Thumb drive size altimeter with LCD display

[jsdemar]

If a barometric pressure sample is taken, just as an example, only once every half second with the sample period itself taking only a tiny fraction of a second to complete, the odds that the altimeter will actually sense the overpressure spike due to the ejection charge is low. Thus, both sample spacing and sampling aperture make it unlikely that the transient pressure will ever be sensed. Even if it did, such erroneously high samples can be filtered out via software.

Minimizing sampled data does not improve noise immunity. In the world of digital signal processing this assumption is one of the typical fallacies that people buy into. For slow sampling, it requires analog filtering before the A-to-D converter in order for this method to work. Therefore, the noise immunity is actually in the analog filter, not in the quick conversion "aperture" and not in the "odds" of hitting a spike. Fortunately for this altimeter, it does not have to be fast because it doesn't have to fire an ejection charge close to apogee.

The second reason I was calling "BS" is that none of this explains the claim that the pressure transducer is not susceptible to damage from ejection charge heat and debris.

-John

 

[NjCo]

The second reason I was calling "BS" is that none of this explains the claim that the pressure transducer is not susceptible to damage from ejection charge heat and debris.

-John

I'm sure a device like this will get some heat and debris from the ejection charge but it's not like there's nothing in between the motor and the altimeter at ejection. Most low power set-ups will have a baffle/recovery wadding as well as a chute/streamer placed between the source of the heat and the altimeter. I'm sure these will go a long way toward providing some protection for the device. I'm no expert but it would seem if you are not frying your thin plastic chute then I'd think the impact of the ejection gases on the altimeter sealed in a plastic case should be minimal.

 

[BEC]

Here are some quick and dirty pics of my new AltimeterOne.

The other unit in some of the pictures is a HowHigh SP. Both units weigh the same - 0.25 ounce - but the AO will fit in a smaller tube.

As for protection from ejection: the instructions show it loaded above the recovery system. The end with the attachment eye is the one where the opening for the pressure sensor is...and that end is shown up in diagram in the instructions. See page three here. In that situation I think if your recovery system survives, so will the AO.

I of course haven't had a chance to fly it, but I did do the quick and dirty functional check as Tim VanMilligan does in the video on the Apogee site (on this page) and got a pretty fair agreement on readings with the HowHigh SP shown in the pics.

I wish the version of vBulletin used here let me caption pictures.....

The first is as-arrived. The second is out of the bubble wrap. The third shows the unit slid partially into a BT-20 motor mount tube for size reference. The fourth and fifth are with a HowHigh SP.

 

[BEC]

Four more pics:

Shown plugged into my netbook and charging. Shown charge complete. Shown plugged in from the top side. Shown on my little digital 0.01 ounce resolution scale.

So far the only thing I don't like is the sealed in battery (the case is glued shut) so it's even worse than an iPod in that regard. Eventually that little lithium polymer cell is going to die (though it could be several years from now). When it does, I'd hate to have to pitch the AO... And yes, I know that by then there might well be something much nicer and less expensive available. But then, this is a guy who's driving a 1988 Mazda 323 with nearly 315,000 miles on it to work every day and who has at least one model airplane in flying condition that's 30 years old..... so I'm odd, I guess.

 

[rockosrocket]

Here are some quick and dirty pics of my new AltimeterOne.

The other unit in some of the pictures is a HowHigh SP. Both units weigh the same - 0.25 ounce - but the AO will fit in a smaller tube.

As for protection from ejection: the instructions show it loaded above the recovery system. The end with the attachment eye is the one where the opening for the pressure sensor is...and that end is shown up in diagram in the instructions. See page three here. In that situation I think if your recovery system survives, so will the AO.

I of course haven't had a chance to fly it, but I did do the quick and dirty functional check as Tim VanMilligan does in the video on the Apogee site (on this page) and got a pretty fair agreement on readings with the HowHigh SP shown in the pics.

I wish the version of vBulletin used here let me caption pictures.....

The first is as-arrived. The second is out of the bubble wrap. The third shows the unit slid partially into a BT-20 motor mount tube for size reference. The fourth and fifth are with a HowHigh SP.

Very informative, thank you. This seals the deal for me, I am ordering one right now.

