Going Aiptek Small, cheap video camera that flies.

Getting old is not exactly my cup of tea.  Despite my dashing, youthful appearance, I'm making a rapid transition from "bright young man" into "doddering old fool."  Best I can hope for is to be a pleasant old fool.  Stiff joints, aches and pains, can't read the funnies without my granny-glasses, and I actually find some of the funnies.... well, funny.  A clear sign of brain deterioration. I'm exhibiting all the things I remember my grandparents complained about when I was a kid.  It stinks.  

But after long and careful consideration of all the options, I think I'm going to go with it.  Unfortunately, "going with it" involves having the body rebuilt in bits and pieces, as best medical science can do, to maintian the Epicurian rocket-scientist lifestyle that makes life worth living.

So over the Winter Holiday (aka "Christmas Break" to the non-PC) I had a part rebuilt.  Which part you might ask?  Well, thanks for asking, but I won't say.  

OK ok.   Since you insist, I'll drop one clue:  whenever some person, object, or event is referred-to as a "Royal PIA,"  I now understand.  

Blessings of this is that I've had many opportunities to hang around the pharmacy waiting 15 or 20 minutes for a highly-trained pharmacist to count a few pills into a bottle and stick a label on it.  Not much for sitting and waiting, especially sitting, I wander around.  After verifying for the 100th time that they no longer keep Saltpeter on the shelves and squeezing all the squeaky-toys, I limp toward the photography department.  

And there I found it.  My next great love.   The Aiptek MPVR  handicam.  

    

Three virtues attract me to it:

1.  It's cute.
2.  It's cheap.  
3.  It's small.
4.  It's all solid-state.

Good thing I'm not counting the pills.  

Video is stored on SD memory cards, so there is no spinning head, disk or flywheel to be perturbed by sudden forces of nature.  

It will run for 90 minutes before the battery needs recharging and my 2 gig card SD card fills up.  Oh, the SD card is extra.  Micro Center had them on sale for $16.00 recently.  The camera has a little internal memory, but only enough for a couple of minutes of video.  

I had already purchased a tube for it.  After seeing the camera rocket that Richard Creamer and Steve Ghioto had flown back in August I ordered a section of "polycarbonate tubing" from McMaster-Carr (Stock # 8585K21, $11.46 per linear foot as of 4/10/07)   This tube is a nominal 3 inch outside diameter, and a tight friction-fit inside the 3-inch Quantum Tube I use in the Sugar Rush.  Inside diameter is 2-3/4ths inches. Steve and Richard had used the "One-Use" digital camcorder available from CVS pharmacies.  It is a remarkable little camera, but the official deal isn't very good.  You "buy" the camera for $29.95, shoot up to 20 minutes of video, return it to the store and they will burn your video to DVD for another $12.95.  And they keep the camera!  No doubt to erase, re-package, and "sell" again.  $43.90 for 20 minutes of mediocre-quality video?  I think not.

But these rocket geeks are clever folks, and both Richard and Steve know electronics.  Richard modified the camera, adding a USB port, so that it is no longer a "one-use" device, and can download, erase, and re-use the camera at will.  It runs on two AA batteries, so power is not a big problem provided that one starts with freshly-charged, high-capacity batteries.  The ones that come with the camera  tend to run down very quickly.  He also modified it to save better-quality video, albeit for a shorter time.  No big deal, just start the camera, run... no, walk calmly back to the safety zone, and fire it right away.  

So the new deal is much better.  $29.95 plus a little work for an all-digital, indefinately reusable camcorder?  Yes!  I'll bite on that!  

But I am not much of an electronics geek.  I can plod along, but haven't made it over the hump.  This kind of project could become very time-consuming, and time is a precious commodity for me.  Also, I've learned that CVS has redesigned the camera to make it more resistant to hacking.  I thought briefly about "renting" one of the CVS cameras as per the official deal, but when I picked it up and looked at the poor little thing, just couldn't bring myself to do it.

Thus the Aiptek revelation.  It's more money, but an outright sale, better quality video, and pretty darned cute.  And being in medical mode I'm used to plunking down large chunks of money, so as an "impulse purchase," I buy one.  It looked to be the right size to fit inside the polycarbonate tube.

Wrong.  The body of the camera fit the tubing, but the lens bezel stuck out about 1/8th inch too far.  I briefly thought about removing the bezel, but was not sure what would fall off when I did so.  

