Los Angeles Submarine scratch build

[John.RedearthParticipant @john-redearth]

I am building a model of a Los Angeles class submarine from scratch. The reason for this a little convoluted. I already have two subs that I have built from scratch and they are working very nicely, both using arduino to control the systems and routines, which are, I must admit complex. I decided however at Christmas time to build simpler diving system using the classic config of two pistons and using arduino once again to control them.

I will do a regular blog on the build and this is the first one. It will be on my site and I will drop it on here also if that is OK.

So to that end I built two pistons and more importantly the gear boxes, and started playing with them. The idea that is central to this new system is to control the pistons using a laser distance sensor so I can control the ends of their travel without micros switches, and to get the arduino to remember the sweet spot of neutral buoyancy, all managed though the RC. That sums it up. Also to use a pressure gauge to manage depth keeping. (I do this with my current subs and have solved this little issue already!)

For pics this is a link to my blog

[John.RedearthParticipant @john-redearth]

First stages of the build

[John.RedearthParticipant @john-redearth]

I built the pistons in 100mm acrylic tube. (BTW I also use acrylic glue to ‘weld’ in bulkheads, which has been another revelation to me – stupid me.. everyone else knew about this)

So I needed a hull as an afterthought and on researching I decided upon an LA because it looks fantastic, was in Red October, (I know laugh on) and most importantly it can be made with a piece of drainage pipe.

So my challenge has become to build a fully automated, cutting edge sub, quickly and cheaply.

Decision making time.

I am going to build in the following stages.

• One: to rough out the hull with fins, connections, motor and prop
• Two: to build a two piston system with complete arduino control.
• Three: get it all going and adjust
• Four: take the gear out and finish the hull nicely

Here is a video, mostly featuring the pistons and arduino as it demonstrates where I am going with one piston at least.

I am going to build the bow and stern sections using the ‘discovered method,’ messy but effective!

So what scale? The ship is 110 meters long and a 10 meter beam. Using the standard scales of 1:50, 1:72, 1:96 it became apparent the best was 1:72. It gives a boat 1.5 meters long, and a beam of 138mm. Perfect, if a bit big for my Ford Laser.

So I needed some 138 mm tube. Quickly I discovered this is impossible. The pipe that looks best is PVC drain pipe but that is either 160mm or 110mm (OD). So after some procrastination and research I bought some 160 OD PVC pipe for $8, and ripped 80mm out of it length wise on the rip saw. (Had to do some maths (2 pi R)). I also went to the tip and salvaged some polystyrene sheet and using a homemade hot wire cutter and then made a series of sections for the bow and stern.

First the hull.

I used some tape to bind the pipe together while I applied heat to the plastic to get it to remember this size and sit happily as a 140 mm tube. This process was a little tricky as my torch has a narrow point and a couple of bulges occurred that I had to fix by making a piece of wood as a shape holder. (pic) Then I cut the styrene, used some dowel to make shashlick and glued them together using wood glue. When that was dry I used wood glue (water based white glue) to glue some aluminium foil to the styrene to seal it (seems like a good idea) and then to apply one strip of fibreglass to the join of the pipe and the styrene bow and stern. I made this with a lot of hardener so it would go off quickly as this is the ‘weak spot’ of the sealing process with the styrene. (If the liquid resin contacts the styrene the styrene will ‘disappear.’ This worked nicely, and then I applied the first layer of fibreglass to the whole bow and stern, sanded it all roughly and then applied another layer. I have now added a layer of car body filler and I am waiting for that to harden to start some manual sanding.

[John.RedearthParticipant @john-redearth]

The planes

In the waiting times I found some 6 ml acrylic sheet and using the power bench saw, cut it to make two halves of the planes. This is for the rudder section, the stern hydro and the hydroplanes on the sail. I have a tried and true method of doing this but this time joined the two halves of the acrylic together using acrylic cement, that not so much glues as ‘welds’ the material together and then used my ‘bench’ belt sander (I got this from a garage sale) to sand them down. This is a very enjoyable process as it is akin to being artist. In the end I now have three hydro dynamically shaped blades with a fixed slot through the middle that is the fulcrum for the pivot of all blades. This was a spectacular success and very easy to do. I now also have in mind a sure fire way of fitting these to the hull. The main issue with the stern is that the single prop shaft goes right through where the axis of the rudder and planes will go, so I have to make an allowance for that. I have that all in my mind and this will not be a problem. I mean it will be but it will be solved.

The sail

Also during the waiting time I roughed out a block of balsa into the dimensions of the sail, and then sanded it down into a hydrodynamic shape. Then I used a tried and true method of fitting two pieces of aluminium flashing to it, (Use on roofs) gluing them together and then removing the piece of balsa. I now have a lovely sail made of aluminium, which is light and strong.

What I need to do now is switch my project to building the motor / shaft seal, and the amazing 7 bladed prop.

Arduino

OK!. Meanwhile I am writing code for the arduino. I have an old sketch that controls my other subs, but they have different ballast systems using an adjusting piston, a pump in/out tank, and two trim tanks. The arduino manages this system, with a pressure sensor very nicely and for a few bucks I have a very nicely automated sub, but they are complex. This boat however will have just two pistons which is an easy, and I may say classic approach.

So I started the code again. It is reading so far three channels of the RC. The channel to move the two pistons in/out, the channel that controls the ‘snaphsot’ and ‘auto’ and another that controls the ‘diving and surfacing routines.’ I am using two distance sensors to remember the relative positions of the pistons, to enable me to limit their travel without micro-switches, and also to remember what I am calling the ‘sweet spot’ which is their positions at neutral buoyancy. The arduino will also remember the water pressure as that controls the depth keeping routine.

[John.RedearthParticipant @john-redearth]

The issue I was having trouble with was getting the arduino to read all three sensors at once. I had to use the adafruit forums, always a risky issue for luddites such as me, to get some leads as I couldn’t for the life of me get them working. (I say risky because they are all so good at coding, and I am sure laugh at what I do behind their hands. ‘What an idiot’ I hear them say).

