IWETHEY v. 0.3.0 | TODO
1,095 registered users | 0 active users | 0 LpH | Statistics
Login | Create New User
IWETHEY Banner

Welcome to IWETHEY!

New IANTT
;-)

I can't find the posted description of your boat. How big is it again? 25'? How much displacement?

You might want to take a gander at the [link|http://patft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&p=1&u=%2Fnetahtml%2FPTO%2Fsearch-bool.html&r=0&f=S&l=50&TERM1=tiller&FIELD1=&co1=AND&TERM2=autopilot&FIELD2=&d=PTXT|tiller and autopilot] patents listed at the USPTO. There is at least one listed for cases when a lone pilot falls overboard. It might give you some ideas for your case.

I suspect that someone has already solved your problem, but finding it may be difficult. A nice challenge. :-)

I can certainly write a controller in C, but I have no idea what sort of hardware to get or how to interface a processor to a 10k pot for position feedback. In short, I don't know hardware at all but I'd like to learn enough to be able to build a controller for this thing.


If I understand your description and the Nook catalog, you want to relate the measured resistance to the rudder angle and then use your code to convert that into a feedback loop to control your heading on the water. There are several ways to approach that interface problem, but it's probably simplest to use an existing general-purpose controller. You can have your software interrogate an analog interface board that measures the resitance and converts it into a 12 or 16 bit value that your program can read. There are simple examples of things like that in this [link|http://www.cs.princeton.edu/courses/archive/fall05/cos436/Lab2/index.html|Princeton computer lab] page. It uses a [link|http://www.parallax.com/detail.asp?product_id=BS2P24-IC|BASIC Stamp] controller (or a variation on this one). It's programmable in BASIC. There are a vast array of similar controller options out there.

Alternatively, if you know the type of circuit you need, places like [link|http://www.national.com/support/techdoclib.html|National Semiconductor] have lots of documents on example circuits that tell you exactly what you need to implement it. Going that route is a bit more involved but more flexible.

Hope that helps.

Cheers,
Scott.
New No - nothing off the rack works
Vessel is 36' on deck - 40' overall and heavy displacement - 12 tons or 24,000 lbs.

Closest thing on the market is [link|http://www.anchorexpress.com/a12006.html|http://www.anchorexpress.com/a12006.html] which claims to be OK for boats up to 13,000lb. Boats like that will be fin keelers and have light control forces.

The RayMarine stuff also gets dinged a lot for being unreliable and frequent failures.

Simrad's largest is the TP32 [link|http://www.anchorexpress.com/simrad-tp32-tiller-pilot.html|http://www.anchorexp...tiller-pilot.html]

Roughly same limits. 14,000 lbs (6,300 kg) and 39 ft.

Its not a force feedback system - there will be a compass readable over NMEA protocol, along with a desired course. A difference will necessitate some control input. The magnitude of the difference will determine how much. Once we attain the heading, neutralize the inputs. There will be some amount of dynamic "center finding" in the algo to compensate for leeway/weather helm/prop walk effects.

I'll also add that I only need the pilot for motoring. I have a great [link|http://www.selfsteer.com/windvanes101/index.php|servo-pendulum windvane] self steering system for use under sail.

However, when I was bringing the boat up from San Francisco, it was motoring for a week directly into 10-12 foot swells. Hand steering that got old fast.

IANTT : INSTITUTO DOS ARQUIVOS NACIONAIS TORRE DO TOMBO??? I have no idea.



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
Expand Edited by tuberculosis May 6, 2006, 03:00:30 PM EDT
New Variation on IANK
I am not Karsten. -> I am not TonyTib. :-)

Oh, [link|http://store.wmjmarine.com/s21113402.html|here] you go:

Simrad HLD 2000 long dual (13" stroke) Hydraulic linear drive 21113402

For boat displacement up to 77000 lb

Stroke: 13.3 inchs Peak thrust: 2310 lb max torque: 28000 lb Tiller arm: 11.7 inchs

Works with AC40


I don't know enough about this stuff (e.g. is AC40 == 40V AC?), but it looks like [link|http://www.simradusa.net/index.php|Simrad] has what you want if you're willing to spend the $$$.

HTH. Good luck!

Cheers,
Scott.
New No hydraulics
too messy, will require engine power to drive it, and besides, I need 24" of travel.

I like the screw jack positioning mechanism - solid - no power to maintain position required.



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New Good points.
Getting the screw drive to work would be a good solution. The robot guys should be a great resource. (Where's [link|/forums/render/user?username=FuManChu|Bender] when we need him? Oh, that's right, he's [link|http://aminus.org/blogs/index.php/fumanchu|here].)

Good luck!

Cheers,
Scott.
New Here we go: http://www.seattlerobotics.org/
Guess I'll attend the next meeting. Hey Thane - wanna meet some robot hackers?