Sometimes you can ram two problems together and get a solution.  I really like it when I can pull that off. The other problem?  Shooting video through thick, curvy plastic. I was not enamored with the idea of shooting video through the tubing wall, since this "clear" tubing is not perfectly so.  It can't help but distort things.  Solution?  I decided to make a hole for the lens.   Using a 1-1/16th inch spade bit, I drilled a hole centered 2-3/4ths inches from one end of the tube.  The spade bit made a rough hole.  Next time I'll try a hole saw or something nicer.  But a few minutes with a round file made it acceptably smooth.   Then the table saw was used to cut a slot down to meet the lens hole.  This allows the camera to drop into the tube easily, with the lens bezel extending about halfway to the OD of the tubing.  It fits pretty well, with the camera "hanging" in its slot.

Here is another puzzle to ponder... The camera turns on when the LCD panel is opened.  Nice.  Open it up, it comes on.  Handy.  When the LCD panel is closed, it turns off.  Problem.  Big problem.  There is no way the camera will fit in the tube with the LCD panel open, and there seems to be no way to keep it on with the panel closed.   I searched the menus for a way to disable this option.  None.  I looked in the cracks and crevices to find the switch.  Couldn't see it.  I started to take the thing apart.  After removing a few tiny screws I was interrupted, and in the interim realized the probability of destroying the camera is close to 1.  So I put the screws back and the camera lives to record another day.  In desperation, I even read the manual.  Can you believe it?  The MANUAL!  No clues there either.

Then when fiddling around, I discovered that the LCD panel can be swiveled all the way around and folded back into its spot, nice and neat!  This returns the camera to a form very closely resembling it's OFF mode, but with two significant differences:  1)  It stays on!  2)  I can see the LCD panel.  That was a revelation.  Now I can make it work, and without even violating the warranty.   See the headphone jack?  I'm using it to play oldie-goldie MP3 files as I take this picuture.  It does that too.

Now how to hold it in the tube nice and snug....

Funny how often the right thing jumps into my life at just the right time.  (I'll tell you how I met Teresa one day.)
 

Cruising around the cheap-tool shelf at Tractor Supply, I succumbed to another impulse purchase.  This time, it's an adjustable hole cutter.  It was only $9.95.  Those marketing folks are so clever.  If it had been $10. I would have walked put it back. It looked pretty cheesy, but the price was right so I walked out with it.  Well that's not exactly true...I stood in line, paid for it, then walked out.  These transactions go a lot smoother when one hands the cashier money or some acceptable substitute.  So I'm told.  I haven't tried it the other way.  But I know some folks who have.  They do Community Service hours under my "supervision." Note that by now I have two of these hole cutters.  That's a clue.  It works better than it looks.  The hole cutter is infinitely adjustable within its range, provided one has the infinite patience to adjust it, which can take several test-holes.  It can cut holes in 1/4th inch hardwood plywood with reasonable ease and a bit of smoke.

A helpful hint:  I got much cleaner holes by cutting halfway through, turning the wood over, and cutting the disk free from the other side.  

So I can make plywood disks any size I want!  Lots of them!  I went kinda wild cutting disks.  
 

Here are a few, cut 2.8 inches in diameter, just a bit too large to fit in the polycarbonate tube.  We will fix that later.   Each disk has a 1/4th inch dia.pilot hole.  This is handy for gluing them together using a 1/4th inch screw and wingnut to clamp them together.  The pilot hole is tight on the 1/4th inch dia. screw, so I cleaned it out with a 1/4th inch wood bit.  Now it fits smoothly and snugly. The stack to the left is 6 disks that have been spread with Elmer's glue.  Or Titebond, I forget.  The screw holds them tight until the glue has set.  Then the plug is sanded on the belt-sander until it slips easily into the polycarbonate tube.   The plan is to put two of these bulkheads inside a section of tubing with the camera in between, so that it can't slide fore and aft.

But just squeezing the camera in between these plugs doesn't seem right.  It could slip and tip against the smooth wood.   So I pull out another lucky find.  I think this came from from Pep Boys:  Bondo Fiberglass Resin Jelly.   It's much thicker than the usual fiberglass resin, much like an epoxy gel - which could probably be used instead. It mixes up just like Bondo body filler, which it resembles, it's just heavier and stronger.   But you know, Bondo is fairly tough and might work OK.  I may try it on the next one.

First, I want to make sure I can get a grip on this thing when it's installed in the tube.  So I drill a recess in one end, install a 1/4th inch eye-bold from the other, and epoxy everything in place.  Then a strip of poster-board is taped around it to create a collar, creating a "cup" about 1/2 inch deep.  This will hold the resin.

