Adding pitch control to a model sub.

[Guest]

In the book model submarine

In the book model submarine technology, there is a circuit for a proportional piston tank controller.

It's essentially a beefed up servo amplifier, with big transistors for the H-bridge to allow control of larger motors. It's ideal for your application.

The circuit also includes a crude failsafe that will empty the tank on loss of the radio signal, via a darlington transistor. This should be omitted for pitch control.

You will need a pot attached to the system to make this work.

If you want to have a go at this, let me know and I'll give you a bit more info.

Andy

[cstranc]

I don't think I've given

I don't think I've given up. It's a day later and I'm thinking about it again. I just think I will put it on pause for a little while. I would like to work on some other areas of the sub for a while.

Davy,
I cannot fault the APC. When I rotate it I know it responds very quickly and will drive a servo, or a motor without load.

By coincidence I have a Mtrornics ESC on order. I'm sure I shall try again when that arrives. Also I have been thinking about pulling the ESC out alltogether and just use relays to drive the weight.

Andy,
I love talking to you electrical folks. I was a mechanical engineer by schooling, but I drifted into a long career as a software developer. I get the distinct idea that I am missing rather a lot of opportunities through my electrical ignorance. Thank you for the offer. If I travel down that path I shall take you up on it.

Chris

PS. If you have any "recommended reading for electrical nerds" please feel free to mention it. I went out and got one of those 300 electrical experiments in a box so I could sit down and boost my knowledge. Just I have not had the time to sit down yet...

[Guest]

Well I work in the

Well I work in the electrical field, although I'm completely self taught when it comes to electronics- so just an enthusiastic amateur.

The circuits required for servo control are not difficult. This is mostly due to custom chips replacing discrete components (e.g. NE544, M51660L).

Many off-the-shelf servos, i.e. Hitec, no longer feature these chips, they're now using small microcontrollers like the PIC range, which are actually cheaper and offer some advantages over their analogue counterparts.

If you're a software bod, then you have an opportunity to design your own controller suited best to the needs of an auto pitch shifter. PIC and ATMEL microcontrollers seem to be the favourites amongst hobby folk.

The former can be programmed in Assmebler, C and Basic. I've been teaching myself the fundamentals of PIC assembler, because I have a need for some special mixers for a couple of subs I've built, and these are unavailable from retailers.

An APC-4 is an ADXL202 accelerometer coupled with a small PIC microcontroller and a few passive components.

If an H-bridge chip was incorporated in with the microcontroller, you would have full control of the motor built in with the leveller.

Andy

[cstranc]

C was my first language

C was my first language (after English, I suppose).

This sounds almost too cool. I think you have sold me. But first I'm going to play with torpedoes, the bow planes and pericsopes. This will also give me the time to order components.

Where would I order the components like:
- NE544, M51660L.
- ADXL2002
- PIC

How do you program a PIC?

Chris

[Guest]

Hmmm, just found this-

http://secure.oatleyelectronics.com//pr ...

Hmmm, just found this-

http://secure.oatleyelectronics.com//pr ... 6b14ec1a4b

Looks great, and cheap too, the components would cost that alone if you made your own PCB.

They also sell servo chips. NE544 is difficult to get, as it's no longer in production. You can pick them up surplus/dead stock or have a fish about in old servos (old Acoms servos often use these chips).

Better to use the M51660L though, as it's a current design, more compact too.

ADXL202 can be purchased from Digikey http://www.digikey.com , although a cheaper alternative is now available, which does the same thing, but is less expensive, and more resilient. It's a sensor made by memsic-

http://memsic.com/memsic/

The levellers coming out of Germany all use these sensors instead of the pricier ADXL202. The difference lies in the way the chip senses level relative to gravity.

The ADXL sensor uses a micro machined beam with lots of little 'fingers' on it. These interact with more 'fingers' which are fixed, and the two together make up a varying capacitance, that's the way I understand it anyway.

