Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length.

[derekwarner]

Hi all....& a great build Andy......... ... I am new to Mayhem but haven't seen any references in this thread to pointing & non pointing Zones...[I too could not download the self installing software as above]

In modern day warships, these complex Zones are controlled by computers, when the Dreadnought was built the same complex Zones were controlled by mechanical computers

ie., to intrinsically prohibit [not limit] firing ordnance over [OR AT] any part of the actual vessel

A better example is say a Doyle class FFG...when she deploys from port...the MK 75 gun is trained at ZERO pointing toward the stern ...also with ZERO elevation, being the axis of the vessel...just aft of the MK75 is the Phalax [CIWS] ...so the non pointing Zone for the MK75 is a pre determined arc centered off the  Phalax when the MK75 is @ ZERO train + ZERO elevation, however a pointing Zone for the MK75 could be ZERO train but with + XX degrees of elevation.....

Sorry to go off at a tangent.............Derek 

[Capricorn]

Andy, I'll give it one more try (the shockwave download/install), I'm sure it's not that difficult.  Sorry I didn't read back much, nasty habit I have, lazy I guess.  I didn't realize you were having it track the target on a heading too, wow, that's pretty niffty.  I guess once you've got the onboard computer it's not a stretch to do that part of it as well, the electronic compass might give you the global reference you need.  Now how come your boat is static?  It seems it should move so that you can watch it track the target, will you build a turntable for it?   (hope I didn't misread or not read enough again).  I'll save up for some microprocessors (I'll need at least two as I tend to melt electronics)   

Derek, your pointing zones are as far as I got and am building a mechanical director that will control the zones of the five turrets (there is only really 2 zones on mine).  As simple as it is, doing it mechanically has been a puzzle, for me anyway, interesting though.  I'll basically have a CW CCW control for the director on the radio, it will be able to turn 360 deg.  The guns will follow the director unless they go to the no fire zone at which point they return to their zero position and wait.  Fortunately I have about two cubic feet of space for it !

[derekwarner]

Hi all...Capricorn & Andy...again this is a little off the original thread but see how we go

My experience is not sea time based, but 1504 days as an above water weapons forman at Australia's Garden Island Naval Dockyard in Sydney many years ago - with British or Australian designed vessels....the turret designation's from the bow going aft ...post WW11 were A , B, X & Y but the NPZ are far more complex than you may imagine [now simply termed as A, B or X...and PZ or NPZ]

The B  may be physically higher in elevation than the A .....but if the A barrels are @ say 30 degrees elevation, with the B turret at Zero elevation......the barrels of the A would be within a pointing Zone of B.... ....so hence these positioning revert to NPZ

In real life warfare, a turret that progresses toward a NPZ does not return to Zero elevation or train.....but waits for the vessel fire control system computers to alert the scenario & has the potential to offer the control & alter the vessels course to provide the required trajectory via a NPZ be it for a shell or missile

Your scenario of ....The guns will follow the director unless they go to the no fire zone at which point they return to their zero position and wait ....certainly be the most practicable course to follow ........Derek

[dreadnought72]

Hi all!

Sorry for your problems with the Shockwave. This is the direct link to the example, which might load for you. I know there's been recent trouble with IE and some Flash players.

The turrets on the Dreadnought were (bow to stern and left to right) A, P, Q, X and Y. I believe there were mechanical limiters built in to avoid any (ahem) "embarassing moments". 

Each turret has an arc of fire greater than normal servo throws, so on the model they'll need to be geared up to rotate the extra distances. The way the shockwave demo works is that the normal, parked position of each turret is at 50% servo throw for turrets A, X and Y (since these can rotate left and right of their rest position) and at 0% servo throw for turrets P and Q - these wing turrets can rotate 180 degrees aft of their parked positions to cover each side.

There are relatively cheap microprocessors that can export servo pulses to a high accuracy, which makes setting the position of each turret straightforward - just a bit of maths - though any sent bearing angle will require different outputs for each turret A, P, Q, and the X/Y pair (which can train over the same angles). While making the demo, it was easy to make a mistake with regards to the direction of rotation desired - but faults like that will be ironed out before the real thing is operational.