The camera is wrapped with several layers of plastic wrap.   I used red wrap so it is easy to see in the photos.  I'm sure clear wrap would work just fine. Posterboard cup is filled almost to the top with resin gel, and mounted on a weight so that it is level and stable.

And the two are mated!   I'm sure that little camera is frightened... I would be! But after about 30 minutes the gel has set, the camera is removed, unwrapped, and found to be OK.  No resin seepage, and the resin did not dissolve the plastic wrap....guess I could have checked for that possibility before risking the camera but I didn't.

It might have been a good idea to use some kind of jig to make sure the camera was square with the plug, thus would fit straight in the tube.  I just eyeballed it, and in fact it is a little crooked but not enough to matter.   Removing the paper collar, I get a rough impression of a rough impression.  A minute on the belt sander improves the flat surface and trues up the outside edge.  I wrapped sandpaper around a 1-inch dowel to sand the groove a bit.

The plug is sanded a little more so that it slips in the tube without resistance.  It seems willing to clamp the camera down, holding it against the bottom of the cutout.   The plug is still somewhat rough and ugly but that should not matter.... Because I'm going to dress it up!

Here I have found a scrap of green green Lycra that will match the shocking-pink Sugar Rush nicely, don't you think?   Sprayed it with 3M Super 77 glue and quickly pressed the plug onto it.   (I use this technique to resurface mouse-pads in the lab... works great!)

   

The plug is inserted in the polycarbonate tube as far as it will go, and a sharp knife used to cut the excess cloth so it can be pulled off.   The cloth creates a snug friction-fit so that the plug slides in smoothly without scratching the plastic too much.  Not that it matters.  .

Removing the plug from the tube, I press the head-end of the camera into its recess, hoping that the glue will make it stick to the plug and thus conform. (This photo is a reconstruction after the event - I covered the camera with plastic wrap during the real thing) The cloth stuck pretty well on the head end, but when I made another plug for the lower end I took too long and the material did not stick in the grooves.  It works fine anyway, just doesn't look as good.

This one doesn't look all that great either, with a bunch of dark spots where excess glue seeped through the thin material, but it's stuck!

As mentioned, a bottom plug is made in the same manner, just using the bottom of the camera to mold the resin instead of the top.   I cut the plastic tube 8 inches long, to allow an inch or so more length than the camera and both of its plugs.   Here are both plugs, friction-fitted.  Astute observers will note that I finished the bottom plug before the top one.   Doesn't bother me.  Since that surgery, I am no longer astute.

As long as I'm just handling it on the ground, those plugs with the cloth skirt are a nice friction-fit.  But I don't think that will cut it during a launch.  

For the lower end, I squeezed the tube into a section of 3-inch dia. Quantum Tube, drilled 4 pilot holes and inserted small 3/4th inch long wood screws.   These are long enough to penetrate deeply into the wooden plug, locking it, the plastic tube, and the body tube in place.   It took a bit of wrestling to get the polycarbonate tube into the Quantum tube - it's a tight fit.  Good.  The camera mount is part of the body tube so it needs to be stiff.

Things are not so tight at the top end.   Because of the slot, the top end of the tube can be compressed a bit, so it's much easier to squeeze inside the Quantum tube.   That's a good thing, since it needs to be squoze and unsquozed every flight.   I will compensate by using more screws.

Since it is a little bit looser, I used eight 1/2 inch long screws instead.  Shorter screws go in and out more quickly.  Guess I should get wedgehead screws and countersink them for better aerodynamics, but this is what I had on hand at the moment.  Flatter screws might cut down the wind noise too. As in the base plug, each screw goes through the Quantum Tube, through the polycarbonate tube, and aways into the wooden plug to lock them all together.   As you can see, I cut a peep-hole for both the lens and for the control buttons on the back.  These were drilled in a solid piece of tubing, which was then cut at that level.  This seemed much easier than trying to drill off the edge of a piece.

So now that the camera is locked down in the tube and the tube is locked into the airframe sections, I can worry about how to get the darned thing recording when that magic moment arrives. The solution is crude.  With the camera mounted firmly, I marked a spot over each button, removed the camera, and drilled a small hole in each spot.  This lets me push the buttons with some kind of poking stick.   I try to remember to take a bamboo skewer out to the launch pad, but often forget.  Fortunately my altimeter-arming screwdriver serves this purpose OK.

