Lovely box full of print-prop parts. First experiments with acetone vapour chamber I might have over cooked the first ones. 2 hours in the tub. Some wobbles in the print are still visible and the edges of the props are a bit rounded- I will give them a wee tweak when they have dried. Next ones will get a quick scrape with a knife to remove more wobbles before vapour chamber.
Really nice to see something real and a big part of the ROV. Thanks Andy and Ole.
The logic level converter arrived to connect the BeagleBone black to Arduino 3.3v – 5v 8 channel. OpenROV cockpit still not uploading firmware to arduino. Even Arduino IDE isn’t able to upload openrov.ino to the arduino mega. – avrdude: stk500v2_ timeout and getsync errors. I’m trying to upload from Linux so am struggling a bit with the learning curve.
Batteries are coming along. The discharger is working but we might need to add more cells to the pack as significant heat is generated while discharging at 2C or 4 amps.- 18650 from defunct laptop packs. I am thinking 45 cells- (15-15-15) 3S 15P which would mean discharging closer to 1C or 2.5 amps per cell when driving 2 thrusters on full power(480W or 40 amps at 12v) The £1 1A lithium charger modules from china have been setup with a desktop PSU and are charging 5 cells at a time for testing.
DCDC converters are a bit of the target of delivering 240w to ROV.
I just noticed the boos DCDC i got has a 8amp fuse on the input- so at 12v it will blow at 200w. Might use 2 of them is that’s possible. And the 18AWG wire that I got from China looks like its got about half the amount of copper it should- need to test to calculate what the final voltage drop will be.
I am also thinking about an array of 3 DCDC buck modules to be used as a charge controller individually charging each of the parallel cell groups from the teather input- This would balance charge the pack and reduce the current output required from each buck module. Whether or not this is possible will become apparent. Much to learn.
24v dc–>DCDC -48v———-teather———->DCDC—–>4.2v
DCDC—–>4.2v
DCDC—–>4.2v
I’m starting to think I might have had to learn less if I had just bought the OpenROV control board for $120.
I would like to share the idea that fishingROV isn’t designed to replace trawling boats just replace the heavy gear that interfaces with the seabed. Could 10 mass produced fishingROVs directly replace heavy nets and be towed behind a fishing boat and catch a similar amount of fish? Obviously they would need to be autonomously catching the fish/scallops.
Planned construction of a hyperbaric test chamber. To test the ROV hull. It will pressurise to 40 bar or 400 metre equivalent depth. It will be interesting to see what happens when the pressure outside the ROV exceeds the gas compensated pressure inside and test the failure pressure of hull components.
The first pictures from inside the ROV. Some blue light noise from the LED on the camera reflecting of the inside of the dome.
The Microsoft lifecam hd 3000 720p worked without changing the openrov code.
I was a bit challenged when trying to write the openrov image to SD card from linux. The windows method worked great. I was also challenged trying to take a screen shot from linux- needed to apply changes to display settings to get one screen shot.
I am still playing with the networking. The first latency test on the camera range from 350-250 miliseconds- might be the screen capture, computer, network?
Pictured is the Arduino and project shield- left and the ACS712s for measuring the current to the ESCs- right. Behind is the new camera housing – 70 5 60mm acrylic tube 4 mm orings and a 8mm stainless connector so the housing can be opened without redoing the thread seal.
The OpenROV software looks great and will save loads of time. This project might turn into an OpenROV in a bigger box. Lots of very clever people have been working on it and that shows. Thank you.
I also made a battery discharger for testing old LIPO 18650s from defunct laptop packs. Inspired by:fixitsan on youtube. With arduino instead of PIC microprocessor.
Ben’s video about battery discharger:
Just mounted the main camera to the etube:
Potential for 3d printed parts- tilt pan mech, disk to hold camera mount. The mount holding disk could be adjustable- center and height. The idea was to slide the mounting disk inside the etube to adjust height. I would like to test the full range of the tilt pan mech before adjusting it to close to the dome- The closer the camera is to the dome the less distortion is evident from imperfections in the acrylic. Need to make a servo signal generator – waiting for delivery from China.
