Depth Test IIMay 25 2014
Accept contributions for your expediton by providing us a few details. We will create an account on your behalf at WePay. If you haven't already registered with WePay, they will send you an email to complete your registration.
Here is the after pic right beside the before pic. Sooo convenient!
Large pressure gauge. It sprang a leak sometime after the cameras were eaten by the sea. (I guess I haven't mentioned that yet) It's failure is not surprising to me. I had to quickly repair it as between the last trip and this one, the epoxy came loose and the endcap came off. sadly gluing it went less than smoothly. I was quite happy it worked as long as it did. :D
Small gauges. all good.
1/2" solid endcap. You can see in the video that it had a faulty syringe seal. full of water now.
Empty battery tube. Failed in an undramatic way, and off camera. Small piece of endcap broke and it filled with water. It was squeezed flat first though.
2.6 battery tube endcap. I don't want to talk about it, or how I seem to have for gotten to seal the hole in the endcap. Too late, now you know. Full of water.
Semi-permanently sealed E-Tube, using hot glue. Holy poop nuggets, it worked like a charm. dry as a bone inside. I threw this together at the last minute and I was pretty sure it would leak from one end. (Woot!)
2.6 battery tubes with fake supports. These are the real champions of this test. This is the standard 2.6 battery tube. The 'fake supports' are just plastic tubes. I didn't have any extra batteries. :) They went to 260m+ depth. One of them was completely dry. it collapsed onto the fake battery, but didn't fail or leak. The other tube had a tiny bit of water in it, not enough to pool but more than condensation. This second tube doesn't seem to have collapsed at all!
2.6 endcap. You can see what happened to it in the video. It Failed at 147m (though calling this a failure is insane. 147 m is close to 50m deeper than my 2.4 endcaps lasted. Very impressed.
2 x 1/4" Again you can actually see what happened in the video. Failed at 180m. That is much worse than I hoped, but not a real surprise. It seems that gluing two pieces together is much weaker than the equivalent solid piece.
2.6 o-ring. The tube was full of water. It wasn't in the frame of the video, so there is no info on when this happened.
Buoyancy tubes. These two tubes held up just fine. ;)
Quick disconnects. The small QDCs both kept all the water out! Very surprised at this. They are $3 each at lees electronics, so not high end. The large QDC with oil in it seemed to be absent of water. The 'dry' one had some water in it, though I didn't see any corrosion yet. The question with that one, is weather is would have been bad enough to cause an electrical short.
Here is the summary:
The water was very calm. Took very little time to travel the 2km from my launch point to the 285m+ target area.
The new rig went together very easily, lessons have been learned and mostly remembered, but lets not dwell on the forgotten mast...
I chose not to use lifesavers CANDY because I needed a fair bit of weight and I feared they would not last long enough to reach the bottom. It was, I think the correct decision as the cage sunk quite slowly at first, and the cameras failed before hitting bottom. With lifesavers I would have no idea how deep the rig went, but with the full weight I have some idea.
Having the clump weight on long (1.5m ish) rope really helped because I could put the rig in the water and get everything set up to my liking before tossing the weight in.
I had connection issues during pre-trip testing so I decided to set all the cameras to save files instead of stream. During the test however, I had a fairly solid connection. Because I was only using the 5v battery pack to keep the switch running, it kept switching off. I think it wasn't using enough current and the auto time out kicked in.
Each time I turned the battery back on, the cameras started recording and saving again, but I think they kept recording over the previous file. The raspberry pie does not have a battery to keep time with, and so the time stamp file name must have been repeated, erasing the already recorded video. Bugger! It didn't do that in testing, perhaps because in testing there was a computer/internet to get the time from? I don't know. The reason the footage I have survived is that the camera was shorted out during this recording. the other camera lasted long enough for the reset to happen and so its last video has no usable footage, just a water logged bubble for a few seconds then dead. I am very happy to have the footage we do have! phew!
The camera module kept running the whole time, but water did get in fairly early, but it did not fill completely. I was quite careful when placing the seal. Don't know what the issue was. Still, if it hadn't leaked, we would have no footage at all. :D
It was quite a battle getting the rig up, so much less buoyancy with all the tests failing. I quickly learned to let the boat do the heavy-lifting with each wave.