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New That's probably a good resource
although typically the robot enthusiast groups deal with wimpy little robots. But, it's likely some have good control experience, analog experience, and motor control experience.

Try looking up PID control (Proportional Integral Derivative) - that's what is normally used for closed loop control, e.g. on all the servo systems we build (and for temperature control, etc) (there's also fuzzy logic). We use commercial motion controllers or PLC's to do the real time stuff.

You'll have to interface to the feedback system (analog front end plus A/D converter to get from resistance to voltage to bits) and jack screw (motor driver of some sort).

I'd recommend using something with a bit of computing horsepower so it's easier to write the NMEA code. There are plenty of little boards available; if you want to use something other than assembly, BASIC, or C/C++, get something that gcc runs on (e.g. x86, PowerPC, ARM) and you should be able to use Objective C via gcc.

Also, it's likely you can visit eBay or other surplus sites to get commercial grade stuff (e.g. brush or brushless motor amplifier) to make your life easier but still be affordable.

--Tony
New I'm happy with C and more familiar with fuzzy methods
for control. Anyhow I envision a couple different modes of operation. The simplest is just "keep the boat pointed that way" and this should be built into a dumb controller that works by itself using just the compass and tiller position data. So the autopilot needs a microcontroller to do this.

A more advanced strategy would be to have a computer figure out corrected headings based on actual boat motion to compensate for current/leeway and allow the boat to track a course. Final goal is I should be able to punch in lat/long and boat simply goes there - using charts and radar to avoid traffic/obstacles.

While I distrust technology - I think a staged and distributed approach will be best with each system fully manual, dumb semi-automatic (state maintaining), and then smart central control to tell the dumb systems what their states should be. This way, if the main computer dies, I only have to babysit the systems from time to time. Ditto if a subsystem (steering, motor control, whatever) fails.

Sort of a smart home approach for boats. Commodity parts for ease of replacement will be favored wherever possible.




[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New Many different approaches
There are many different ways of building such a system. Three examples are using a PLC (more industrial approach, but more costly), using a single board computer (ZWorld, New Micros, and many others), or using an industrial PC (e.g. PC104 system)

It might be good to see what the Seattle Robotics people like - then you hopefully can good feedback and help. Or pick a system with an active mailing list.

Protection from the elements is going to be a big concern. It may be better to build your own enclosure - you can get splash resistant systems (e.g. IP67 using sealed connectors), but they cost considerably more.

BTW, they're probably too pricey and I don't have any experience (not a good fit for our applications), but I've always found [link|http://www.exlar.com/|Exlar's] linear actuators interesting.

--Tony

New I've gotten a few good ideas
from the robotics people. Someone pointed me to this:

[link|http://www.digitalmarine.co.kr/prod_mif_e.htm|http://www.digitalma...kr/prod_mif_e.htm]

Unversal NMEA converters. That should make things like rudder angle/rate easier to read.

Other advice I've gotten

"By using an off the shelf PLC controller like a Siemens s7-200, or a
micro logics Allen Bradley, both available in 120V or 12VDC and other
voltages, could receive analog input via the NMEA universial converter
on desired compass heading from a remote navigation program, the PLC
controller utilizing Proportional Control (PID control build into these
small controllers) would send left or right digital outputs to an
electrical actuator (like those pictured at Nookindustries, but they are
also availble from Graingers in South Seattle off 4th Ave) An encoder
could be attached to the rudder that provided a 4-20 mA signal relative
to position, BCD or several other codes could be sent back to the PLC
for feed back. Both the Siemens and AB conrollers have small LCD touch
screen interfaces that are easy to configure and setup."

So I'll have to research this stuff some more.



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New Avoid A-B; PLC has good and bad sides
We've used A-B, Panasonic, Automation Direct, Omron, and Siemens PLC's.

The PLC approach does have its advantages. They are industrial quality, with various input (24VDC, 120VAC, analog in, etc) and output (24V, relay, analog out, etc) options, while the C-programmable stuff is typically bare boards with TTL (5V) inputs and outputs. If you need 4-20mA, you'll have to check that the PLC has an option for it.

Some do have built-in PID; we've done PID temperature control. If you want to do PID, make sure it's built in.

Older PLC's are programmed solely in ladder logic, which is a fine fit for combinatorial logic ("when these inputs are in this state, do this"), but sucks for everything else (e.g. sequence of actions). Newer PLC's support the IEC61131 standard, which supports five langagues: ladder, function block (ladder written with AND/OR instead of wires), structured text (our favorite; a mix between Basic and Pascal), flow chart, and ladder assembly.

PLC's are horrible for any kind of data handling, so if you do use them, that recommended approach (use PLC as PID controller, don't do NMEA conversion) makes sense.

The price will be higher, and you'll have to pay for programming tools (A-B's are very pricey, and have copy protection; Siemens, Aromat, Automation Direct are more reasonable) or know someone who has them.