 

[BEC]

Unfortunately you'll have a bit of a wait. I just got an email newsletter from Apogee which included the following:

Altimeter One news

The first batch of altimeters came in last week, and we sold out almost immediately. We are currently on back-order until about the beginning of the second week of June. Sorry about that. I didn't expect the demand to be this high for the altimeter.

If you would like to be included on our list to be notified when that next batch comes in, please leave us a note at: https://www.apogeerockets.com/Information_request.html

 

[rockosrocket]

Unfortunately you'll have a bit of a wait. I just got an email newsletter from Apogee which included the following:

Weak.

 

[NjCo]

Yeah, I just got the email for the latest Apogee video and in there it mentions that the demand for the altimeter was unexpected. It looks like almost all of the product he had went to preorders to preferred customers and newsletter subscribers. It was only up on the website for two days at most I think. More stock is expected in the 2nd week of June it said in the email.

By the way, the video is on how baro altimeters measure altitude. Check it out here:

https://www.apogeerockets.com/Rocketry_Videos/Rocketry_Video_40.asp

 

[bobkrech]

Bob, can you explain this "sampling aperture"? When I heard that it was a baro altimeter that you could just stuff into the rocket and tie to the shock cord, I dismissed it as crap. How can they expose a pressure transducer to the ejection charge without deteriorating the electronics?

I know Adrian's done it with the Parrots in some tiny rockets with a small shield on the cord between the motor and altimeter, but I'd be concerned about damage from continued use that way.

This would certainly be handy for itty bitty rockets, but in general I'll stick to a separate ebay.

-Ken

Ken

I might be mistaken, but the common MEMS pressure sensors that I am familiar with are susceptable to zero offset errors and possilble damage if exposed to an over-pressure that exceeds the yield of the internal beam. I'm assuming that by proper acoustic design, the manufacturer can restrict the conductance of the inlet so that little airflow will enter the innards of the altimeter in the 1 or 2 milliesonds of ejection charge overpressure. A sampling aperture with a time constannt of 0.2-0.3 seconds would be adequate for apogee detection, but would restrict the effects of the ejection charge over-pressurization.

Bob

 

[jsdemar]

Ken

I might be mistaken, but the common MEMS pressure sensors that I am familiar with are susceptable to zero offset errors and possilble damage if exposed to an over-pressure that exceeds the yield of the internal beam. I'm assuming that by proper acoustic design, the manufacturer can restrict the conductance of the inlet so that little airflow will enter the innards of the altimeter in the 1 or 2 milliesonds of ejection charge overpressure. A sampling aperture with a time constannt of 0.2-0.3 seconds would be adequate for apogee detection, but would restrict the effects of the ejection charge over-pressurization.

Bob

The manufacturer's website states 20 samples per second in one description and 10 samples per second in another. That would make it impossible to have a sampling aperture of 0.2-0.3 seconds. Another factor is that the typical ADC of a microcontroller does relatively fast successive approximation conversions (shorter than 1 msec). They also state a 19-bit resolution which would require an ADC external to the microcontroller or built into the pressure sensor.

The correct way to handle the noise and spikes is to oversample and filter digitally. For this type of device, however, it would make sense to have a 50msec analog filter time constant and do slow sampling, without concern for aperture time or pressure spikes. Cheaper slower processor and less power consumption. That method, however, would not scale up to an active altimeter that detects apogee and fires charges.

-John

 

[Gus]

BEC,

Thanks for the great writeup.

It's no wonder Tim has sold out the first batch. That thing's the iPod of altitude reading altimeters. Clean, simple, and so brain dead easy to use that a 4 year old could do it, and have fun doing so.

I expect that should get NAR certification pretty quickly.

Kudos to whoever designed it.

 

[BEC]

BEC,

Thanks for the great writeup.

You're welcome. I suppose I should post something about it over on YORF, too...but it would be kinda pointless until there are more available.

I MAY get a chance to fly mine this weekend, even though I'm going to an electric airplane meet rather than to a launch. In fact, I may well put it up on my instrumented airplane (Eagle Tree GPS and barometric altitude, among other things) as well as launch it a time or two.