It was very dark by the time I got the rig back up. I saw 2 shooting stars and the milky way was bright and clear. The rig light up the water below my boat (the Bouncy Castle) before I could see it. I couldn't get a good picture of it. I need to bring a better camera for that. :D
As for how deep the rig went:
The network cable I was using is 295m long and at the end of the descent, it was going 'straight' down. It was very heavy so though there may have been some angle, it can't have been too much. When I was puling it up, at one point a tangle in the cable let go and I pulled (worried that it had broken) a very light cable for some time before I felt the weight again. At another point I pulled up a portion of the cable with another tangle in it. It is hard to be certain, but I am estimating the total length of the two tangles to be 20-30m. So the total depth is likely 265 - 275m
If anyone would like the video file, I have shared it here:
The advantage of the file vs youtube is that with VLC you can play it frame by frame.
To enable frame by frame you need to make the 'advanced' controls available. Just click on the 'view' menu and click 'advanced controls'. It is the same control bar you need to record a stream using VLC. There are 4 new buttons. it is the one on the right. each click will advance the video by one frame. Mostly that would be very tedious, but right around the failure point it lets you see things you would otherwise miss. ;D
This link starts you 10 seconds before the 2.6 endcap fails. According to the large pressure gauge, it failed just shy of 150m! Nice! 147m is what it looks like.
I will compare this to what the small gauge, but first I need to go get it as a reference point. I will update this post once I do that.
And I made a checklist. So I don't forget lifesavers candy, or say... lifesavers candy.
I have shared the list. If you think of anything I missed feel free to add it. :)
You will have to cut and paste it as it seems to be too long to link properly.
Last time, though it too me a while to notice, a little water got into one of the battery tubes. Boooo!
the batteries were a bit rusty but they all still held their charge. One is so rusty and just nasty looking, that I won't be using it unless i get really desperate for one more battery.
As a result, here are the face seal rings I laser cut at VHS. (Vancouver Hacker Space)
For orings, I just use the same ones we use normally and they just stretch. I made the battery tube endcaps with an extra inner disk to leave room for the o-ring.
I didn't take a lot of pictures while building this rig.
Originally I didn't use something like a garbage can because I wanted the cameras to be able to see out into the water past the test tubes. But you can't see much past the mosquito net any way.
Working with a garbage can is nice. it is like working with wood, in that you can just drill a hole any where you want or mount any where you want. Hooray!
I needed a net for the clump weight and I was having trouble finding what i wanted. The prawn trap had a net and the frame is perfect for sitting at the end of the garbage can to hold the test items.
I cut out the center of the garbage can lid and laced in to the top of the prawn trap to hold the can in place. Works better than I hopped. I also have bungees and a 'safety' rope in case it gets hit by jaguar shark in the deep and comes apart.
The camera module goes at the bottom between the 2.4 style battery tubes. ;)
Here is the layout of the test items
Unlabeled and underneath some other items are 2 pairs of quick connects of 2 different sizes. I put oil in 1 of each size and left the other dry. They are very cheap so I don't expect them to hold the water at bay, but we shall see.
I finally collected all of the parts needed to get back on the road with the expedition. Sorry it has taken soooo long! I accept full responsibility.
I was ready to go, then during testing it over heated. It does that when sealed up but not in water. I have 8 - 15 min of dry work but I had not considered the helium I filled the tube with, which doesn't transfer heat as well. Too bad.
It isn't all that bad though, I just had to re-write the SD image and it is now working again.
If the weather does not stop me, I will go this evening (since the it is predicted that said weather won't stop me at that time.)
This giant flashlight/Dalek looking thingy is the new rig:
The wifi router I was using ran out of juice and so the rig turned off. I was unable to connect to the top camera and see what was going on, so I didn't notice until too late. So no exciting implosion videos. Don't worry. I will need to make a third trip so there is still hope for cool implosions.
Here are some screen shots of the video.
Top left: The Squamish river flows into Howe Sound. It takes a while for the fresh water to mix with the salt water. It is known that, fresh water makes light go wibbly and salt water makes it wobbly.
Top right and bottom left: Greeny yellow sea water makes vision tough. One reason why dolphins are so keen on sonar.
Bottom right: This shot is just to cleanse you senses and let you see what a clear image looks like.
So I must set a new date for the test.
How about 1 week from today?
I have also decided to do the test in Howe Sound instead of Harrison Lake. The reason I was testing in the lake, is that my original plan, of renting a boat to test in Howe Sound, failed, due to boat rental season ending.
I have concerns that my battery packs will not hold up. That wasn't a problem in the lake because "its only fresh water". Possible solutions are:
laser cut some disks. Glue them to the end of the battery tubes for a face seal.
Seal face of tubes with silicone.