Anyway I don’t care as one of them got me out of gaol and I now have my arduino reading two laser distance sensors, and one pressure gauge. Of interest is that the pressure gauge is a solid as a rock and as I put a little hose into a bottle of water it shows the pressure increasing beautifully, but the distance sensors jiggle around with a 4 mm variation. So I added to the code for each sensor a routine that takes a set number (whatever I choose) of readings, adds them together and averages them. If I use a factor of 40 – 100 it settles the measurement down beautifully, so I now have three solid, dependable readings. Now the coding fun starts! I need to add a control for trimming for and aft, and then writing the diving and surfacing routines and getting the whole thing operating on the bench. Not far away now.

For pics this is a link to my blog

John

[John.RedearthParticipant @john-redearth]

Los Angeles Update:

This is happening part time, unfortunately I have to work, and unfortunately my family does deserve time. I haven’t even broken the news to my wife that another sub is happening. She feigns disinterest, no that is wrong, she IS disinterested, but it is the time that is the issue. At the same time we are starting up a new business in another area.. OK, enough of my personal stuff.

I use ‘dive in to submarines’ on facebook, as place to bounce ideas. One of the issues was that of how big the motor would have to be for this sub. The prop is 80 ml across, with seven blades. So I threw some ideas at the crew and one gave me the idea to use a brushless motor direct drive. Needless to say, I am on board with this, and bought a .. It is not tested yet, but the issue is going to be mounting it, as it is an outrunner and will have to be secured from the base (back). Not a problem really but I will get to that when I get to it.

So I started with the prop this week. I have a jig for making props but it extends to 6 blades only, and I considered making the prop six bladed, but honestly after going to all that trouble I think it should be 7, and it such a prominent feature of the sub. So I used a metal cut off wheel in my bench saw, a sheet of 1.2ml brass, and cut seven blocks of brass for each prop. Then I used the bench saw to set the cut for the base of each blade and make another cut to remove the vast amount of waste, and then set to the grinder, and ground one down to the right shape. This became the template for the other seven. I am at the point now of getting the file and possibly using the belt sander, to start to put the aerofoil shape into the blades.

This is an interesting point. I have always done this with props and generally ground the ‘lip’ of the blade down to a knife edge, or nearly so. It clearly gives the blades better efficiency and by putting an aerofoil shape in the dynamics are so much better. However this is a seven bladed, 80ml prop, and I am going to sharpen these blades. Honestly it will be like having a scythe behind the boat. Ducks legs, childrens’ fingers, shins, fish, all watch out. Who needs torpedos? I just have to back into another boat and machine a hole through it. Anyway, that is a task for today.

[John.RedearthParticipant @john-redearth]

The real work this week was with arduino, and it is finished.

Each stage of the build of this ‘sketch,’ which is the techo name for ‘code’ in Arduino speak was an achievement. I started with a basic piece of code that received the signal from three channels of the RX, and tested it all using the serial monitor, which I use all the time to make sure I know what is happening in the code. Once that was working I saved it, then saved it as the next step and gradually built the code up step by step. It is like climbing a mountain. Each step is secured before moving on, and yes, twice I have to go back to the previous set up and start again.

I haggled with Adafruit until someone gave me some lovely code that allowed me to get readings for the three sensors. (That is where we were up to).

I then added instructions to run two pistons and added the second one. Not hard, I just had to rename all the piston instructions spiston (stern piston) and then create functions called bpiston, and duplicate the arguments. Done, and saved.

Then I added another channel from the RX called trim, which is attached to a knob on the Tx. This became the channel to control the trim. The next step was to write instructions to make this command move both pistons in the same direction together, thus moving the air inside the sub toward the bow or stern, giving for and aft trim. Done

Then I used the ‘routine’ channel to create the actual dive and surface routines. This means that on ‘surface,’ the bow piston starts first, followed by the stern, in order to get ‘bows up’ on surfacing. As this boat is ‘bows up’ when on the surface, I use the same to dive but this is just to get the hull level. I know all this will need to be adjusted when the boat is built but that is just changing a number in the code. Currently they have a 2 second lead. To do this I also had to borrow from some ‘old’ code I have that added a delay without using the ‘delay’ function so as not to stuff up everything else. Done and saved!

When diving, using the routine, the pistons withdraw to the sweetspot that I recorded through the use of the snapshot on the Tx. Have I covered that?

OK, when I use the controls to get the pistons to the point where the boat is at neutral buoyancy, perfectly trimmed, I flick the snapshot switch. This takes a snapshot of the positions of the pistons, and the pressure from the pressure gauge. I did something new here in that I want this ‘snapshot’ to be remembered, so I researched and added some code to make sure it is recorded in the little hard drive on the arduino so that when I go to the lake next time, the settings will be remembered. This is in fact the real breakthrough of this coding! The boat will remember the sweetspot each time I go to the lake. I can then do some final adjustments for the ‘day,’ then snapshot it again.

The final stage was to add the ‘bow’ servo. This is the servo that is on the sail on this boat, or would be a bow servo in another boat. This is only controlled by the arduino. When the snapshot switch is thrown the other way, from neutral into ‘auto,’ and the servo will adjust to keep the boat at the pressure that is being read from the gauge and the snapshot. The coding for this was pretty easy. For those of you with a background in this, I had previously used a PID algorithm to control the depth keeping, (and the auto course control on my surface boats). I now use a much simpler method which takes up much less arduino energy. Also, when the auto is turned off, the servo centres and then is disconnected.

With it all l working, I videoed it in action and this can be found on my website blog. I also have the code and the circuit diagram on that page. Everything to do with arduino is open source!

So I now have the complete system set up for the new boat. I just now need to get this hull finished to the point of being useable, and build the scythe!

BTW, my blog, pics and video are here

Cheers

[John.RedearthParticipant @john-redearth]

First the prop. I used the belt sander to start the process of developing an aerofoil section on the back of each blade, and then continued with a file. Being an ‘old’ fitter and turner, I feel very comfortable with a file. I worked on the aerofoil and then polished them up. The next part that I usually do with props is to bend a concave into the opposite face, but this time I have 7 straight slots in the hub, and decided to leave this part out. Instead I used the dremmel cutting wheel to cut the bases off the blades, and then the belt sander again to get each blade fitting neatly into the hub. I put it together and here is the result. It is 80mm across and quite a monster. I soldered it up with silver solder.

