the altitude that the altimeter beeped out, and the tool will tell you the actual altitude was if your altimeter measured the pressure perfectly. If you're fine with 5% or 10% error in your altitude reading, don't bother. But if you are interested in accuracy better than that, this tool can provide it. There will still be some error due to the fact that the balloon wasn't launched at the exact time and place of your flight, but a quick comparison of different balloon soundings shows that the error from these effects are on the order of 1%-1.5%. So with a good barometric altimeter, that measures pressure with 1% accuracy, you can get an overall accuracy of around 2%-3%.
The tool I was referring to is available here:
http://www.rocketryplanet.com/forums/attachment.php?attachmentid=937&d=1215289086
When you can get a GPS that fits on a 1" x 1" board, costs $100, can be powered from a tiny battery, and never loses lock during boost, then barometric altimeters will be obsolete. That day will probably come, but until then, here's a way to get 2%-3% total accuracy from a good barometric altimeter.
I think we really stand at a crossroads in the altimeter industry. Small, cheap, and reliable altimeters have been probably the single biggest boon to the hobby. There are some incredible technologies coming out, and some incredible products introduced by our friends (and members).
But here's my take, harsh as it is..... I don't care about 1% vs. 3% vs 10% accuracy with respect to true altitude. I am not landing a plane on a runway. I want 1-2% accuracy between altimeters - and I don't want to send my altimeter back in for re-calibration every 6 months.
If I fly a Brand X altimeter, and someone else is flying a newer model Brand Y that uses a different algorithm, or external compenstion, then it is not a contest.
Temperature compensation, particularly with additional components open up a whole can of worms that vendors would probably do very differently. When a rocket can travel thousands of miles an hour, the latency of the component's ability to measure temperature can be a huge issue.
GPS may not be the solution either. They are not fast enough, they lose lock, and there can even be stray data that appears correct but is way off. I have had earlier GPS data-loggers the had stray LAT/LON number that were way, way off. If the position can jump, then the altitude probably can too. NCR is fortunate that everyone is using the Astro, but if a new model comes out with a different data filtering process then we may see disparities with GPS too.
One thing I can be very sure of is that I glad I'm not on the Tripoli Altimeter committee. I think it will be a difficult year or two as we move to the newer, and more capable, components.
My thoughts on contests in the short term would be to use NAR or Tripoli approved altimeters for the small diameter stuff, and GPS (perhaps limited to Astro or SiRF). Yes, I know this is basically passing the buck, but what else can we do? Should we force everyone to fly the same altimeter for contests?
Soooo much depends on the rocket being flown. Adrian, you made some comments about size, weight, etc. that are clearly appropriate to a 24mm rocket that weighs 100-200g. As one moves up the alphabet, size and weight quickly become less and less of an issue - in fact, the pendulum can swing the OTHER way (at least for mass).
Case in point - the last rocket I flew had enough room to fly a substantially larger payload than what it did -- the nosecone had a lot of area that got foamed. If any simulation of any software is to be believed, I was somewhat under optimal mass... so in that scenario, size and mass of the GPS really didn't come into play. Disclaimer - My ebay was about the size of a can of tuna - I didn't have any extra room in that part of the rocket. The GPS was up top.
Overall LENGTH of a rocket is always key - shorter is better, less drag. As you get into 98mm, 75mm, and probably many 54mm builds, there is a fair amount of room to play with both geometrically and with regard to mass.
Regarding the mass issue - I have indeed flown the same rocket with identical motors and the heavier rocket has gone higher. There is, of course, a "sweet spot" but there is indeed an optimal mass for any particular rocket / motor / environmental condition. But there we go again, steering dangerously close to another thread. My point is that overall mass is definitely germane to this discussion.
Ed,
I agree with your goal, and I agree that if you could get two identical flights to read within 1%-2% without worrying about absolute accuracy, that would be fine. No argument there. The problem is that the same issues that affect absolute accuracy also affect relative accuracy.