 

[bobkrech]

The manufacturer's website states 20 samples per second in one description and 10 samples per second in another. That would make it impossible to have a sampling aperture of 0.2-0.3 seconds. Another factor is that the typical ADC of a microcontroller does relatively fast successive approximation conversions (shorter than 1 msec). They also state a 19-bit resolution which would require an ADC external to the microcontroller or built into the pressure sensor.

The correct way to handle the noise and spikes is to oversample and filter digitally. For this type of device, however, it would make sense to have a 50msec analog filter time constant and do slow sampling, without concern for aperture time or pressure spikes. Cheaper slower processor and less power consumption. That method, however, would not scale up to an active altimeter that detects apogee and fires charges.

-John

John

I'm suggesting that the hole in the altimeter case may be sized to permit an acoutstical filtering of 0.2 Hz. That's totally different than an electronic sampling rate or digital filtering. The reason why this works for apogee detection is that a rocket is within 4' of apogee for 1 second, and in 0.5 seconds the change in pressure will have equilibrated to 8% of the actual value resulting in negligble error in the apogee measurement.

The hole in the altimeter case limits the conductance and therefore hole size can be used to limit the rate of pressure change inside the casing, however I don't know if the altimeter uses this method of rejecting the ejection charge overpressure.

For example if the rocket has an apogee of 1000', 0.5 seconds before apogee the pressure is 0.015% higher than apogee. In the 0.5 seconds it takes to get to apogee, the pressure would have dropped 92% or 0.013% with a 0.2 second time response aperture, an error of less than 1'. If we assume a 100% ejection charge over-pressure for 0.002 seconds, the pressure rise with the same aperture size is only 1% which won't damage the pressure transducer, but a 1% change represents ~250' lower altitude at a 1000' apogee.

Bob

 

[NjCo]

For example if the rocket has an apogee of 1000', 0.5 seconds before apogee the pressure is 0.015% higher than apogee. In the 0.5 seconds it takes to get to apogee, the pressure would have dropped 92% or 0.013% with a 0.2 second time response aperture, an error of less than 1'. If we assume a 100% ejection charge over-pressure for 0.002 seconds, the pressure rise with the same aperture size is only 1% which won't damage the pressure transducer, but a 1% change represents ~250' lower altitude at a 1000' apogee.

Bob

I'm not an engineer so maybe I'm entirely missing the point here but if the rocket is within 4' of apogee for an entire second then why does this 250' lower altitude matter? Assuming 10 samples per second as another poster mentioned then 0.1 seconds before ejection the rocket will be essentially at the same altitude as the moment of ejection (lets assume ejection at apogee). Let's say ejection happens at 1000 ft. So what you are saying is that at the moment of ejection the altimeter will record an altitude of 750 ft. Is that right? But 0.1 seconds prior to (or 0.1 seconds after ejection when the altimeter is outside the rocket) that the recorded altitude was 999.5 ft or whatever and that altitude is what will be recorded.

So what about this vacuum issue that was mentioned previously? Someone mentioned that the overpressure from ejection wouldn't cause an issue but the vacuum generated when the nosecone leaves the body tube might cause an issue. Is this possible even with the recommended three sampling holes that are supposed to be placed 4 body tube diameters below the base of the nose cone? Aren't the sampling holes (in the rocket), the ejection charge and the vacuum from the nose cone all working against each other at the same time?

 

[jsdemar]

John

I'm suggesting that the hole in the altimeter case may be sized to permit an acoutstical filtering of 0.2 Hz. That's totally different than an electronic sampling rate or digital filtering. The reason why this works for apogee detection is that a rocket is within 4' of apogee for 1 second, and in 0.5 seconds the change in pressure will have equilibrated to 8% of the actual value resulting in negligble error in the apogee measurement.

The hole in the altimeter case limits the conductance and therefore hole size can be used to limit the rate of pressure change inside the casing, however I don't know if the altimeter uses this method of rejecting the ejection charge overpressure.

For example if the rocket has an apogee of 1000', 0.5 seconds before apogee the pressure is 0.015% higher than apogee. In the 0.5 seconds it takes to get to apogee, the pressure would have dropped 92% or 0.013% with a 0.2 second time response aperture, an error of less than 1'. If we assume a 100% ejection charge over-pressure for 0.002 seconds, the pressure rise with the same aperture size is only 1% which won't damage the pressure transducer, but a 1% change represents ~250' lower altitude at a 1000' apogee.