Then it was on to do the ‘drive’ train. I used the same methodology I used in the Nautilus to develop the connecting system to make a direct drive from the brushless motor onto the drive shaft.

What I did more successfully this time was to work through on the white board everything I needed to make. When I started building I then used the lathe to do the similar parts, and did all the drilling at once etc etc which reduced production time dramatically. Just a little planning…

The process involves making two pieces of brass (square to fit in the lathe), and screwing them together and drilling through the centre of both. Then fitting the motor to one side and screwing it on. I needed to reverse the motor to do this as it is an outrunner and fitted a longer 4mm shaft. That was not so easy as it was a tight fit, and I used my vice to press in the new shaft. The other side of the plate is then bored out in the lathe to fit to a brass rod (OD 16 mm) that I bored out to fit the shaft and connector. The far end of this has a 5.5 ml outlet. The stainless drive shaft is 5 ml, so this end is designed to be the seal. I drilled two holes vertically into the end and threaded them with 3 ml thread. I needed some ‘meat’ to do this. I stuffed it up first time so had to do it twice. Pic Then parted off a bit of the end to be the cap, and bored out a space for a little o ring that fits nicely to the shaft. Screw it together and low and behold you have an easy and cheap seal. I use these in both my boats and I have never even had to change the o’rings years later. (maybe I should)

Then I made a coupler that will join the shaft where I cut it (I have to do that to avoid removing the prop to remove the shaft). I also made a bush for the end of the shaft where it goes out of the hull, and a plastic bush to sit inside it.

Then I made the direct couple. I drilled a 5ml hole into a small piece of brass rod and silver soldered it onto the shaft. Then i put the shaft into the lathe, aligned it perfectly and shaved the outside of the brass rod to make it line up nicely with the shaft, and then drilled a 4mm hole into it for the motor shaft. (trying to make it before soldering will end in tears. The soldering process will be hit and miss)

[John.RedearthParticipant @john-redearth]

Next the end cap. I have a tried and true method of making these and it took about half an hour. I tested it with the o’ring and a piece of acrylic tube and it was perfect first time. (Using about 15% compression) Then bored a hole in the centre to take the shaft housing. This brushless motor will take no space at all, and the next step is to start building the structures to hold two servo’s, get the holes sorted and to put two bolts into the end cap so that I can bolt it onto some heavy brass that will lock it into the hull. My thought is that the speed controller and the self leveller will also go in here. All the wires that will be needed to go to the unit will be a nine wire ‘band’ using coloured wire from the electronics shop, for the two servos and the speed controller, and two power wires to go to the speed controller.

Arduino

I am not sure I covered this, but last week I finished the coding. This week I made a couple of pin changes to it to fit with the use of a little ‘Nano’ instead of the bulky ‘Uno’, completed the circuit diagram, and loaded them onto my web page for general use. Click here. Then I created a new video which goes for about half an hour which includes how I built the code from the ground up. Please look if you are a wonk. If not, I refer you to my video ‘arduino for boats 1’ before you start. I mean it! Seeing this without prep is a sure fire way to be put off!

Hull

Then I got to work with the filler. I absent mindedly bought ‘builders filler’ from the local hardware shop. I tried it and it is really not the right type for this work and had to get the real stuff, ‘car filler’ from an auto shop. Then came some elbow grease sanding down the filler and re-sanding and I am about to fill the little holes and low spots and then use some very light sand paper to get it close to finish.

my blog with pics is here.

[John.RedearthParticipant @john-redearth]

Blog link. Was an error above

[John.RedearthParticipant @john-redearth]

I have been working on the drive train, which includes the seven bladed prop, the prop shaft, the end bush, the seal running into the stern end cap for the motor room, the connection and connectors for the motor, a structure to hold the servos, a place for the self levelling device and the speed controller and exit slots for the nine radio wires. (6 for two servos, and 3 for the speed controller) and two larger exits for the power wires. Plus some bolts so that I have a method of fixing it and removing it as needed.

While I am doing this I am making videos of the build for the web site. It is fun as I am multitasking, going from job to job, and then when ready, adding material to movies but splitting it between subjects. So we have the hull build, the drive train, the prop, the planes, the Sail and the arduino stuff.

But then came a disaster. A WTF stupid disaster. I have been working on the hull, sanding the bow and stern sections and getting them better and better. I used my hand sander (for car panel work) to get the overall smoothing of the hull and then applied a new layer of filler and reduced the grain on the sander to 150. After a good smooth out I sprayed with grey primer and sanded with 400 grit to get the 150 grit scratches out and thought that, ‘it is where I want it.’ Not finished but then I only want the hull to be rough so that I can start on the internal workings. The finishing process will be last.

So with enthusiasm I prepared to gut it lengthwise and produce a top hatch. I have done this before with success, so chose to ‘rip’ a piece of timber the exact thickness of the hull, and screw it to the top of the sub hull. This becomes a solid guide to use against the rip saw fence. Then I measured the depth of the cut, and where to start and finish it on the rip saw table. (I never video these moments. I like to show how I do things but the idea of someone using a rip saw table unprepared and cutting off fingers is something I do not want to be responsible for!)

So I made the cuts length wise, pulled it off and put it onto the bench and realised my huge mistake. I had ripped from the base and not the top, thus making the hatch very deep. Then came my second mistake as I should have just re ripped and filled the ‘wrong slot’ but instead I used the hand saw to cut the section out, and cut through the foam within. I realised at once this was no good. The base would not be strong enough and would twist and bend over time.

[John.RedearthParticipant @john-redearth]

Then I finished it all by applying some petrol and watched the foam disappear, leaving a very impressive bow and stern shell. (I don’t video this either for self immolation reasons)

In moments like these there is always time to reflect, consider why the hell you started this project, and maybe, just maybe , how to do this better.

So the idea was this: I went to the car store and bought a tub of car body filler which is impregnated with fibreglass. This stuff is really really tough. Then I went home and cut 40 ml strips off the base of the upper hatch. Then I got on the lathe and turned down two plastic discs that would be a neat internal fit into the hull. This latter idea is also going to be the method I will use to assist with aligning the top and improving the fit.