Let's take an example where John sets a record on a hot day in July. Just to make sure that there is no issue with the altimeters affecting the records, you borrow the same altimeter that John flew for your attempt. You get your rocket ready and fly on a chilly day on Oktoberfest. Your rocket design is nearly identical to John's, such that it goes to the exact same altitude. Your altimeter will beep out 5% higher than John's flight even though you're using John's altimeter and it actually went to the same altitude. Here the culprit was the pressure vs. altitude changed between the two flights, an effect that could be eliminated using the balloon data.
Now take another example. Two people are flying their single shot attempts on the same calm, sunny morning. They have the same make and model of altimeter, and in fact they even went to the trouble of putting both of them in a vacuum chamber together and verified that they had the same readout. No way for the altimeters to read differently now, right? Well, one guy is flying a white rocket and the other one is flying a black rocket, so the interior temperature of their av-bays are 30 degrees different as they sit there on the pad. If the altimeter they're using has a temperature sensitivity of 2% over 30 degrees (like the Parrot would if it didn't have on-board temperature compensation), then one rocket's altimeter will read 2% higher than the other, even if they went to the same altitude. Here the error mechanism is that the altimeter's pressure readings shouldn't change with the altimeter board temperature, but it does because of an altimeter design weakness that could have been caught with chamber tests over temperature.
There's nothing wrong with ignoring these effects, but if you do, your expectation for even relative accuracy should be in the range of 5%-10% instead of 1%-2%.
John,
I agree that in the larger rockets, GPS isn't a volume or mass problem, and if you're flying over 30kft, the extra $150-$400 shouldn't be a show-stopper, either. But reliably being in lock at apogee is still an issue. Maybe there's already a GPS model that reliably keeps lock at apogee on those sort of flights. But since some of them are limited in altitude, it may be awhile before there are enough flights over 30k-50k to get confident in any particular model.
... But reliably being in lock at apogee is still an issue. ....
Adrian, I think everyone concurs that the GPS will not return the exact apogee of the rocket – it will almost assuredly return something a bit less, because odds are you will get a lock on the way up or on the way down. Personally, I’m willing to live with that because I know I went at least that high, and I know that other subtle nuances like the color I painted the rocket or the air temperature on the day I flew it didn’t inadvertently return something more than I got. I'm good with that.
John, with your high flight, did it look like you got any dropouts in the Astro data?
No real gaping holes (other than most of the "up" part) 😯 Values were returned from on the pad, then a large gap occured during ascent.... and then ?30 seconds later more values came in that were incrementing higher, and then they started to decrease. As such, I know I got lock on the way up.
My handheld GPS, also made by Garmin, loses it's lock when I use it on faster aircraft. I once read that many of the units in commerce have an auto shutoff at something like 200 knots? I believe the thought was that they could not be modified for terrorist uses -- though I'm guessing you CAN get units that lock at higher speeds, because surely jets use GPS?
BTW, someone told me the Garmin DC20s can be set to return values at shorter or longer intervals. The shorter the interval, the less data you can store and the quicker the batteries deplete. I believe the default on the DC20s is 12 seconds, but I don't have the stuff in front of me now -
The DC20 default interval is 5 seconds, but can be changed to either 10 or 30 seconds. I wish 1 second was an option.
All I have to say about the altimeter/GPS argument is that neither is absolutely accurate. However, even an absolutely accurate altimeter wouldn't help a lick in finding a rocket you've lost sight of. That, to me, is the true value of the DC20/Astro combination.
The DC20 default interval is 5 seconds, but can be changed to either 10 or 30 seconds. I wish 1 second was an option.
All I have to say about the altimeter/GPS argument is that neither is absolutely accurate. However, even an absolutely accurate altimeter wouldn't help a lick in finding a rocket you've lost sight of. That, to me, is the true value of the DC20/Astro combination.
I wish we had 1 second intervals too!
I use a combine both to derive altitude in my own mind.
Before I had the minis reprogrammed, they read right with the dog coller (within 75 ft.anyway) After, they were 10% low on several shots. So.......
Am I supose to take my SSSS shot on thier report? Or the dog coller report? 😕
Hey!
I will tie a 3 mile cloth tape measure to my dart, and we can get a true reading! 😉
Latest Post: Hello World Our newest member: amiecannan4271 Recent Posts Unread Posts