Bob

Yes, I understand the acoustic/flow filtering of a mechanical aperture, but you had said "sampling aperture" which provokes thoughts of sampling time in the digitial domain. Sometimes it's hard to remove the EE/DSP hat I've been wearing for 30 years.

As you mentioned, this is just speculation without seeing the unit. If BEC could take a close-up of the opening at the end of the AOne, we might have a better idea. However, the flow aperture concept requires the unit to be sealed everywhere else... unlikely for a unit in this price range.

Both the mechanical aperture and the signal filtering (analog and digital) will all combine for a net delay between actual pressure minimum and the ejection decision point. Add to that the lag from ematch firing to actual deployment. Without a predictive filtering algorithm, apogee ejection is always late in practice.

Good discussion, thanks Bob.

-John

 

[bobkrech]

John

I'm use to gas sampling and flow restrictors as well as digital sampling theory. I'm not always as clear as I should be.

Vacuum doesn't damage mems sensors, overpressure does. The ejection charge is a positive pressure so there is no vacuum produced when ejection occurs. My comments concerning the conversion of external pressure variations to internal pressure variations and then to altitude were for a situation where the acoustic frequency response was 0.2 Hz.

If there was a 10 PSI over-pressure when the ejection charge fired, I assumed an ~ 2 millisecond pressure burst. That's that's much higher an acoustic frequency that a 0.2 Hz acoustic frequency vent aperture, and the maximum level of response is only ~1% of the overpressure or 0.1 psi which represents a ~250' lower altitude reading reading below apogee near the ground. This could be used as the stop trigger for the unit.

Bob

 

[nu_blackshirts]

You guys are giving me a headache. :bangpan:

 

[NjCo]

This could be used as the stop trigger for the unit.

Bob

I saw somewhere that the unit doesn't "stop". It think this was on the Apogee website somewhere. The example stated that if the rocket caught a thermal and went higher under chute after ejection that the altimeter would keep recording. On the Apogee website it does say that the unit shuts down after an hour of inactivity.

 

[BEC]

If BEC could take a close-up of the opening at the end of the AOne, we might have a better idea. However, the flow aperture concept requires the unit to be sealed everywhere else... unlikely for a unit in this price range.

I can try. It's hard to get a clear picture of a hole in something that's shiny and black!

The rest of it IS sealed, or at least tightly glued together. The actual opening is about 1/16 inch deep and as wide as the tab which projects out of the end for the attachment point. This tab is actually part of the circuit board.

I don't see how a nose cone - ejected by the ejection charge - would pull a negative pressure on the altimeter inside and fool it into thinking it is higher than it is. Now if you pulled the nose cone off quickly enough after turning the altimeter on and putting it in - but before it was blown off by the charge - then I can see a need to reset the zero before going to fly.

I could experiment with that, I suppose.....

 

[BEC]

I don't know if this will tell the story that you want, but here are a couple of attempts to show the opening in the end of the AltimeterOne that accesses the sensor.

 

[McKailas Dad]

Now this thing is cool! I've been looking at a few basic (i.e.,inexpensive) altimeters, and this one seems like the cats meow. I'm going to keep an eye on this thread to see how it works out for you guys...

Has anyone put 2 or more different 'brand' altimeters in a rocket, just to compare accuracy to one another?

 

[BEC]

Over the past weekend I was able to fly the AltimeterOne on both an RC airplane (which was equipped with its own altimeters, both barometric and GPS-based) along with one of my Winged Shadow HowHigh SPs as well as a standalone Eagle Tree altitude sensor.

On the airplane (where I took off, then climbed as fast as I could to alititudes where flying the airplane became "interesting" - 1200-1600 feet) the AO and the HH SP gave very similar results (on one flight reporting a peak only one foot apart) and with good agreement with the other measurement tools. I tried them in the fuselage and also taped to the fuselage and then the wing tips.