The Memsic sensor uses hot air. Miniature heaters inside the chip heat an air bubble. As hot air rises the sensor detects which side is hottest and biases the output accordingly. Because this sensor has no moving parts, the chip is virtually bomb proof, and also cheaper to produce.

One slight disadvantage is that the chip is slightly less sensitive than the ADXL beam design. However in practice it doesn't seem to make a dot of difference (I've played with designs that use both types of sesnsor).

Programming a PIC?

You need a basic spec PC (486 or better) and an interface (DIY or buy readymade £10-15) which plugs into either a serial, parallel or USB port (depends on the programmer, USB tends to be most expensive, but better for later machines which may lack a serial port).

You will also need software to blow the code into the chip, this is available free from Microchip, the company who make PIC's, it's called MPLAB-

http://www.microchip.com

It's a steep learning curve for a software novice like me, but you'll probably get the hang of it in no time.

There are a bewildering array of PIC's available, but for most of our uses the low end devices are best. Most of the devices are now equipped with EEPROM, so no longer the need for UV erasers and other such nonsense.

The smallest PIC is a 6-pin device, but even this would be adequate for an auto leveller-

1 input for the channel input
1 input for the sensor
2 for the h-bridge to the motor
2 for the chip power

Dave Forrest (davy) knows a lot more about this stuff than I do, and he's played around with DIY levellers too.

Andy

[anonymous]

It is very kind of

It is very kind of Andy to say. Though could I recommend a friend of mine from the Mid Thames Model Boat Club:

http://www.kleefeld.freeserve.co.uk/mod ... _speed.htm

He is not only very good with PIC's but also the power switching end. He could sort something out for you at a reasonable price, I'm sure.

Before you do anything like that, though I would suggest that you try a simple switcher with relays. A screw driven system is not very demanding in terms of needing a rapidly responding or proportional system IMHO. An on/off control sytem (like your central heating sytem) may be OK as a first step and relays are easy to understand!

Having said that if you want to start with PIC's, and they are very useful in radio control, you have come in at just the right time.

The current 18F pics can be programmed in C using a free compiler (See the Microchip web site).

On the basis that the best programming language is the one you already know - you could steal a march on we assembler types

Davy

[cstranc]

I was thinking about this

I was thinking about this while walking the dogs today.

The stern dive planes will have a APC controlled servo. I could mount some micro switches around that servo. Those switches control relays that drive the motor (full voltage) in the appropriate direction.

It's simple, but what do you expect from a mechanical engineer

I would still need to have a seperate enable / disable because I would not want this engaged all the time... It also makes use of the level sensing device that is in the APC already....

I shall have to learn about PICs just for the practice, because they sound like an awesome tool to use in RC.

Thank you both for all the info and ideas.

Chris

[Guest]

I would keep the static

I would keep the static trim and hydrovane control discrete.

Andy

[cstranc]

Well the torpedo tubes, and

Well the torpedo tubes, and front dive planes are getting closer, so it was time to look at the trim thing again. This time round I started working on the bench to see if I could get something that worked nicely.

You can see the reciever attached to a ESC via the APC. The APC is mounted on a 2x4 which is pivoting on a copper tube. The setup is driving a motor that turns a threaded rod that moves the lead weight back and forth.

Here is a side view:


So what's different then before? The motor. I found these really neat motors from Tamiya. They are for hobby robotics.


Depending on the set of gears you install you can get a reduction of 400:1. I am using around 100:1. This gave me two advantages. The system should be less "twitchy" since the weight moves more slowly. The 100:1 will generate will require very little torque from the motor, so a smaller current should get it working.

So when I put it together and tested it the 2x4 would rock from side to side. Most un-happy. I thought about this for a while and looked at my setup. Eventually I realized it was quite un-realistic. That copper pipe the 2x4 balanced on had no "give" to it, so the ballance point was too precise.

I replaced the pipe with some blue foam.