I may mention here my ‘ripping’ process always give me one problem. A three mil slot that needs to be filled in some way. On my other boats I have made a 3 mm acrylic shell that I have glued to the base, but now I have two slots in this hull. So I need a method that fixes this issue at the same time.

So I put the ‘reduced’ top hatch in place, packed it to fit, and carefully marked on the hull where it came down to at each corner. Then I got the hot glue gun out and hot glued the 40 ml strips in edge to edge, but made them fit perfectly to the marks so they will mate with the top hatch without the need of a gap filling process. In fact, it looks really good! This of course left a 6 mm gap at the bottom, three quarters down the hull. I then I mixed up some fibre glass filler and layered it onto the internal joints which were previously only hot glued. When it had hardened I put the disks in place and they lined up the lower strips nicely and tacked them along the length of the sub using the filler. Now I have a six mil space to fill, three quarters down the hull, which I then did with ordinary car body filler.

The idea will be to get the disks, and gut them so that they become round shells. Drill a hole vertically into the sides where the hatch meets, glue them into the lower hull and then cut through the top parts, and use brass plugs as locaters to fit in the holes, fit them and then glue the top down onto them. Perfect hatch locators, hull strengtheners and possible the mechanism for holding the WTC in place.

All this because of a mistake. I fact this may be a good approach for the future! I am an upbeat person. I think there is always a way out! Maybe I shouldn’t have mentioned the WTF moment, but then these things happen.

BTW, I have now been purchasing off ebay the outstanding parts I need for this build and my account, not including batteries and radio, is still under $500 aus.

Blog with pics **LINK**

[John.RedearthParticipant @john-redearth]

I am now concentrating on the motor room. I have already built into the end cap, the seal and housing for the seal of the prop shaft and a very efficient (space wise) mounting plate for the motor. I made some brackets, bolted them to the end cap, bent them to the shape of the hull and using the fire glass filler, glued the whole assembly into the hull, ensuring it was nicely lined up. (The bolts obviously meaning it is now removable). The prop shaft line up was helped by the disks I have previously made as I could ensure that the motor shaft is right in the middle of the boat. (Pic in blog) Unfortunately in doing so I mucked up the horizontal alignment in that the holes I had made for the servo arms were now out of whack. I moved one of them and filled the other hole. (Another WTF moment.. I must be trying to do this just too fast).

But, it’s all good, lined up and looking nice. I used piece of 6mm acrylic and bent a piece (after heating) to the shape of the inner end cap and screwed and glued it in place. Then added a removable shelf behind the motor for the servos. The servo arms and the holes in the end cap are now a real treat. Line up perfectly.

So now the attention turns to the planes and rudder. I decided to get to the lathe and make some small parts to assist with the inner configuration. I am referring here to the fact that the planes and rudder fulcrums run right through the prop shaft. I have sorted how I am going to fix this, but not the detail so I made some little sleeves out of round brass, drilled it through to fit the 1/8 inch brass rod that will be the shaft, and then turned down the inner part of it so that I have got room for a decent grub screw in the top, and yet it is not too thick in the middle for filing. I should also say here that when I made the planes and rudder I set the fulcrums up to be 10mm unaligned, to reduce the trauma of the conflict a little!

I have a good way of making these planes. Firstly I cut two rectangular pieces out of 6 mm acrylic on the rips saw which will fit the profile of the whole plane, and also rip a shallow cut lengthwise which is the fulcrum of the planes. Then I glued them together with acrylic glue, which actually welds it now into a 12mm block. Then its off the belt sander to get the outline shape right, and then to grind the aerofoil into them.

[John.RedearthParticipant @john-redearth]

The next step is to use a nice piece of brass rod (1/8^th in this case) and a snug piece of brass sleeve. I drill through the hull with a drill of ‘rod’ size and put the rod through and work it until it is square. Hot glue is a good idea to hold it in place. Then I took the hydroplane and used the profile of the hull to mark onto it the centre which is to be removed, cut it out, and now we have two matching hydroplanes, port and starboard. Then drill out the fulcrum slot with rod thickness drill and slip them onto the shaft and grind the planes till they fit snugly to the hull. Once they are fitting I use a set square, and marking opposite to the fulcrum, mark where the planes will be cut so that I have a ‘stub’ which will be glued to the hull.

Then I use a drill that fits the outer sleeve, and drill partway into the planes, past the spot they will be cut, and also open out the hole in the hull. Then I cut the planes along the cut off line, in this case by hand using a fret saw) and then start assembly. I have already turned down my little linkage for the centre of the hull which looks a little like a dog bone. It is drilled out to fit the rod. So then, taking two pieces of sleeve, and the dog bone, the stubs of the planes and the rod, I glue the whole lot in, in ensuring the rod and the dog bone turns freely, but the stub and the sleeve is glued in. The sleeve protrudes out of the stubs, and will fit into the planes. In this process I use a long piece of rod and make the final adjustments to ensure the planes will be square (bow stern, topside and keel).

Then I have generally drilled and tapped a grub screw into the planes alongside where the shaft will travel to enable a fixing once I am happy with it, and then slide the planes onto the assembly and check out the movements, fit and slide.

The next step is to work on the dog bone to make it miss the prop shaft. I have done this by silver soldering onto the middle of the dog bone a piece of rectangular brass, and then putting the assembly into the vice and using a round file, filing through the dog bone to get a clearance for where the shaft will go. This is a process of trial and error but eventually I get a nice c section with a perfectly lined up top and bottom. I then solder (using ordinary solder) a ball joint for the servo arm at an opportune place, and then drill and tap two grub screw holes in the top of the dog bone. Phew!

Then it is off the fit and file and fix. In doing this I used apiece of brass and silver soldered in the end of an allen key to get better reach onto the grub screws, attached a servo arm rod, the cup for the servo arm and eventually got it working. Tried out adding the prop shaft and ‘oila,’ the first set were in. (In this case, I started with the rudders, as they are the ones that are further back in the hull (Remember the 10 mm off centre fulcrum)

Then it was off to tod exactly the same with the hydroplanes. Then it was off to connect the arms to the servos. In doing so II cut the servo linage arms and fitted a couple of sleeves with grub screws so I can remove parts easily, and also make fine adjustments to the length of the arms.