The next day (Monday) I flew the AO five times on three different rockets including an Alpha III (where it fit inside the nose cone). In each case I had put static pressure vents in the upper body tube as directed and did not use a separate payload section. In each case I got plausible results from the AO (based on prior observations and/or simulations of these rockets) and even though a couple of landings were a bit rough it performed just fine.

Rough landings were the result of parachute tangles. Having an additional item in the form of the AO clipped to the screw eye on the nose cone and stuffed into the rocket above the 'chute along with most of the shock cord increases the chances of tangling, or so it seems based on those flights. This is something I'm going to have to think about a bit....

But - that said, I think I will order another and will use the one I have often. It sure seems to work.

 

[John Beans]

About a year ago I decided to develop a new altimeter, with the very deliberate mission of answering the "how high did it go?" question in as easy-to-use a way as possible. In a lot of ways, the AltimeterOne is not an altimeter for me (an engineer) or perhaps some of you, but for my kids and their friends. Their joyous moments aren't spent in front of a computer studying pressure and acceleration curves, but rather in those moments right before launch when we banter about "how high this one's going," and then the laughter and chase of recovery.

We needed one altimeter that would work for all of our rockets, not need a lot of care and feeding, and "just work." That was the impetus for what you see today, the AltimeterOne.

I'll try to address a couple of points that you've collectively raised in this thread. I'll also post my email (or you can comment on my site) so that you can ask me more if you feel like it.

Issue #1: Rides in the fuselage. Really? C'mon. Really??
Yep. We've been field testing this thing for months. If you try hard, you can mess it up. But you have to *really* try hard. It's rugged, and as for over-pressure, it can handle 10x atmospheric. (In general, when your electronics can handle more than most fuselages, you're in good shape.)

Issue #2: I still can't get over that it rides with the chute. What about false readings?
Probably 80% of what we worked on in flight testing could be roughly termed as "filtering," though it has less to do with Nyquist limits and Kalman filters than it does with handling sporadic events. It takes a lot of work to discover all of the little corner cases, combinations of conditions, and odd "once every ten launches" type events that can make software look stupid when they crop up. The list is long: breezes, popping open the rocket to take a peek right before launch, mach shock sweeping across a vent hole, delays between loading and launch (thus the 60 minutes), etc. Every time something like this cropped up, we figured out a way to filter it. You do need to be aware that you need to let it breathe to get a reading (whether in a payload or a fuselage). I'm a big fan of "more venting is better," and I think in practice it's hard to over-vent, a symptom of which might be difficulty with proper ejection. Hmmm. There's a joke there somewhere...

It wasn't discussed much yet, but the fact that you can use this in a rocket, plane, or kite is pretty cool, and somewhat unique.

Issue #3: How good is this thing? Is it just all show?
Don't let the "pretty face" fool you. This thing has some really advanced components, and is wicked accurate. I have a photo I still keep of the first four prototypes, along with a competing altimeter, as viewed through the window of my pressure chamber in its maiden run. All four altimeters read 14,390 feet. All four. The same. In a world where most common pressure sensors and analog to digital conversions occur with AT BEST 1.5-2.5% accuracy (including noise), this is a pretty cool result.

Issue #4: Why does it cost $49.95? I was hoping for $30...
Well, mostly see Issue #3. This is not "the cheapest way to make an altimeter." Instead, it is the cheapest way that I know to make it right. For me, that's tiny, rugged, accurate, and simple. "Easy" may be different for all of us, but for me, it also includes never having to buy and keep a spare set of odd $5 batteries on hand. And--though I realize that this is highly subjective--I think the flash and beep counting is kind of kludgey.

All of that having been said, it is reasonable to expect that the altimeter will come down in price over time, though I hope you agree that this is a pretty good value already.

Issue #5: I wish it had data logging and staging outputs.
I hear you, I'm like that, too. But that's not who this was for. There are some fine units out there that do those things now, as you know.

Feel free to ping me. I love to hear suggestions, comments.

Have fun,
--John

John Beans
President, Jolly Logic
[email protected]
www.jollylogic.com

 

[Gus]

John,

Thanks for stopping by and giving the great explanations.

As I said back in post 71, you've designed the iPod of rocketry altimeters.

Really a wonderful piece of engineering.

Congratulations!

You're going to sell a ton of these.

