State of the Art sub controller, any interest? - Fully integrated submarine regulator

[jsl]

Hello Submariners!

The Submarine Controller university project I have reported on a while back has been completed. Two of my fellow students and I completed this project at the Technical University of Denmark.

The objective of this post is to uncover if there is any interest among you out there, for this sub regulator. If there is, we might be three students interested in making these to support our studies.

Please to observe, that our test submercible did not have any dive planes at all. The regulator can maintain depth and angle solely using piston tanks.

Let me start pointing out, that the datasheet presented below, is the full version of the regulator. As it is likely overly advanced (and costly) for model submarine application (no offence), we're thinking of splitting the regulator up in three versions.

The Professional Version: Only relevant for Hardcore submarine/electronics freaks. Full complexity as in the datasheet. The pricetag will reflect this.

The Standard Version: Serial communication and data logging left out. Regulator will control depth and angle on planes during motion and change to depth and angle control by piston tanks during static. Pressure sensor will be changed to improve accuracy on scraficing maximum depth. Max depth will be around 4 meters.

The Economy version: Serial communication, data logging, one piston tank controller, one servo controller (front or aft plane), and associated software will be left out. Pressure sensor will be changed as above.

The economy version will only operate one central piston tank as both ballast and trim tank. Angle while static will depend on the static trim of the sub, while static depth will be controlled by the piston tank. Depth and angle control will transfer to either aft or forward diveplanes under propulsion.

[url=http://www.student.dtu.dk/~s011263/T1-SSR%20datasheet.pdf]Datasheet]

The important issue of pricetag... Leave it out for now. This is (for now) about interest, not economy.

We tried to make the regulator a nice little package. Advanced, while still easy to hook up and operate. Hope you like it.

You are welcome to comment, and ask questions if the answers you seek cannot be found in the datasheet.

[guillermo pelaez]

Jacob,

Good to hear you guys

Jacob,

Good to hear you guys finished your project successfully. Congratulations for you and your team-mates!

A couple of questions I am curious about:
1) Have you measured the RC output error due to interruptions handling when sensing the RC input and generating the output?
2) For changing an RC output value (once this have been solved by the software), are you using incremental steps to reach the target value or just put the value at the RC output?
3) Seemingly you guys used an 8bit A/D converter right? How did you achieve the 5cm tolerance for depth control? Or did you measure the error from the test results?

I like how you put together the datasheet. It has a very professional look. Very impressive angle keeper for not using hydrodynamics flows (not using planes).

Again, congratulations!

Cheers

[Guest]

Hi,

Good work. Absolutely fascinating stuff,

Hi,

Good work. Absolutely fascinating stuff, and it seems as if you have something unique here.

Put me down as interested for the advanced version.

Thanks,

Andy

[bob the builder]

Sign me up next to

Sign me up next to Andy!

[jsl]

A couple of questions I

A couple of questions I am curious about:
1) Have you measured the RC output error due to interruptions handling when sensing the RC input and generating the output?
2) For changing an RC output value (once this have been solved by the software), are you using incremental steps to reach the target value or just put the value at the RC output?
3) Seemingly you guys used an 8bit A/D converter right? How did you achieve the 5cm tolerance for depth control? Or did you measure the error from the test results?

I like how you put together the datasheet. It has a very professional look. Very impressive angle keeper for not using hydrodynamics flows (not using planes).

Again, congratulations!

Cheers

First of all, Thanks guys.

We did have our problems getting back to ground after we got the grade for the project. We scored a straight A+ (if it even exists). And attempts have been made to lobby us back and do our master projects at that institute.

Ahhh Guillermo... Cutting right to the bone again, ehh?

Well, here's your retalitory broadside!

1) RC output timing error was zero. The algorithm was timed in such a fashion, that it was processed in some free holes in between the main regulator routines.

However.... The RC-input algorithm needs some tweaking before any regulators would be put up for sale. A periodic timing error results in slight servo twitching, a fact that must of course be addresed. As we used only serial communication and pre-programming, that anomaly was not addressed propperly within our limited timespan.

2) All RC servo-out signals are generated by a very nice little routine that uses next to nothing in processing power, as all PWM generators are used for driving the H-bridges. And we change directly from one value to another.

3) No.. We use the very nice and in fact extremely accurate 10 bit ADC's build into Atmel processors. Using mean value algorithms we managed to reduce the precision to effectively less than one bit. In that noisy enviroment from three H-bridges and running a processor with a 10 bit ADC on the same battery supply, that was quite a feat. But I can assure you that we did use every trick in the book, and then some, to get rid of the noise on the ADC reference.