I have now made a very nice intro as I have finally worked out what I and doing here!

Also if you click here for the blog you will see the pics of the build.

[John.RedearthParticipant @john-redearth]

So progressing with the linkage I now connected them through the rear of the end cap directly to the planes and manually had them working. The ‘dog bone’ connectors work very well but were hard to get into place in the narrow stern. I found I needed to make an allen key extension for the grub screws and I can tell you that once installed, these ‘dog bones’ will not be removed except for necessity.

The hydro dog bone hit the side of the hull (whee the ball connector was). I ended up using my dremmel drum sander to cut into the side, and in doing so went right through. This is not such a disaster as it seems as it is ‘under’ the stern hydroplane, helps me to see where the fit the dog bones more easily, and can easily be covered if necessary. All good and now there is complete freedom of movement.

I then connected the servos to the receiver and for the first time discovered where the ‘centre’ was and had them working nicely. Once the relative positions were established I super glued the 1/8^th shafts into place into the rudders and hydros, and tested again. There is not too much of an issue here as the couplings on the servo arms give some adjustment.

As always, getting to the fitting and detail part of such complex systems as hydros, props, servos, levers, and then making it water tight is not an easy process. There are lots of corrections to go though. What I thought would be a few hours turned into a day of fiddling and adjusting.

Then came the motor. I connected this up with a solid connector for direct drive, wired up the speed controller It was very exciting to get it running! But, I found an unacceptable amount of vibration. This lead to a lot of fiddling. I realised I had forgotten something I learned using this technique for the Nautilus. Namely that when the shaft is silver soldered into the brass rod, where it will connect to the motor, the next step is to place the shaft in the lathe, line it up perfectly and give the outside of the rod a light cut. Once this is done I needed to open the jaws, line up on the machined brass rod before drilling as there is too much flex if I do this from simply clamping onto the shaft. So with that done, the vibration slowed, but not completely. I then found my prop shaft was not as straight as I wanted, and then used the lathe to centre it and then used the dial indicator to find the run out. That was fiddly.

There is still some vibration, but not much. This however became a problem for the next step as I decided to add the self levelling device to the motor room. The device wandered when connected and I am sure it is to do with the vibration so I removed it and it will go in the control room when it is built. On its own it works very nicely and as always you have to adjust it to match the movement of the servo.

I also found the ESC was only ‘one way’ as it was designed for heli’s. The instructions showed that it should do forward and reverse so I emailed support and Hobby King, to get the unwelcome news that it won’t do anything but forward and I need another ESC. Mmm. I do find the instructions for escs and motors a bit opaque! At least I found the mistake and have ordered another that will do what I need.

Once this was done, I finished the motor room shell. I cut a pieced of acrylic tube down to minimum size (there is a reason for keeping this small) and then used my mandrel to machine a piece of acrylic for the end of the room and glued them using acrylic glue. After drying I tested it for water tightness and it was fantastic. I moved on and then fitted to o ring to the motor room, unbolted it from the boat, added the rubber boots for the servo controls and, with a full bucket of water, completed the first pressure test. Amazingly, it showed all the wires and their outlets were completely water tight, but there is a pinhole leak in a rubber boot. Once again, annoying but I will leave for now.

Then for the first time I ran the whole thing from within the sealed motor room. Very nice and very satisfying.

I am now going away to Germany for three weeks. My plans amongst others is to use some of the time to develop my videos and also possibly to visit Engel, the Mercedes of submarining.

click here for the blog

[John.RedearthParticipant @john-redearth]

On holidays at the moment, which does not give me any more time to work on my boat, but I am able at least to put some videos together. So, this is a nice video on the discovered method if you haven’t seen it before.

Edited By John.Redearth on 07/04/2017 23:44:40

[John.RedearthParticipant @john-redearth]

On holidays at the moment, which does not give me any more time to work on my boat, but I am able at least to put some videos together. This is a nice video on the discovered method if you haven’t seen it before.

[John.RedearthParticipant @john-redearth]

So I am still on holidays but that has not prevented me from thinking, which is one of the great benefits of this hobby. So many things can be redesigned in the head, during boring lunches with in laws and while waiting for them to start speaking English. (Don’t tell them that)

What I have had time to do though is to compile several videos for the website and to better plan the video titles, subjects and labels. I uploaded 8 or so new videos on the build to date.

Most interestingly I decided to change my main game plan. The arduino code for the sub is already done, but I realise now that I can add a compass to the boat. I really enjoy my surface boats having a compass and it is not hard to add to the code. This means I can run the sub, in this case, to a place on the lake, direct it onto a specific course and then to turn on auto, and it holds this course. I know how to do this and it will take a bit of recoding but not much. The idea of holding a course submerged is quite enticing. This will require the addition of a compass ($8).

The second change I want to make is that of ‘shorting’ out the connections on the self levelling device keeping it turned off until I flick the ‘auto’ switch. It makes sense to do it and so why not. It will require the addition of one more relay. ($2.00)

Secondly I have now decided to rethink the WTC. I have decided to make two new ‘couplers’ which can be glued to the ends of a central tube which will enable the attachment and removal of the two piston units. They will be attached by o’rings. This will require some careful measurement of the central unit to ensure it will hold the two lipo batteries, and also a tray onto which the electronic parts can be placed. Also, the 6mm spindles from the pistons will need to move through this arrangement when the pistons are in diving mode. The piston tube lengths will be calculated by adding the central cylinder, to the two piston gear box segments, and also the motor room. In general we want 25-35% of the total space provided for ballast, which will give me the length of the piston tubes in total.

The new idea here however is to make two plastic discs of 132 mm which will fit snugly inside the hull of the boat. These will then have a 100mm cavity machined into them, but not through them, that will snugly hold the ends of the pistons. The plastic will of course need water and air breather holes. The discs will have three uses.

• To make up a unit in which the discs can be bolted together using a long threaded rod, thus holding the WTC together to enable the management of pressure and to ensure the WTC does not ‘pop’ apart when the pressure builds up
• To enable a fastening system (yet to be designed) so that the whole unit can be slotted into place in the bottom of the boat, and held there.
• To allow for a screw on system to hold the top of the boat nicely in place.