 

[BEC]

John,

I agree with Gus.... it's a great little device and I am quite likely going to buy at least one more (and probably stop using some of the competing devices I have been using - yours is just so simple to use).

I showed it to my students this morning and described in a little more length what I summarized in my previous post (#82). I had to stop as it started to sound like a sales pitch .

I will be using it with their rockets (after retrofitting vents where needed) at our end-of-year launch next Friday. It will also get rides in more than the one airplane it's already been flown on (in case anyone here cares, this is the airplane: https://www.rcgroups.com/forums/showthread.php?t=740824).

 

[bobkrech]

John

It looks like you really did your homework, and adequate flight testing.

Congratulations.

If the reproducibility between units is as good as it appears to be, it probably should become the standard NAR contest altimeter because it apears to be precise, rugged, simple to use, tamper-proof, and virtually indestructable.

Bob

 

[BEC]

Looks like the second batch has sold out at Apogee as well.

I've been flying mine every chance I could...and remain impressed. I'll have a little story to tell later....I found the limit on "virtually indestructible", I think, today. But I want to get in touch with Apogee and Jolly Logic first before telling the story here.

 

[BEC]

OK - here's the story.

Yesterday at the monthly BEMRC launch, I put my Altimeter One into one of my Alpha IIIs (it fits in the nose cone) and for some reason I now cannot fathom chose an A8-5 rather than the more appropriate A8-3 for a little flight.

It was a great straight boost, then it pitched over and started back downhill....and the ejection charge fired within 10 feet of the ground. Impact was on the gravel/dirt access road to our little site, not in the grass. The nose cone with the AO in it (along with the rest of the Alpha) hit hard.....and when I got there I expected to find smashed bits. As you'll see in the first picture, the body of the Alpha III is a writeoff (it'll get replaced with an appropriate length of stronger ST-9 tubing). The altimeter looked undamaged, but oddly it was reading zero rather than the expected about 180 feet.

I extracted it and noticed that the seam between the case halves seemed a little opened on the top but otherwise everything looked fine. I decided I needed to test it to see if it was still working...and stupidly didn't do the little test described in the video on the Apogee site but rather to immediately fly it again.

So, it was clipped to the screw eye of my BMS School Rocket, reset, an A8-3 was installed, and the rocket was launched as the last flight of the afternoon. It was a picture perfect flight with a good deployment of the streamer and a grass landing a bit south of the launchers.

When I got to the landing site I saw a perfectly reasonable "224" on the display, but also that the front half of the case had departed some time between ejection and landing. I have no idea where it went.

Last night I wrote to both Apogee and to John Beans and told them this story. I heard back from both of them by this morning (working on the weekend, eh, guys?) and will call Apogee to order a new case for the princely sum of $3 tomorrow. I will probably also order another Altimeter One so as to get one from the June 25th batch.

John and I are still trading email on ideas as to why, on the crash flight, I got a zero reading. Right now the working theory is that the power button got bumped either by the flexing nose cone or by an object on the ground before altimeter had written the apogee value in flash memory. Since the whole flight was about six seconds long (assuming a nominal delay) things happened pretty durn fast.

The pictures are of the crash site (recreated), a closeup of how the AO fits in the nose cone of the Alpha III, the now front caseless AO after that School Rocket flight and a closeup showing the reading.

This is a tough little gadget.

 

[John Beans]

Bernard,
Thanks for the great field report. My intention was to make this altimeter as tough as possible, so that people could just use it and not have to molly coddle it. There have been some truly epic "recoveries" over the last couple of weeks, most of which the AltimeterOne has emerged from intact. But we've also lost a couple of cases (like your flight).
Right now I'm working on a case modification to make it even harder to crack open. The failure mode it seems most prone to is seam splitting from a hard plastic-on-plastic "slap" (either from lawn darting or from an over-enthusiastic ejection charge in larger rockets that propels the altimeter against the bottom of the nose cone). It's much easier to shatter the glue in the seam from these types of sharp slaps than it is to crack the plastic case, because the case is a tough ABS type plastic that isn't brittle.
Thanks again for the report!

 

[BEC]

You know, it might be as simple as a less brittle glue (it looks to be CA). What about something like "Goop" applied from a small syringe?