Fill the tubes with fresh water (or oil, but no.)
The next day, the tow truck driver, (who owns Rayne towing and auto body, and never told me his name) picked me up at the hotel and towed my truck to the OK tire shop. Thanks Man!
The top left picture is the rock from my right tire poking through the tread wall. Right top is same tire but from outside.
The Bottom left is the rock that was InSide my left tire. They call these bad touch rocks arrow heads. This ones looks match its name.
This sign states the distance to the beginning of the logging road. But the logging road is about another 10 km from town. I think 'No Problem!' because I am occasionally optimistic...
It was about 8:20pm
I only walked a few Km, checking my phone every few seconds for a signal, before a nice family picked me up and gave me a ride into town. Hooray! Even driving it took 20min. I called a tow truck and they agreed to get my truck even though by then it was dark.
It was a long drive in (work trucks have stiff suspensions for the handling of heavy loads, makes for jarring trips down logging roads) and an even longer trip out. We were moving no faster than a slow jog.
They didn't have any tires that would fit my truck, so I had get a hotel room.
The problem was weight. too much of it:
The plan required a lot of rocks. So I had a bag of rocks from the shore.
The rig is heavier. The empty rig may weigh slightly less but there is a lot jammed onto it.
The Kayak was low in the water on the last trip too, so it doesn't take much extra to put over the top.
Pie. I don't want to say too much pie. Because that is nonsense, obviously. But I have eaten a lot of pie since the last test...
Also, I was sitting further back than before which concentrated the excess weight in the back.
This picture doesn't do the scene justice.
This is a picture of the roots of a tree, flooded from the spring melt. It is the only underwater image from this trip.
I arrived really late. I got up late (I mean it WAS Sunday after all) and had a number of difficulties. I was not worried because I didn't have to get home early, and it was pitch dark, before I got back to shore, on both previous trips here to Harrison Lake.
It took almost 2 hours to set up. I ran into further difficulties with my console. I had adjusted the Steel tube a couple inches closer to my seat just before firmly epoxying it into place a few nights before. I had been using an empty spool of the exact same type, for fitting. But I forgot I had added a piece of wood to the end of the actual spool, to attach a handle to and give better leverage. It didn't fit, I didn't have a saw, and it was attached with glue, also I didn't want to remove it. Instead I used my drill to cut a bunch of holes in a line and hack it into usability. It sort of worked, but not smoothly. I figured it would be fine and continued with setting things up.
Eventually I got into the water but then I found it impossible to position my self to control the motor. I had things set up the same as the last time, but the top camera on the rig was going to take a beating, if I tried to control the kayak in that configuration. So I switched the rig to be in front of me and started cruising out. But when I looked down, I noticed that the water was only an inch from covering the rear pontoon. The water was fairly calm where I was, but I had about 3km to travel and I could see waves. I tried shifting my weight forward, but the rig didn't have any where to go. It was clear that this wasn't going to work. I turned the boat around and headed back to shore.
I got about a quarter of the way up the logging road when I realized i forgot the lifesavers! Ack!
So I headed back down to the Harrison Hot springs Husky gas station. They didn't have regular lifesavers. Just mint flavor. I bought a roll of mint and a pack of Halls too just in case the mint version was weak and un-believable.
Later, I got hungry.
I forgot to take a picture of the rig with the next on it, so you will have to use your imagination. Note the net is bigger than it needs to be, so it properly, with parts folded here and cinched over there.
The nets job is to catch any endcaps or other shrapnel that might try setting up permanent residence in the lake.
Lasers and Servos and Wifi, oh my.
I used the stairs and hallway (very neatly as you can see) to align the lasers which I didn't have until now.
Helpful hint to future laser mounters everywhere; clamp your camera mount to something. You might need to remove it, but it will save you much time if the camera mount can't move, so is not easily nudged, every time you adjust one of the lasers or touch anything at all... :D
I felt, at the time, that they were "close enough!" but I will have to check them again as I may have been feeling strongly that "I don't want to do this any more". They were a little closer than 9cm apart at 4.5m distance.
I also replaced the plastic geared servo with a metal geared servo. It was a slightly smaller size, therefore, hot-glue. Lots.
The D-link dir-506L is a wireless battery powered router that lets you charge your phone. Instead I use it to power the Topside adapter. ;) It works great, un-tethering your computer and allowing you to view your rov from your phone, or even control your rov from an iPad. The smaller tplink box is also a router but it needs to be powered by usb. It is quite in expensive though.