A bloody fantastic idea, that I am sure someone has already come up with, as is always the case.

For the blog with pics and links view here.

[John.RedearthParticipant @john-redearth]

Still on holidays, and still thinking.

One thing that bothers me about the LA sub is the lack of traction on the rudder, which sits before the prop. (Everything is before the prop)

I note that when I built the Nautilus and put it in the water for the first time I put the rudder hard over and it went straight ahead. Not having output from a prop to flow over a rudder really reduces its useability, and that is particularly the case for slow ahead situations. With the Nautilus I found that if the boat was at speed it would maneuvre with the rudder, but it was a delayed process. It seems as though the very act of turning began a moment of turning which could be built upon. I have to say though that in the world of small ponds, and other boaters around who do not necessarily see a sleek black hull down low on the water it is a nightmare. Subs clearly used tugs and other ships to assist them and are simply not thought of as requiring maneuverability. None of this works for a model.

So with the Nautlilus I added a piece of clear acrylic to the lower portion of the rudder that has a little ‘turn up’ to allow it to function in the wake of the twin props. No-one notices it, and I have the capacity to turn, really well.

The LA is very different however. There is no way I can do what I did with the Nautilus so after considerable thought, and the necessity of wanting to manoeuvre the boat I have decided to break the rules of the model and add a bow thruster.

The first thing to consider is how to control it. So I have decided to use the ‘knob’ that now controls compass headings to turn on the bow thrusters. This means that when I want I can turn it on, and through a relay, power the bow thruster. The thruster itself will be a brushless motor with speed controller, attached directly to the ‘rudder’ channel. Thus when the boat is still I can turn this on and turning the rudder will have the effect of incrementally increasing power to the bow thruster.

So the next question is ‘how’ to build a bow thruster. It is possible to get them on the net for less than $60 or so, but they are brushed and this clearly does not include a speed controller. I want a speed controller as the power will be on/off and will need to be managed, plus much better control. Also a brushless unit with speed controller is cheap and is really where the world is now.

But, I am a scratchbuilder,

so the time has come to make my own. My experience in building pumps for the U-boat and Nautilus will come in well here and I know I can build a very powerful and effective bow thruster. So more on that as we go forward. Come to think of it, as this one will sit in the water it will need to be sealed. A similar model, inside a surface boat, will be much much easier to build and adapt. The advantage of going brushless is that the motors are very easy to modify with a longer shaft. The brushed motors have fixed shafts that you have to work around, or to. Plus, a brushless motor will be so small!

Lots more to come on this. I have spare small brushless and I need to test it out with a 3s lipo. Lots of fun when I get home.

Blog on **LINK**

Cheers

John

[John.RedearthParticipant @john-redearth]

The build of the Los Angeles is coming along, and most recently I built the main WTC with the seals and end caps, and also the bow thrusters for the bow of the boat. But I have to digress and tell you about another problem I was confronted with which I have not addressed properly before concerning submarines.

Radio Control

The problem all sub modellers face is that the RC that is needed for subs is not easy to get. There are different rules /laws in different countries re RC frequencies, but the simple truth is that only low frequencies work under water, and only low frequencies can be used with subs. In Australia that is 36 MHZ. In other countries the frequencies run from 29-72 and there are probably variations within that grouping.

The issue is the discovery of 2.4GHZ. Nearly everyone has one now. They don’t require discreet frequencies, and they work flawlessly, and they are cheap and they dominate the hobby, particularly for planes and everything above the water line. In the days before 2.4GHZ, modellers would gather to share their hobby, but would have to use ‘peg boards’ or similar to make sure that frequencies were not clashing. For example in the 36MHZ range there are 20-30 sub frequencies. If you turned on a transmitter when someone else was using that frequency there would be a control problem, and with planes it would be a very nasty and fast end. I can’t even imagine how many flyers have come to blows over this. Or the occasional eye squewered by a crashing plane!

The 2.4GHZ eliminates this by ‘frequency hopping’ and means that no clashes occur. As a result the old 36mhz is no longer used at all and no-one makes them. The sub community is not big enough to attract the investment of building old units.

So, many years ago before 2.4 I bought a 36 MHZ radio and three receivers that had 7 channels specifically for subs. (I still have this now and if anyone wants it let me know). When I build the Nautilus however I wanted to do more but by this time everything was 2.4GHZ. So I searched ebay and magically found a guy (old airplane guy) who had a 9CAP 9 channel Futaba with two 9 channel receivers at 36MHZ. I have to say this was purely good luck, but there are people with these on the shelf that are not being used anymore, but this also highlights the problems for the submariner. Namely, you have to look around second hand. So I used the two receivers and built them into my two subs and they work beautifully.

Then comes the LA. So I put a permanent search on Ebay for a 9 channel receiver and suddenly got a bite, for the USA of all places. (They don’t use 36 MHZ in the US). I tried it out and it worked nicely, until last week. It died when I was trying out the brushless bow thrusters. I am certain this had nothing to do with the bow thrusters. Anyway, it came back to life, then died again. Needless to say even if it comes back to life, I can’t use it in a sub!

So I took it to the local model shop. The young guy there looked at it and waggled his eyebrows and sideburns. This was clearly a dinosaur from the ancient past. He said, I would buy another one and throw this away. That was helpful but also not quite that helpful as there is no-where to go to buy these. There is a Hong Kong on line shop that seems to sell a range of Futaba receivers including old ones, even ones at 36MHZ, and nothing with 9 channel.

So I went to my favourite place on the facebook page ‘Dive in to submarines’ and asked the brains trust there.

The result is I am going to try something new, that I have read about at other times but as I smugly had my equipment I have not thought of using. This is called 433 MHZ LRS system. This system is close to a radio ham technology, modernised and used for ultra long distance drones. It is a higher frequency but has channel hopping and from all experiences thus far works very well with subs. Is 433 MHZ OK in Australia? Perhaps. I find the best way to deal with this question is not to ask.