Actually, only one of these holes is necessary.  That's the one on the top-left, allowing access to the "record" button.  The top-right could also be useful, if I remembered to use it.  That's the one that takes high-res still shots.  I should remember to do that when the camera is up on its stand ready to go.  This little camera takes very good stills when everything is right.  It takes very bad stills the rest of the time.  But they are always better than frames captured from the movies.  

It looks rather nice when fully assembled.  But I gotta get prettier screws.

   

So here is the anatomy of the Barbie rocket.  

Nose cone is attached to an 8-inch section of Quantum Tube with 4 short screws. The 8-inch tube slides over the upper 2 inches of the clear camera tube, where it is secured with 8 short screws.  These screws penetrate into the forward wooden plug, securing the head end of the camera. Camera lens pokes through a slot, so that video clarity is not compromised by having to view through the "clear" tubing.   4 inches of clear tube is exposed, allowing view of the camera's buttons.  This makes operation of those buttons much more certain, and provides a view of the LCD screen to determine the camera's operational status.   The aft end of the camera tube is mounted in a 15 inch section of body tube, secured with 4 longish screws.  The aft camera plug is mounted there, and a parachute shock cord attached to its eyebolt.  The open area of this tube serves as the forward ejection bay. Electronics bay is modular, with a wooden "cage" which holds 2 electronic devices.  Each device is mounted on its own panel, has its own battery, and the different panels can be interchanged rather easily.  Currently I am using an ARTS on one panel and a G-Wiz on the other for redundant ejection.  (No more core samples for me!  I hope!) Fin section has a 54mm motor mount which can take up to a 7-grain motor, plus enough open space for another parachute.   The longest motor I've used in it so far is a 5-grain.  Hmmm....

So here is the awkward part.  The camera must be TURNED ON when I install it in the rocket.  It doesn't need to be recording, that can be controled with the poking stick.  But it must be turned on when inserted because once in the rocket, there is no way to access the power button.  

It's not a real big deal, but it does take a little time, and at a time when time is precious.  I get the rocket all ready to go, altimeters, parachutes, ejection charges, motor, all installed.  Then go check in with the RSO, get them to call in for a waiver, go to the launch pad and install the camera.  That means turning it on, wedging it into the lower plug, pressing in the upper plug, positioning the nose cone section, and screwing in all those little screws.   It takes 2 minutes, not counting time spent looking for dropped screws - I've learned to take a few extras, and a magnet.

Then when the ignitor is installed, and the rocket is up on the rail, I would stand on a tippy little stool, poke the RECORD button, and strain to see if it were actually recording.  My fear is that I would "bounce" the button, stopping the recording.  More recently, I've started the camera recording with the rail horizontal, where it is much easier to work on and to see.  I can even cover the camera section and my head with a towel to reduce the glare, see if it is recording or not, and get the guys at the flight line to laugh at me.  Then it is raised to firing position.  I've gotten some good footage during the raising too!  

But I may be overdoing it.  After all, the battery and SD card can record for an hour and a half, so it isn't all that necessary to wait for the very last minute to turn it on.  I just might need a spare battery or two for subsequent flights. They are $20.

For clarity, here is a visual. If you are still astute, you will notice that I'm arming the altimeters, not the camera..  Once the ARTS is beeping cheerfully and the G-Wiz is singing it's song,  I'll start the camera.  Then Mike and I will raise it into position.   Just imagine me being 18 inches to the left and poking the camera button with a small screwdriver.

Hey!  Wait!  I'm a photo-geek! I'll do a quick/dirty edit.   No sense using up imagination for something this trivial.     Mike Harris is very patient to put up with this nonsense.

Since the upper section of this rocket can be rotated, I have the option of turning the camera to view whatever is most interesting.  Sometimes it is the launch control area with everybody waving "bye bye!"  But I often prefer to aim it at other rockets so that it can record them taking off.

Here are stills of other rocket liftoffs, captured by Aiptek as the Sugar Rush sat on the pad waiting its turn.   Nowdays I request that it be fired last, so as to catch as many liftoffs as possible.

To date (May 07) I have launched this camera 10 times.  Every time it got good video and lived to tell the tale.  This included one hard landing during its fourth flight at Orangeburg where the main did not deploy and it landed on hard soil with just the drogue.  The rocket suffered minor damage, but the camera was fine.  With the help and encouragement of Tom Binford and Jonathan Carter, I patched it up and flew it again that day.  