Making the target bearing independent of the ship's heading was the next challenge, and I'd had thoughts of using model helicopter gyros or a home brewed gyro compass - but neither seemed suitable or practical. The idea of a solid state electronic compass wasn't one I'd thought of, so full kudos to Capricorn Joe for raising this as an option. There's not much metal in the model, and I think I can distance a compass like this and faraday-cage the motors in order to prevent unwanted magnetic effects, but as you can imagine the director and guns should then be able to track a "target" (albeit one that's stationary and at infinite range) independent of the ship's heading changes.

Which would look good, to my thinking.

Andy #1963#

[Capricorn]

Interesting stuff especially since I'm in the midst of it right now, thanks for all the info derek.  Andy I sure know what you mean about the mistakes in rotation direction, I'm counting on them being ironed out by reversing wires as needed.  I print out the wiring for the director and sit down and try to check it and I see the polarity is wrong in one circuit, so mark it up, then I follow it through and decide no, the polarity was right, then the third time...  Just curious on the servos, you mention gearing them up, are you concerned with them rotating the turrets a bit fast?  I'm sure you have that covered, but if not you can imagine taking the pot out or the servo and gearing that to match the turret limits you want, and then gearing the servo shaft down to turn the turret slowly.

I could not tell you if the compass would work inside a model boat with all the motors and metal parts, seems you know enough - faraday cage...  I will say that I'm concerned about all the servo's I'll have on my boat, with quite extended leads, mainly the steering servo way at the back, that's an important one, but I'll use them to elevate the main guns and probably for the bofors.  Any thoughts on that?  Cap

 

[victorian]

Would GPS be a better way? Apparently some receivers output course data in a serial format called NMEA. GPS course data is derived from position changes and is immune from magnetic interference.

[dreadnought]

Hi Andy

Well now that the Christmas rush is over I am again getting down to business on the

Dreadnought build.   As you mentioned earlier the planking on the compass platform

runs for and aft not the other way as shown in the AoTS Dreadnought drawing.

I am now about to proceed with the navigating platform and was wondering  which way

it was laid..

Checked out your build today,  did not know how to do it up to now. (not very computer

savvy) sure is impressive.  I will be up to date on it now though.

Hope you have a good new year.

Dan  (Dreadnought)

[derekwarner]

Hi all......Andy......your representation as per the link is very impressive...

http://personal.strath.ac.uk/andrew.goddard/dreadnought/turrets.swf

but how will it work?

a) does the director have a preset program of target locks?...to which the respective turrets follow?
b) will the barrels elevate?

You may see old footage (Iowa Class] of each individual barrel in a turret returning after firing to Zero elevation at different times....this essentially was a function of shall we say basic manually controlled loading  & the requirement that one gun could be taken off line without affecting the other two

Many years ago at work, I had the pleasure of witnessing a multiple dummy load + train + elevate in the STDB compartment of the FWD turret on BB663

From Capricorn........"and then gearing the servo shaft down to turn the turret slowly".

I understand the Dreadnoughts were 12 " as against 16"...... so scale is different here, however would consider both were a majestic but relatively slow process of full train & elevation

So Andy, Capricorn & now Dan.....keep up the good work  all very interesting ........Derek

[snowwolflair]

"and then gearing the servo shaft down to turn the turret slowly"

If it is any help, you can get a piece of electronics that will do this.  It just fits inline on the servo lead and slurs the signal so to speak.

[derekwarner]

Hi all...........snowwolflair .......do you mean a servo slowdown?.....which are adjustable & maintain high torque? & about AUD$20.00 each......what a great idea.....Derek

[snowwolflair]

Yes, thats the ones.