The stuff I have seen on this describing how to use it was very complex and I never delved into it, but the gurus (Tim and Andrew) gave me a pointer so I purchased a receiver and a unit for my old 9CAP for $50 Aus, including shipping. Amazing value. So when I get it, I will delve into the dark arts and produce a video on it that is hopefully a tail of simplicity and success. Story of my life.. mm.

So while I wait, I will get on with the hull.

BTW the bow thruster works amazingly. The full build description is found on my LA page.

BTW the blog and pics are here

[John.RedearthParticipant @john-redearth]

Main WTC

I noted that one of the reasons I wanted to build the LA was to use a classic design of using two pistons. This is the tried and true method used by Engle, and is relatively simple and reliable. The difference is that I am going to scratch build the pistons, and I am adding the laser sensors to measure where they are and get an arduino to automate the whole process.

But, despite that, I have my own quirks. I built, as I have before, a discrete motor room, with the motor, speed controller and two servos in it, to control the rudder and the hydroplanes. The main WTC therefore does not require these and can be built as a unit without them. I also like to build the end cap for the motor room into the hull, (removable with a bolt) and connect the working gear onto it. The WTC cover is a sleeve that fits over it and is removed between sails for ‘airing.’ This is my design and is a bit different.

So the time has come now to build the WTC, and I am keeping the ideas I thought about in the blog …This means I am going to have three spaces. The central space is the main area for electronics and also holds two lipos. This by definition (lipos are tricky) means that it has to be accessible and the lipos need to be easy to get out for charging. The two other parts of the WTC need joiners and each will have a bulkhead glued in that the gear box and motor drive can be connected to. The other side of the bulkhead is the piston. So the two connectors are the issue and I decided to make a two way connector for the bow section, so that I can remove the piston section easily, but also remove the other side of the connector from the central section so I can get full access, and also therefore full use of the space inside.

However I decided that the connector for the other end of the central section could be glued onto the central section, and only be ‘breakable’ to remove the piston section. This is so that I can build into the central section the horizontal bulkheads for mounting the electronics, and also to make a fixed space for the lipos. Of course the structure I build will be screwed onto the joiner, allowing me to remove these parts as I wish. The other issue is that I need an exit point for wires, and this structure will give me more room to have a ‘wired out’ section in the joiner. (The bow one has a very thin section as I use all the meat of the joiner for the two o rings). There is only one on the stern section.

So with the structure in mind and sorted I then had to decide on dimensions. This involved using the ‘rule of thumb’ method of having 25-35% of the total WTC hull, including the motor room, for ballast. I have made full video on all of this and it can be found here!

So I made the connectors, cut the WTC material with the dremmel, glued in the piston bulkheads. (Although before gluing these in I had to make a removable framework to hold the piston sensors. To do this I used my favourite material ‘acrylic’.) That done I completed the joiners and tested them with o rings.

Then I added two disks for and aft that fit across the entire inner part of the boat, with cut outs that fit the WTC. This has purposes. One to enable me to bolt the system together to stop it coming apart due to the pressure of two pistons, two to enable me to fix the whole system into the hull, and third to provide wholes in the top I can screw the top of the boat to.

See blog

**LINK**

[John.RedearthParticipant @john-redearth]

So following on from last week, I added the o’rings to the WTC connectors and mounted the system in the hull. This was in interesting process in itself.

After I ‘sat’ it in, it was a matter of standing back and seeing how it looked by eye. It needs to be central, but the LA does have a ‘bow up’ stance when surfaced so a little shift toward the bow is quite reasonable. I am not too worried about this as when I get the boat going finally I will set the pistons to fully dive, or at least to where I want neutral buoyancy, and then add ballast to the hull till I get it sitting perfectly using mechanical adjustments to the right spot; namely 10 Centimetres under the water. Only when that is right will I empty the pistons and see how it is sitting above the water, and I can then adjust the pistons (remember the laser control system) to get it sitting exactly how I want it, and then locking the pistons in to that setting.

But now it is a pretty much course adjustment. We have way to go. So I set it to the best spot and then heated up two pieces of acrylic to act as fasteners. I wanted them to be 90 degree brackets. Using a heat with acrylic is easy and results in a light, and strong bracket. I added to the bow bracket a screw that points to the front of the WTC holder and then drilled my WTC holder to accept it as a slot. I used some fibreglass filler on the bow bracket, and mated up the WTC and holder to it and let it dry. The stern bracket was screwed onto the stern WTC. Then I drilled a hole into the base of the hull at the stern bracket and threaded it so that I had an easy fixing model for the WTC into the hull. Now all I have to do is unscrew the stern holder, and the entire structure can be slid out. Done.

Once this was in, I could now start to work on fitting the top hatch to the base of the hull. I placed the top hatch in place and then drilled and tapped screws into the bow and stern WTC holder. (Remember the multipurpose holder concept?) This now reveals how well the top fits to the hull. This is a revealing process, doubly so for me as I have done some work on adjusting the lower hull due to my previous cutting error. I won’t dwell on this! I realised as I played with this that the lower hull had too much play. The truth is that PVC pipe is not that rigid. It is very strong, but does bend. What to do?

The only way to fix this was to do it properly. So go back to the lathe and build three new pieces of acrylic (I have lots of this stuff) that fitted the entire inner hell. This is a little like what I did to make the holders. I ended up with three pieces of acrylic (6 mm) which fitted perfectly to the dimensions of the inner hull. I took the wtc out and adjusted them to get them right. The plan is to fix one under the motor room WTC, and two under the larger WTC placed to achieve maximum strength to the hull. Once I had the positions right I then cut them so that all I was left with was a thin bulkhead that fitted under the WTC’s and travelled up the sides of the lower hull. I used fibre glass filler to hold them in, added the WTC’s and let the filler set, then removed the WTC’s and used screws on the top of the lower hull into the top of the bulkheads. Then I added the top hull and adjusted the screws till the lower hull matched nicely, and then removed the upper hull and used filler to fasten the bullheads in place, and ‘threw’ the screws away.