So here are the launches, including videos from most of the flights: NEFAR 2/07 - 2 flights, one flight video on web page NEFAR 3/07 - 2 flights, 2 videos   Tripoli SC 3/31 and 4/1/07 - five flights, three on page NEFAR 4/07 - one flight, one video

$ New Camera $

It seems that my wild spending spree was short-lived.  During my short hospital stay lots of concerned, friendly, strangely-dressed folks converged upon my bed.  One after the other they poked, prodded, examined, questioned, informed, pierced, injected, and poked some more.  Every word, every gesture had a price tag hanging from it.  Every one of them wanted a tip.  Not right away, mind you.  Didn't have to.  They have a talented and enthusiastic billing department to back them up.  

So I endured a period of forced frugality while the tide of bills rose, fell, and rose again, washing my accounts and leaving them to bleach in the sun.

Funny how things turn around.  The bills receded, paychecks continued to arrive, and several unexpected wads of money came my way.  So yesterday, while heading home from the populated lands to the south my extravagant mood returned.   I diverted my Ford Tortoise back to Walgreen's, where another love lay waiting.

This model is the MDV-CB It's the same shape and size as the old one, but a little better.  It offers slightly higher resolution, an internal battery charger, and ...(click here for drum roll, please)...a remote control! Good news is that the remote works through the clear tubing, so it may not be necessary to drill poke-holes in the next mount I make.  I suspect that much of the howling wind-noise in flight videos is from those holes.  Guess I could put tape on them for the next flight and find out. But there is a serious omission!  There is no POWER button on the remote.  So I still have to turn it on before mounting it in the tube.  Oh well.  Maybe on the next model.

What else?  

This system is not yet optimal.  Here are some possible imprvements.

1.  An immediate desire is to arrange for an external power switch.  The camera has a power switch that can be used to turn it off and then on again, but it is mounted under the LCD screen!  So the LCD panel must be open in order to access in in its current position.  But with a careful disassembly, the switch might be relocated to a more accessible spot.  Given the compact nature of this device, it seems unlikely that much rearranging could be done inside.  So maybe the switch wires could run to a jack that dangles outside the camera, which in turn could be plugged into an appropriate switch mounted in the plastic tubing.

2.  I'm hoping someone will come up with a hack or firmware patch that allows the camera to save uncompressed video.  I am convinced that the lower quality of video is due to in part to high compression.  I would be happy to give up long recording time for higher quality.  The new camera model can take up to a 4 gig SD card.  That should be enough for about 15 minutes of uncompressed .AVI.

3.  A look-down capability might be nice.  I like the current horizontal view, but to see the ground recede at high speed would be a blast.  I have a prism from an old pair of binoculars that might work, if I can figure out a way to mount it.  Or a first-surface mirror might be better.  Downside is that there would not be much video from the descent phase, since the camera will be looking at the sky.  Unless I reconfigured the airframe so that the camera section descends base-first, or perhaps horizontal.  I might try this next time it needs rebuilding.

4.  A haze filter could provide clearer video, since this thing tends to be looking through lots of moist air when it flies.  Hey!  That might be cheap and easy!  I wonder if filters are available for this camera, or if I could adapt one from another camera?  Some research seems in order.

5.  Separate parachute for the camera section of the rocket should allow it to capture better video while descending.  At present, most of the descent is with a small drogue trying to slow both camera and aft sections of the rocket.  Result is a wild ride, with the camera tumbling every whichaway.  In fact, I have the rocket set up for separate deployment like this, I just haven't had the courage to try it.  The camera would be on its "main-and-only" parachute from apogee on, thus it could drift a long way.  

6.  Thus I have thoughts about dual deployment dual deployment.  Instead of having both altimeters in the lower section, put one in each section.  Both the camera part and the propulsion part would have its own drogue and main.  Both altimeters could be wired to fire the apogee ejection charge, to provide a degree of redundancy.  This would require a serious rebuild, or perhaps a new airframe.  

7.  A way to turn the LCD off.  This would increase battery life dramatically, probably more than double it.  

Even with these wishes, the camera has been fun, and works pretty well.  I find it useful for other tasks, such as recording presentations I've been doing so that I can wince at them later.  All I have to do is find a shelf, turn it on, stand it up, and do my thing.  It is inconspicuous, and records well enough for an informal review, or documentation of what happened.  

I hope to get in a few more flights soon.  I've seen glimpses of alien spaceship in several of the videos, so I know they are about to take the bait.  Hopefully they will return my rocket with some interesting video.

---

Jimmy Yawn
jyawn@sfcc.net
rev 5/20/07
Return to Recrystallized Rocketry home page