Sorry about this people, but he started it. I hope I have not driven anybody crazy with this heavy duty tech talk.

Oh... One other thing. Our test sub was not much but a tube with very little water friction at all. Any sub with planes, rudders, appendages and so forth, will likely result in even nicer depth and angle precision. Furthermore, changing from a 25 meter pressure sensor, to a 4 meter pressure sensor will also increase depth precision by a factor six. It's gonna sit dead still in the water.

[guillermo pelaez]

Jacob,
Thanks for the prompt

Jacob,
Thanks for the prompt response! Very clarifying... Indeed high class engineering mate!
It seems to me that switching to the pressure sensor for 4 meters (about 12 feet) will suit best the RC submarine applications. Sacrificing depth for accuracy shouldn’t be an issue, and you still have up to 4 meters...!
Perhaps you guys could write an article in the SCR about the project...
Cheers,

[jsl]

Perhaps you guys could write

Perhaps you guys could write an article in the SCR about the project...
Cheers,

Well...

Before we write anything, let us make darn sure it works as advertised on a commercial kit before we get anywhere near marketing it. A little thing I have learned from past experiences..

Problem is that as students, we're on a tight budget. I'd hope we could either find a sponsor, or somebody in Denmark or Germany, willing to get into a cooperation. Something like a free regulator in their sub, in exchange for helping us perfect the RC sub versions of the regulator.

Someone with a boat already installed with a single centered piston tank. That would make a nice and easy test platform for the Economy version, for starters. And it doesn't really hurt having somebody else's word, that it works as promised.

It's my impression that the twin piston tank system is not very common, and that a WTC might have to be build/rebuild to house the system.

[guillermo pelaez]

Well, you have the boats

Well, you have the boats from Gotland and others at SONAR, using trim piston tanks with a central ballast tank. I would guess that in these boats the operator will fill in the main (centre) ballast almost at periscope depth, and then the controller will handle trim and depth from there on, right?

[mylo]

JSL.

Am I interested

JSL.

Am I interested ? ...of course, anything "sub" interests me.

Are you asking for cash to fund your project ? Is that what "interested" means ? If so, mail me your mailing address and I'll send you a few bucks, or Euros I guess. Nothing big time but, I'm always game to help a good cause......kind of like cancer research or something. Heck, even if I just buy you and the boys lunch in show of my appreciation that somebody actually puts effort into designing this kind of thing.

Myles.

[jsl]

Thanks a bunch, Mylo.

But this

Thanks a bunch, Mylo.

But this is not an undercover fundraiser. It's just a probe to see if there's a market for this regulator.

It just mean, that we don't have the cash to buy an entire boat with all extras, nor the time to put it all togeather, make it work and THEN install the regulator.

But we would very much like to get togeather with someone that has the sub, so that we could concentrate on just the electronics. It would speed up the testing process a great deal.

But I think we're gonna try and get in contact with some of the northern-most germans. Beeing so close to Alexander Engel KG, makes them prone to using piston tanks I guess...

But I'd still like to take you up on that dinner offer, when I get to Canada someday.

[mylo]

JSL,

Yep, just stop in,

JSL,

Yep, just stop in, we'll do lunch, my treat.

Mylo.

[fbradasc]

Hi Jacob,
will you publicate the

Hi Jacob,
will you publicate the wole documentation (electrical schematics and software code) in an open source form (under GPL or some other open source license) ?

[aeroengineer1]

I am getting interested in

I am getting interested in using piston tanks, and your system looks to be ideal. I really liked your documentation, I think that the teachers here would give that paper a bad grade due to the fact that the information was too easy to read and not complicated enough. Anyway, what might the price range be for the standard model. Just a price range, I know that you sould probably not have a final value. Also, what is the approx size of the unit? Looks great, keep going.

Adam

[jsl]

Hehe... That datasheet was just

Hehe... That datasheet was just a little something extra. If you want it complicated, the final report is a 135 pages, though in Danish.

It looks like we have found the money to get ourselves a sub. We should have agreed on something by the end of the week.

With regard to the Regulator, we're thinking of doing a revision now that we're gonna make the two other model sub variants of the regulator. It will become somewhat smaller so that it will fit into smaller diameter WTCs.

So... The pricetags are extremely hard even to make an educated guess on right now. On the other hand, we're likely to document the follow-up project in here anyway, and as soon as anything can be said about a possible price tag, I'll post it.

[Guest]

How about a kit, for

How about a kit, for those of us handy with a soldering iron?

Or at least sell the boards with some docummentation and a programmed microcontroller.

Andy