The next problem is that I needed to adjust the positions of the holes that I use for the screws that bold the WTC to the holders lengthwise. (If you remember, these are the bolts and screws that hold it all together and give it strength to managed the internal pressures in the WTC without it popping apart). I needed to adjust them so then were very close to the sides of the WTC to allow space got my new bulkheads. (Nothings is ever easy@)

Now I had a strong lower hull, and I added the upper again and found I needed to do some work to get these to mate nicely. This was now the time to get the filler hot, the sander and to start ‘working’ this hull. Due to my previous build mistake, this resulted in some work, I might say. I used glad wrap under the top hull and added the filler and worked from there. The magic spot, which is when filler is starting to go off and has the consistency of ‘cheese’ as a real help, as I could pull it apart easily and cut the filler with a knife to get it right before it went off.

I like using filler as I have done some panel beating in the past when people still ‘did’ that. It’s messy, but very forgiving and eventually I ended up with a perfectly mating upper and lower hull.

The next step is now to build in the bow thrusters, and then to fit the sail and get the upper servo connected up to it. Unlike other builders I see no problem in mounting a servo in the water, especially if it is doused in plasti dip, and has a small o’ring added to the top servo arm.

Pic are on the blogs

**LINK**

[John.RedearthParticipant @john-redearth]

Hi all. I have a very unexpected update. I have been working away on the LA and have gotten to the part of finishing the arduino code but have hit some very unexpected problems which have cramped my style.

Because I am writing this newsletter from hospital. In three days I will be under the knife for the removal of a brain tumour. I finally know why it has taken so long to wire up the LA Submarine, which is possibly the best boat I have made. As if anything else matters, I hear you saw? Well life goes on, and I have no interest in stopping.

Today the surgeon gave me news that is as good as can be so far, and as I am a congenital optimist, I believe I will get there. Fabulous friends and family are right with me.

To the LA? I wrote the arduino code and got it working on the bench. I finished the videos about getting the project to that place. Then I started the relatively easy job of wiring all the parts into the WTC. While I did this I managed to blow up my 433 receiver, one of the distance sensors, the txa9548a multiplexer, and possibly the compass. For some in explicably reason the nano arduino is indestructible. Go figure. I didn’t want to blog to you about abject and stupid failure? Always good to admit to problems and mistakes but this was ridiculous!

Writing is not my strong point so please read my blog.

[Bob Abell 2Participant @bobabell2]

Hello John

Can I on behalf of all our members, wish you the very very best for your forthcoming operation

You are being very brave and optimistic about it all and that is half the battle

Get well soon after your op and finish that Sub

Bob and the Gang

[John.RedearthParticipant @john-redearth]

So now I am back into the submarine. I mounted the hydroplanes onto the sale using the space just above the WTC and below the top deck for a low-profile servo. When I wired it up to the circuit I discovered that it turned in the opposite direction to the one that I wanted but that was simply fixed by reversing the code in the arduino script.

I then built a bow thruster using a piece of PVC and a brushless motor. The details of that bills are all on my website. The result is an extremely powerful Bow thruster which works in accordance with the rudder but only when a particular switch is thrown on the transmitter.

BTW during my experimentation I used a cheap 2.4 gig transmitter and receiver with to avoid any further catastrophic damage to the 433 MHz with unit!

Once this was done I then wired everything up on the bench. This was a very difficult process during which time with I had numerous problems which were eventually resolved when I stopped using the arduino breadboard with and simply wired everything up directly with. The complexity of the system plus the numerous connections with the breadboard just simply threw up too many errors to make this a successful process. As I wired it up stage by stage, testing each stages I went, I ended up with the whole thing working on the bench which was extremely exciting. Nothing was easy!

The next stage was to build the shelving that sits inside the central WTC to enable me to fit the two Lipo batteries underneath. This dictated how much room I had at the top of the WDC to fit all my electronics, of which there are a lot. Of course adding with extra components such as a Bow thruster and a compass into the mix has taken up some extra space and added a lot of complexity for which I make no excuses and if you're following me with this you can leave these components out!

So then I started to wire up of all the electronics into the WTC shelf. That process began by placing the component side-by-side and working out where they would ultimately best fit. It took some time to find the best spot for the receiver and a couple of occasions I wondered how all parts were going to fit. But once the spots were selected I started a logical process of wiring each part up piece by piece.

One part that went in early is by ‘arduino failsafe’. This is a small servo from the RX that sits in the WTC that throws a micro switch at a particular point of travel of one of the knobs on the transmitter. When the switch is thrown, it resets the arduino. This is just fantastic.

I also revisited the pistons as I noticed that the distance sensors seemed to be acting unreliably. I noted that the distance being measured changed marginally if a blanket was thrown over the whole system keeping it in the dark. I put a piece of black plastic on the back of each piston so that the censor had something stable to look at (the back of the piston is white.. this may be a complete furfy) and I ensured that the hole between the bulkhead, the censor, on the back of the piston was completely unobstructed. This has assisted with the stability of the measurements.

With so I began the wiring by connecting the sensors with the TXA multiplexer board (allows you to put multiple sensors into the arduino) and each stage of the build I tested with each component with before moving onto the next stage with. One of the best ways of testing with arduino with is to use the serial monitor on your computer loaded with the arduino software to measure each component and its performance. So gradually with I added the relays which power with the pistons, and for the first time will use the MOSFET to use with the self leveling device be with.

Let me backtrack: As I have been going on I've been adding with ideas to how the sub will work and realised that when I hit the ‘auto’ button with on the transmitter I could make it serve a number of purposes, one of which would be to turn on the auto leveling device. I was previously going to use with a relay to do this however the MOSFET is very simple to add and works flawlessly with the arduino. Using this enabled me to use one extra relay to turn on and off the Bow thruster. The relay with simply turn the power on and off with to the speed controller with that powers the bow thruster.

So I now have all of the components in the boat as I speak. I have a fair amount of work to do to now but it is all working with and I hope on Sunday to have it ready to begin the manual process of adding led to the Hull to get it to sit properly in the water at neutral buoyancy. With I might add that my process forgetting a submarine ready starts with neutral buoyancy not surface buoyancy. This is just my preference and it is no better or worse than other methods but I like to get it right first and then get it sitting nicely on the surface afterwards is going to be fun.

Busy busy

Full blog and pics here

John

[Author]

Posts