The foam compresses a bit. As you can see the motor can now do a great job of bringing it to vertical. I have tried adding/removing bits of weight and it does a great job returning to vertical. I can also use the stick to get it to balance off vertical too. Just what I wanted.

This is an end-on view of it.



Just because I am a curious type I did check how this would respond if I took out the ESC and used a servo that activates micro switches driving the motor back and forth. This was the setup:



The APC causes the servo to move which slides a plexiglass sheet. The sheet has "cams" mounted to activate the motor. It turns the motor on/off and reverses the direction. There are actually 3 micro switches mounted next to the servo. 2 to control direction, 1 for the on/off.

It's a nice setup for activating micro switches, but it was nowhere near the control that the ESC setup had. I think I shall save it to control my ballast solenoids... Sticking with the ESC control.


I was a little nervous about plastic gears moving this big lead weight. My worst case scenario was "what if the sub went vertical?" I tested this by holding the 2x4 vertically and the gears had no problems lifting the lead.

I also tested the motor that came with the tamiya kit showed no ill signs after water emersion over a few days.

Me thinks this is the setup I shall be installing in the Sierra!

[Guest]

I gave some thought to

I gave some thought to this system, and considered a good compromise would be to set the servo controlled weight manually, but have a leveller trigger indicator LED's (which could be the nav lights) i.e. flash the LED's when the boat isn't level, steady when it is.

This would provide a pondside spirit level. Micro controllers and leveller chips are so cheap now, it wouldn't cost more than a few quid to implement such a system.

Andy

[Guest]

Hello Chris

I am that "guy"

Hello Chris

I am that "guy" that uses the GSI and Pellican as well as the Otter Boxs for wtcs,they work very well and the best are Otter and Pelican,the lexan is thicker and ribbed for better crush resistance and are garrenteed not to leak to 100ft.There many sizes and are inexpensive and work well in most submarine aplications.
There is however another alternative and thats the single or double bulkhead pull out tray wtc that I use on the 5ft alfa and other subs.Made from stainless steel and a silicone seal and perchased in many lenghts all with the same diamiter of 4.75 aprox in. is the coffe container.Being round the pressure is evenly distributed and will not deform at great depths for an rc sub I mean about 25 ft.The lexan ones above especialy the GSI unit would compress enough to missalighne the motor and shaft to the point that a drag would be caused and hence rob the sub of running time and heat up the motor eventhough thats not an issue with my watercooled motors.I know you dont like to have all the shafts and tubes ect exiting at the end cap but its realy one of the best "all around acess" methods that I have found,if somthing breaks or there is a leak its easy to find and fix!
The nice thing about the round wtc is that it can be removed and opened easily exposing the entire guts ,a must for easy maintenance.
I like the shifting waight idea and have tried it but I still like to have mini trim tanks of syringes and a gear pump operated iether by manual or automatic means.
But definatly dont give up on your ideas ,find that answer ,that component or unique way to do things that make it happen,just like I do subs are my passion and building your own wtc is esentialy the sub.
Check out my posts on the WTC and the DeepDive VI reborne.
And there is nothing wrong with overengineering as long as it works and does not cost an arm and a leg as some of those WTCs do.
David Santivanez

[cstranc]

Sorry for the delay in

Sorry for the delay in reply. I was away over the holidays.

Andy,
I am holding the direct servo controlled option as a last resort. It's should be easy to take something complex out and replace it with a manual version (I hope).

David,
Thanks for the pointer on the round wtc. The Sierra is 8" abeam, so it fits the GSI boxes. At some point I am going to have to build something a little smaller and having an in-expensive round profile would be very handy.

I wish I was at the point to actually commit to working inside my WTC. I only had three days of vacation that I could spend time in the shop and decided to use it renovating so I have a little more modelling space.

PS. If my sub goes more that 6' deep I will be swimming after it. Call me a coward, but I cannot even imagine running it with over 20 feet below the keel.