The microsoft LifeCam HD 3000 is a tiny camera which should make it posible to mount very close to the dome or even in the smalles acrylic tube I have been testing. Also it seems to work in ubunto so it might be compatible with the OpenROV software.
The etube was quite easy to realigne- center at the top by filling a bit of the base. It wasn’t cut very square.
Given the relative complexity of blowing an acrylic dome and the impact it has on the housing I am thinking about a separate main camera tube. Modeled on the openROV etube- clear acrylic tube with oring sealed HDPE-endcaps.
Floats are good- glass bubbles with hot glue- has the consistency of Plasticine when hot. It dosn’t sand very well but can be cut with a knife- and presumably could be melted with a heat gun/ blowtorch and resurfaced/shaped.
The tube with the orings in the tube seems to seal reliably at 150psi or 100m depth in water.
The endcaps could be made with a lath but i would like to keep it kitchen table tech if possible.
There is not much room on the endcap for the 1/4″ plumbing fitting so I will try 1/8″ fitting and maybe a 70mm tube?
Left trying the oring in a groove in the outside piece of endcap- the oring keped popping out while under pressure. Right oring in a groove in the internal piece of endcap- tested to 230psi then the pump started to leak. A washer on top of the endcap and tightening the rod stopped the endcap leaking. I was expecting the acrylic to fail.
I will also try the dome as camera housing to see if the additional work is justified by the greater viewing angles and less lateral distortion on the image.
Just uploaded a update video on youtube:
Ideas to dehumidify the air inside the ROV to reduce fogging on the windows. The gas compensation air that will be pumped in at the start of every dive is likely to be wet. So I plan on circulating the air through a loop of copper tube that also connects to the AUX_pod:
The spear targeting laser would be easier to setup of the spear so I potted one to make it waterproof:
Idea of using an OpenROV style etube to house the camera. It would be easier to make and would reduce the impact on the housing. But would reduce lateral vision and make it difficult to incorporate a tilt and pan mechanism for the camera.
Idea to create the AUX_pod from standard plumbing bits: 50mm MDPE pipe union, acrylic disk, rubber seal and tank connector. The connectors are also available in 75mm, if more space is required.
Idea of using cathodic protection for the copper, brass and motor metals. A small current could be passed through the metal -cathode through the water to a sacrificial anode. Would a passive anode protect the metal as much as an active system?
The Ole Hermann Godø 3d printed props are looking really promising. They should provide more than twice the thrust compared to the props I was testing.
I came across an OpenROV test on foam for ROV floats:
I was thinking about using the extruded Styrofoam- the pink one on this test. Unfortunately this test shows it shrinking under pressure. So I am know thinking about making the tether floats out of glass micro balloons mixed with resin or hot glue stick. If the hot glue and glass micro balloons combination is sufficiently buoyant it might be possible to make an extruder that could cover the whole tether?
Got the electronics tube mounted. The tube is 110mm uPVC.
A slotted diamond disk on the 115mm grinder made all cuts nicely. The tube is attached to the four studs that pass through the hull the studs will be used to make a ROV kill switch accessibly from outside and possibly removable teather connections. I might need to separate the electronics tube from the housing with a rubber seat to allow it to be adjusted- at present the tube is 10mm of center at the top.
Experimental acrylic cutting technique- Cutting on a wet tile saw- small diamond disk running in water. The results were surprisingly good almost un-distinguishable from the factory edge no chips and little melting. A slight wobble because i didn’t use the fence. I also has a go at using a thin metal cutting disk on the 115mm grinder results less good but it did get through it without chipping but a lot of melty burs and rough.
I haven’t tried the table saw yet and the chops saw make a bit of a mess with its blunt blade.
I will use the wet saw again. I would like to make a better dome with less flaws than the current one. I came across someone who was blowing acrylic upside down to get a better shape-might be an idea?
–
Thoughts on laying the ROV out- on its side to access the e-tube without damaging the thruster or scratching the dome. How much more drag would be produced from a permanent cage attached to the ROV? Could I make a stand for maintenance/transportation?