Some companies have starter deals which include a PLC and software. Aromat has run specials on PLC's and Operator Interfaces occasionally. Siemens runs them often (including both PLC and Operator Interface); you will need to contact your local distributor.

I recommend looking at [link|http://www.sea.siemens.com/automat/product/plc/auov.html|Siemens], [link|http://www.nais-e.com/plc/|Panasonic], and [link|http://www.automationdirect.com|Automation Direct] PLC's. Automation Direct has all their pricing on line, and it's competitive. (A-B does has on-line pricing, too, if you register, but it's quite high, even after a 30% discount). Then you can use resources such ebay and [link|http://www.plccenter.com|PLC Center] to see if you can get surplus or used equipment at much better prices.

--Tony
New Maybe you could barter
an improved web site for PLC Center for some of their surplus stuff. Check out their "advanced search" web page, takes forever to load - I think because they're loading a whole long list of all the possible companies to search for.

--Tony
New Prolly a dumb question, but...
...Have you tuned the rigging? By creating the right amount of mast rake (and placing the mast "properly"), you can minimize the amount of weather- (or lee-)helm, which in turn tends to minimize the amount of force on the tiller.


(But you already knew that, dincha?)
jb4
"So don't pay attention to the approval ratings that say 68% of Americans disapprove of the job this man is doing. I ask you this, does that not also logically mean that 68% approve of the job he's not doing? Think about it. I haven't."
Stephen Colbert, at the White House Correspondent's Dinner 29Apr06
New This is for motoring only
I have a wind vane. Problem when motoring is prop wash strikes rudder and makes control forces much higher. Observe vessel diagram.

[link|http://www.capegeorgecutters.com/36/|http://www.capegeorgecutters.com/36/]

As for rig balance - Aurora is exceptionally well balanced - which is why the wind vane is so often when sailing - but it is worthless motoring.



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New I see...thanx for the visuals
I understand now what yer looking for.

OT: Did you hear about the wildfire in the Bosque?
jb4
"So don't pay attention to the approval ratings that say 68% of Americans disapprove of the job this man is doing. I ask you this, does that not also logically mean that 68% approve of the job he's not doing? Think about it. I haven't."
Stephen Colbert, at the White House Correspondent's Dinner 29Apr06
New No, when was the fire?



[link|http://www.blackbagops.net|Black Bag Operations Log]

[link|http://www.objectiveclips.com|Artificial Intelligence]

[link|http://www.badpage.info/seaside/html|Scrutinizer]
New last couple of days...
[link|http://www.koat.com/news/9159092/detail.html|here] and [link|http://www.koat.com/news/9169141/detail.html|here] are some (rather terse) references to it. (They are from KOAT-TV, so you know what you'll get...). Seems it was a less-than-controlled burn that got away from them, south of the Bosque. Apparently, a small portion of the southern part of the Refuge was scorched, but damage was said to be minor. The Owl Bar is safe (whew!).
jb4
"So don't pay attention to the approval ratings that say 68% of Americans disapprove of the job this man is doing. I ask you this, does that not also logically mean that 68% approve of the job he's not doing? Think about it. I haven't."
Stephen Colbert, at the White House Correspondent's Dinner 29Apr06
     Hardware hacking tips requested - tonytib? - (tuberculosis) - (23)
         IANTT - (Another Scott) - (16)
             No - nothing off the rack works - (tuberculosis) - (15)
                 Variation on IANK - (Another Scott) - (9)
                     No hydraulics - (tuberculosis) - (1)
                         Good points. - (Another Scott)
                     Here we go: http://www.seattlerobotics.org/ - (tuberculosis) - (6)
                         That's probably a good resource - (tonytib) - (5)
                             I'm happy with C and more familiar with fuzzy methods - (tuberculosis) - (4)
                                 Many different approaches - (tonytib) - (3)
                                     I've gotten a few good ideas - (tuberculosis) - (2)
                                         Avoid A-B; PLC has good and bad sides - (tonytib) - (1)
                                             Maybe you could barter - (tonytib)
                 Prolly a dumb question, but... - (jb4) - (4)
                     This is for motoring only - (tuberculosis) - (3)
                         I see...thanx for the visuals - (jb4) - (2)
                             No, when was the fire? -NT - (tuberculosis) - (1)
                                 last couple of days... - (jb4)
         bah, new fangled things, whatever happened to lashing it? -NT - (boxley) - (5)
             Tends to drift off course after a few minutes when motoring - (tuberculosis) - (4)
                 under power eh? let me get back to you -NT - (boxley)
                 What kind of electrical power do you have to run this? - (boxley) - (2)
                     12VDC -NT - (tuberculosis) - (1)
                         add $40 for a decent transformer -NT - (boxley)

Then again, I think our walls are made from the salvaged hulls of exotic alien spacecraft so YMMV.
67 ms