My Truck is going into the shop for regular maintenance on Thursday the 22nd. Because the site on Harrison lake is down a short stretch of logging road (about 40km), I am going to make sure that the truck is working properly before the test. So the test will be delayed until either Sat/Sun or until next week.
Because this is the weak excuse of a quitter(!), I will do a test run on Wednesday the 21st (the original date I posted for DTII) at Deer Lake in Burnaby, BC. it will be a 30-50min drive instead of 2.5 hours and will not take me out of civilization. :D
Hooray for compromise!
The Ethernet cable will travel, from the water, through a plastic ring, through the front pulley, up to Mr. LongArm, down through my hand, under and around a wooden dowel (with a plastic tube for slippage) and onto/off-of the Spool. Having a wide dowel, a fair distance from the spool, allows me to move the cable back and forth to fill the spool properly. Last time ,it was energy intense, pushing the wire left or right to fill the spool, and I didn't have any energy to spare. This was another lesson learned from DTI. Hooray for learning!
This shelf is where my computer and iPad will go. I haven't finished it yet. It still needs a blackout hood or housing of some sort. I will use a pair of goggles as a view port, so that all I need do is put my eyes into the goggles and I will see through to the screen. It is a real pain to view a computer screen when the sun is shining. Having this system built will allow mother nature to be cloudy or rainy or whatever she wants, without missing an opportunity to make my life miserable (by shining bright sun onto my screen and making me upset.) :D
The Long white tube near the bottom is a wooden dowel with a length of 2.4 battery tube around it (for slippage). The dowel can slide out the left hand side so the ethernet cable can be put through.
The short black tube with a dowel in it and an ugly mass of hot glue is my "saddle horn". During DTI I would have killed for such a simple device. Sadly there was no one nearby small enough for me to kill, that also had a saddle horn. So I suffered instead.
The Spool you see, and the steel pipe, are both just place holders for fitting.
With the rig running, the next step is to sort out some of the other aspects of the mission. Mr. LongArm here is being used to hold a pulley above me so I can use my weight when pulling the cage (though my plan is that I not need to do much pulling this time around).
Now that the rig is positively buoyant, I need something to hold rocks (for weight) and connect to the lifesaver(s).
I attached this old kitchen towel under/behind the camera module. It will fold underneath and a wire attached to the other end will pop up through the big hole (see picture from previous post) to connect to the lifesaver(s).
And the Experiments are attached, mostly with elastic bands. There are 4 (e-tube) endcap tests:
Red: Testing the 2.6 o-ring. Will it hold a seal all the way to the bottom? If it does, that will simplify things greatly!
Clear(behind Red): 1/2" endcap formed from 2 x 1/4" disks. If this holds up, it could make thicker endcaps easier to come by. Cutting 1/2" acrylic might be a challenge for smaller laser cutters and not having to buy a fourth variety of acrylic would also be nice.
Green: 2.6 endcap but made with a different variety of acrylic. Tyson Haverkort bought his acrylic from a different supplier than me. I noticed that it had a "softer" feel to it. I traded some other endcap pieces with him so I could test it out and see if it held up as long.
Yellow(behind green): Regular 2.6 endcap
Not Pictured is an full sized e-tube, it is there both for buoyancy and to confirm that it will hold to the bottom. The endcap is one piece, made of solid 1/2" acrylic.
In addition, there are battery tube (BT) experiments:
In the foreground is an empty 2.6 BT. It has been sealed at both ends with epoxy and has no support inside.
At the back, on the left is a 2.6 BT with a thick tube inside as a placeholder for actual batteries. It does not fill the whole tube but stops short, as the batteries generally do. This test is to see if the tube will fail along that unsupported section.
On the right is another 2.6 BT. It also has a support inside but it fills the whole BT from top to bottom (which makes it hard to detect its presence).
at the back left and on the right are 2 shorter tubes. I had planned to use these tubes as buoyancy on my 2.4 ROV but I don't know how deep they can hold. They are make of the same PET-G material that the regular BT are made of, but the material is twice as thick and they are a slightly smaller diameter.
Another minor test is the pressure gauge. The large one at the back has a gauge on both faces. The one you can't see is the one that was showing during Depth Test I. It took a beating from each implosion, though because it was facing the camera, we can make the adjustments to read it more accurately. The gauge in the picture, was facing the camera that failed during DTI and I need to see if it is accurate. The small one up front (and another small one behind it facing the top camera) are new. I am relying on them to report the depth we reach.