[Capricorn]

Servo slower?  What a great idea, I heard you could get slow servo's but they are expensive.  I'll look for servo slower, I'm not up on all the electronics obviously, still back in the industrial era.  Cap

[Capricorn]

Andy, Since it's all on the computer, it seems it might be easy enough to have a secondary control for the P and Q turret parked position.  It's just a minor thing, I see when you point to 185 the P turret points to 185 and Q to 0, when you switch over to 175 the P turret rotates all the way back to zero, while the Q turret rotates to 175, if the target is at 180, you have a bit of turret rotating going on when it moves either side of 180.  A simple solution would be to have the P and Q parked position at 180 with the target is anywhere 45 degrees from 180.  Anyway, just thought I'd throw that out there, without off center turrets to not have that exact issue but the rear turrets will osillate around 45 and 315, and the front ones around 135 and 225.  I'm ignore it though . Cap

[Colin Bishop]

You might like to have a look at ACTion's Servomorph. It sounds as if it might be what you are looking for.

http://www.action-electronics.co.uk/pdfs/P96.pdf

Colin

[Capricorn]

Colin, that's a the most useful gadget I've ever seen  .  I've been searching a bit and haven't found much, the Dionysus servo rate reducer, hyperion servoslow, but they must not be very common, thanks for the info, I'll look into that too.  Cap

[dreadnought72]

Hi All,

A bit rushed, but answering as many points as I think were raised:

Re: turret rates and directions. I don't need to swap wires or add in servo-slowing controllers (though servo slowing will be needed - it took the Dreadnought about a second to train a turret four degrees link here ). The easy way to do all this is in the coding on the microprocessors.

The servo output command on the very cheap Picaxe microprocessor, for example, is set to a digit between 75 and 255, which outputs a signal between no-throw and full-throw for the servo (an accuracy of about half a degree on the servo, and maybe one or two degrees on the turret, depending on the gearing necessary). So in the programming loop, I can simply send the turret towards the desired end location at any rate I like - the servos will then run like (smooth) step motors to a final position. Since this rate will be slower than the servo rate, there won't be any chasing.

Gun elevation - it would be very nice, but I'm not sure it's as visually important or obvious as tracking. And - on a more practical level - I'd be needing another ten servos. 

Parking positions for P&Q turrets. I see the meaning here - and there's another issue with regards to X, which has a dead zone aft, while Y doesn't. Possibly the easiest thing is, as you say, to leave a turret at its "last useful train" until that train moves such an angle away that it's worth parking the turret. Again, this will be software controlled, so it will be easy to adjust/augment this once the system works.

Setting the director position. This is going to use three functions from the TX. One stick has been "desprung", and can be placed and held at any position in the square box that it can move around. This will be used when the "train" function (an on-off switch on the TX) starts the main programming loop. The position in the box of the stick will be converted to an angle of target bearing, then the turrets will do their thing until another train function is sent or - now I think local bearings can be calculated using the compass - they will track the target independent of the TX, and relative to the turns of the model, until another train function is sent.

And finally - if I can acquire and track, then a "fire" option is becoming an idea. I've seen a fabulous photo of a (UK) model of a pre-Dreadnought two-turret battleship "firing". I need to think about this some more, but I have a few ideas - and a sack of Christmas Party Poppers to play with! 

[derekwarner]

Hi all......Andy, Capricorn & Dan...re elevating turrets

Many years ago, I saw an example of turret elevation by a simple V shaped cam ring as per the attachment

I we consider the A turret, the shaded section turret stationary follower is a cutout of the ring & could be centred @ ZERO train & then diminish in height to any required arc to STDB or PORT - the barrel is spring loaded to the stationary follower ring

Depending on the method of construction...the turret stationary follower ring could have a negative cutout or a positive opposite profile lip etc

So from ZERO train...any + or - train results in a + elevation either side of ZERO....does not require any additional servo function....Derek

[dreadnought72]

Statistics!

I see I've spent nearly a week on Mayhem, and made 254 postings. Hmmm ... time spent when I should be building!!!  But at least it means I read more than I type!

When I set up the Dreadnought site about four months ago, I stuck on a stat counter to monitor its usage. I'm pleased to relate that the count (which is invisible) passed 500 in December - no bad figure - and that nearly a quarter of all visitors have hung around for more than twenty minutes. Which would be a success for any commercial site, and must mean they've either read the lot ... or maybe fallen asleep at their keyboards ...

The bad news? Five of the visitors were using Netscape. Ugh. And a few are using screen resolutions popular in the eighties.

The good news? The visitors have come from all around the world. If there's any interest, I'll bung up a map. The bulk have been (no surprise) from the UK and western Europe, but there's a fair few Americans, Australians, and a couple of Russians stopping by. Interestingly, after the url appeared in a Polish model making forum, Eastern Europe is well represented.

Perhaps I should have a "sponsor a plank" week? I think it's the only thing that'll get me to finish the decking! 

Kind regards and a Happy New Year to all. Except to my Chinese visitor, who doesn't get his well deserved 新年快乐 until the 29th of January...

Andy #1963#

[Capricorn]

Derek, that's a clever method to elevate the guns especially for smaller scale models where there isn't room for much else.  Andy, congrats on your site activity, I may have skewed the results slightly, I have a tendancy to read a couple sentences and then wander off scratching my head.  I originally intended to fire BB's from the 5" turrets on my model, then I saw a tank model that fired .22 cal blanks and I was hot for that (quite an impressive design, complete drawings for it posted by the fellow), I've now sort of decided to use ring caps (the type for toy cap guns).  Basically the gun barrel is the hammer, an envolute shaped cam releases the sprung barrel which detonates the cap, as the cam continues to turn it "slowly" recoils the gun.  Main problem is rigging up the cylinder to turn and get it to all fit in the turret.  Obviously a single shot would be fairly simple, a single cap would have to be manually loaded each time though which could be quite a project depending on how your turrets are fastened down.  Cap  (attached schematic of where I'm headed)

[dreadnought72]

Now that's a neat solution. And I suppose once you've got a cam in place, you can gear down a rotor to carry strips of caps into the right position for subsequent shots? (Trying to remember my cap pistols of <ahem> some 40 years ago...)

Andy #1963#

[Capricorn]

Andy,  There have been some changes in the cap gun world over the last 40 years.  They have ring caps, a plastic ring with either 8 or 12 shots, quite a bit more pop compared to the paper ones.  I choose the 12 shot naturally, that means I need a 12:1 ratio between the cam and the cylinder, however12:5 works too. The toy guns use a ratchet of some sort so when you pull the trigger it turns the cylinder 30 degrees, that should be simpler than a 5:12 bevel gear but it hasn't panned out for me yet .  I mean to pick one up but most common here is the eight shot, I think they are cheaper.  Anyway, it's a ways off yet for me.  Cap

[Capricorn]

Now look what you've done, I've been divererted from the task at hand with this talk about gyros .  I always like to start from scratch with everything I do, sort of neanderthal that way, but will allow myself to be steered as well.  Andrewh helps that way, I got on this idea of adding a gyro to my boat now that would theoretically allow the gun elevation to be keyed to a gyro rather than the boat hull, so as the boat heels the guns remain pointing at the same elevation to the horizon (nothing new or particularly complex there).  I just don't know if it's something I can build easily enough or not.  Let me know what you think of the idea, as I mentioned to Andrew I need a servo tester that has two qualities, a) the knob is synced with the servo (ie turn the tester knob 5 degrees and the servo turns 5 degrees) and b) has a pot knob that doesn't take a gorilla to turn (ie the pot turns easily enough that a gyro can hold it in place when the boat heels).  Any input great (if I'm invading your string unwanted let me know) Cap

[dreadnought72]

Hi Cap,

Reading Patrick O'Brien, and back in Aubrey's day, they'd have "fired on the roll" ... which isn't a nice thing to do to bakery products, but there you go... 

Gyro-stabilised guns? Like the things modern tanks have?

I don't think it can be as simple as a one-pot-shop. For example, if the turret's trained to port, the movement of the barrel will have to be the inverse of when it's trained starboard, for any given angle of roll. And if you were to take pitch into consideration, then you'd need two-axis calculations, plus info on which angle the turret's trained, before you could generate a figure for a suitable "level".

But it gets worse: once the turret axis is off vertical, and if the ship is rolling, pitching and yawing, then you can calculate a gun angle which will remain locked on a particular vector, but it'll be having to train and elevate/depress constantly in order to do so - and you'd best be in the world of <gulp> quaternions. 