The Camera module needs to be firmly attached so I have epoxied it in place. I drilled holes for the screws, filled them with epoxy and pushed the screws into the holes.
Masking tape doesn't adhere to the foam, it might as well be just a regular piece of paper. Hot glue holds ever so slightly better. Which is to say it doesn't really stick either. It may be that it needs to be very well cleaned of dust (from drilling) or it may just need personal space. Because of this, I put enough Epoxy in the holes that a little bit squeezed out the bottom. That way if the bond fails it should still hold (like an epoxy rivet).
I decided to use the syntactic foam. Because adding the foam to the rig would almost double the weight, I am going to swap out the metal shelf I'm currently using with the syntactic foam board instead. It has a feel similar to some counter tops. It is quite rigid and seems quite strong as well.
The electronics are basically done now and I am working on other things.
- I re-wound the top 3rd of the spool (on the last test I was exhausted by the time I pulled the cage all the way up and so the rope and wire spools were not wound very well.)
- The rod I rested the spool on last time was bent from the weight of the reel and also from my leaning on it heavily. I found a steel pipe to replace the bent brass one.
This time I am not using a rope. I will add buoyancy to the cage so that, when all the test items have failed, it is slightly negatively buoyant. It will still be Positively buoyant with all the test items intact so, I will add weight to the bottom, Tahoe Life saver style. (see the Lake Tahoe Expedition openexplorer.com/expedition/shipwrecktahoe )
This way I can stay at the bottom if I want, but when it is time to recover the cage I need only fight the weight of the ethernet cable plus the bit of cage excess weight.
It occurred to me, in the midst of making the reed magnetic switch, that what I should use is a transistor as a switch, using the reed switch to operate the transistor. Then it occurred to me that I should replicate the OpenROV 2.5/6 method of running the transistor switch from the 5v added to the tether and skip the reed switch all together.
So that is what I did. It was so easy and worked so immediately, that it makes for a boring story, and for that I apologize.
Not only does it work, but it is a far far (far far far far) more usable switch, allowing me to reset the modules down-under* if necessary.
I was being optimistic (which has been know to happen from time to time). The switch worked twice, turning on then off each time, but when I ran the full load (cameras all streaming/recording) it transformed from a switch to a wire. Boo!
The switches were, I believe, rated to .5A each (the guy at the shop thought they might take as much as 1A). The whole system uses 2.5A with all lights and cameras going full. But, I was testing with only 2 battery tubes and that is too much power to draw from 6 batteries. Drawing that much power lowers the voltage (in this case it dropped almost 4V. From 12.5V down to 8.75V) which means it draws more current. With 2 Battery tubes I was drawing 3.5A, which was way too much for the switch(es). As a note, it is also very bad for your batteries to draw too much current. It increases the internal resistance much faster than normal which reduces battery life. In this case it also dropped the voltage too low for some of the LEDs I am using to run. That was a helpful clue as I hadn't yet figured out that 6 batteries, is not enough batteries.
I am now using 4 battery tubes (12 batteries) and the system is running at 2.5A.
In Waters 230 - 276m Deep, I will be Testing the Max Depth of:
- v2.6 Endcaps strength
- v2.6 O-ring Seal (using a ~1" thick endcap)
- v2.6 Battery Tube/Endcap Strength
- v2.6 O-ring Seal
- 1/2" (Solid) thick modified v2.6 Endcap ('under pressure' software suggests this will survive.)
- 1/2" (2 x 1/4" glued together) thick modified v2.6 Endcap
- Thicker Battery Tube/Endcaps modified v2.6 (these are twice the thickness of v2.6 Tube/Endcap, so 1/8" tube, 1/4" endcap, and are powering the Test Module and Lights. It is possible they could fail, while the standard v2.6 hold, as the standard tubes are meant to flex some under the pressure and the thicker tubes may break instead)
- 1" Diameter, 1/8" thick tube, 1/4" Endcap ~7" long. Same material as v2.6 Battery tubes, PET-G (I have 2 of these that I intended to use for buoyancy)
- Various Pressure Gauges
- LED Modules, Sealed in Clear Epoxy. Hysol E-30CL.
- Test Module. 1/4" thick tube with ~1" Endcaps. Contains: 3 Raspberry Pi's, 3 Raspberry pi Cameras, Network Switch, Home Plug Adapter, DC to DC power converter. Overheats from cold in 8 to 15 min when sealed in Tube. (It will run longer in the cold lake water but I am unsure how long)
You can get 'Under Pressure' for free: deepsea.com/under-pressure-upwin