A not-very-brief aside (and apologies to all if this is stuff you know) -

Back when NASA did proper manned stuff in space, rather than just ferry toilets the eight minutes to low Earth orbit, there was a thing called gimbal lock. It makes an appearance in the film Apollo 13. When the axes of a gyro line up, the gyro's information becomes useless. Your heading information is lost. That's gimbal lock, and it's best avoided if you want to fire a thruster with any idea of where you'll end up.

Because modern jet aircraft and smart ordinance can end up at any old angle, and we have computers that can do impressive sums in real time, there was a move towards using quaternions - which do not suffer a mathematical version of gimbal lock. These were invented around the end of the 19th century for no reason whatsoever, and lie quite squarely in the "insane" area of mathematics. But with them, you can happily add or subtract any angles you like from a start position and rest safe in the knowledge that the final angle is "correct".

I got an A-level in Maths back in 1981 (and by 'eck it were tough back in my day) and I use maths everyday in my work. I needed to use quaternions for some 3d animation modelling I was working on a year or so ago. Now, maybe it's my age, but it took me a good fortnight to begin to get my head around what was happening in the calculations, and I had to assume the use of the square root of minus one had any "reality" (they appear in quaternions), and finally FINALLY! got the model working.

Within a day of writing the script, little of the maths made sense. I had to just accept it. A month later and it was like looking at ancient Mayan. Today I've returned to the blissful ignorance of not understanding them any more. Have you ever read any H.P.Lovecraft? That's quaternions.

In conclusion -

Gun stabilisation would look extremely cool for modern single-gunned frigates and the like, cutting speedy S-curves and rolling in the bends. But it's a can of worms I'd approach with extreme care! (And wouldn't gyro-controlled stabilisers be better - a one-axis, and therefore considerably easier & more practical application?)

Andy, fitting gunwales to the 42' sailing launch...

[andrewh]

I got an A-level in Maths back in 1981 (and by 'eck it were tough back in my day) and I use maths everyday in my work. I needed to use quaternions for some 3d animation modelling I was working on a year or so ago. Now, maybe it's my age, but it took me a good fortnight to begin to get my head around what was happening in the calculations, and I had to assume the use of the square root of minus one had any "reality" (they appear in quaternions), and finally FINALLY! got the model working.

Andy, fitting gunwales to the 42' sailing launch...

Andy!  When you have used the Quaternions, PLEASE put them back.  I went to look for some t'other day and the jar was empty.  I eventually found enough way up the i axis
andrew

btw - aren't they something to do with malaria?

[Capricorn]

Andy, I don't know what quaternions are, it's not in my old (Websters) dictionary so it doesn't need to exist and can therefore be put in the Oxford dictionary   (or in a jar).  As far as gimble lock goes I will ignore that too because I only have one gimble so it can't get confused.  But you are probably right about it, let me just make sure I conveyed how it would work.

The gyro would consist of a brass disc mounted directly to an electric motor shaft (high speed), the back of the motor would be mounted to a piece of wood with a rod through it and a ball bearing each end.  So it hangs below the horizontal axle (more like a pendulum then a gyro maybe).  Turn the motor on, it spins and hangs straight down.  That is supported on a bracket to suspend it and the bracket is sitting on a turntable, the turn table is geared to turn with the director.  The axle of the gimbal is always then parallel to the axle of the gun trunnions.  When the guns are trained to port and the boat rolls the gyro will stay vertical (and adjust gun elevation accordingly).  If the guns are trained dead ahead the axle of the gimbal is perpendicular to the axis of the boat, if the boat rolls the gyro will roll too* but only if the boat pitchs will the gyro change the gun elevations.  *here's where the other missing gimbal could be a problem, the spinning gyro will have to be rotated - is that causing a problem with quaternions?  Or are you telling me to use 3d cad?  Except for that problem (which might be solved by adding the missing gimbal back in) I see it as flawless !   Also after looking at geometry it's sensible to connect the end of the gimble axle to the pot with servo arm and link so disparity in angle of rotation of the controller pot and servo's can be adjusted for (we think). Cap