1/500 2199 Yamato

[Madhatter]

thanks bud

Yes, I have finally watched the entire series from start to finish and I really enjoyed it. I've gone back to it a couple of times now for references.

I spent a lot of this afternoon making the base for the engine LED's. Finding room for 12 3mm LED's in a circle only 2 cm across isn't easy and it took me 4 attempts to get it right. I went to install the LED's only to find I didn't have any Yellow ones, so I had to order some for arrival early next week. I'm not in any hurry.

I hope to try finishing off the guns tomorrow but given how fiddly they are, I doubt it. I am also going to make a mould of the engine nozzle tonight - so we'll see how that goes. If it moulds nicely and correctly, I will have a crack at making the live action version without having to worry about screwing it up.

Anyways, thanks for looking

MH

[Misrule]

As I recall one of the observation decks in the hull is a steam bath, vital kit for a battleship I am sure

[Madhatter]

LOL! There is no way I can replicate a steam bath I'm trying to find decent head on shots looking through the glass into each of the observation rooms to make into decals to go on the back walls of the light boxes.

If you have any or know a particular scene in the series that maybe of use, please feel free to share

On a separate note, I have successfully cast an exact copy of the engine nozzle using a 2 piece mold and have recommenced casting of the petals so it looks like I should be able to make the live action version nozzle after all - or at least, something closely resembling it

[Misrule]

This is the crest I believe

http://vignette1.wikia.nocookie.net/yamato/images/d/d0/Faeb7d47a4.jpg/revision/latest?cb=20131114013710

I'll get some images of the inside of the observation rooms, they are split over two decks with pools on the two lower sections, ward room to starboard upper and observation deck to port upper. Pretty plain and empty I think so you might have an easier time of this at least

[Madhatter]

greetings once again

Well, I'm feeling somewhat more human again and have managed to get some progress happening.

I managed to cast a great copy of the engine exhaust and finished casting the petals and I even put them onto the nozzle. However, I am glad I took the extra step of casting a copy of the nozzle because I am far from happy with how the final item finished up. It looks terrible! This is all my own fault of course - I should've just made 16 petals from plastic instead of trying to take a short cut and cast the others because I ended with varying thicknesses of resin and this led to a wonky final result. Well, it was worth a try and at least I can say I gave it a go!

I even had a small moment of triumph when I successfully managed to get nav lights into the wing behind the funnel. I realised that the panel line that runs at the rear of the support also connects with the rear line of the wing, so I carved out the lines and ran a fiber through it and down the support into the base. The fiber even produced light at the end. I was chuffed! But, I swear I am destined not to have nav lights in this beast because this afternoon I accidentally knocked the superstructure off it's perch and it managed to break the wing from the support and snap the fibers. So no nav lights there, unless I rip it all apart and try again, which would be a PITA because I had also added detail to the support and if I was to take it all apart again, I would loose the detailing. It wasn't fun getting it all on there to start with. Needless to say, that happiness that I had achieved something some said to be impossible vanished in an instant.

Still, life goes on and so does this build. I managed to get the lighting into the third bridge and that is now constructed ready to go on. The fibers run right through the middle and it was fun threading it all through in a nice straight line.....not.

I also have worked out how to get the SMD's into the range finders for the main batteries and have done 4 out of 6 with the last 2 soon to be done. this is very fiddly work with the SMD's having a hard time being glued to the rear of the range finder wall. Again, my patience gets tested but the results are worth it.

The observation deck light boxes are also together and ready for the LED's to go in. I will do that tomorrow

And last but not least, the engine LED disc is done but no sign yet of the LED's that are to go onto it. I expect them to arrive in the next couple of days. once the arrive, I should be able to get them on easily enough (famous last words)

Well, that's it for now, but thanks for looking

MH

[Will Vale]

Sorry I haven't commented, I've been stuck in my own figure corner (plus being a single parent for 3 weeks hasn't helped) - this is AMAZING. You're definitely raising the bar, the odds of my picking up my 1/500 Yamato just decreased somewhat

Will

[Madhatter]

No worries mate I reckon you ought to bust yours out! You know you want to

I got my first gun done the other night and it looks great. The SMDs even work! Yay. I'll post pics later

[Homerlovesbeer]

I'm following this for sure.....so good!

[Madmonk]

All that fantastic extra detail, some would say you were as mad as a hatter!!!

[Madhatter]

time for a quick update. I have been busy finishing off the main batteries and getting the very small amount of detail added to those. I'm happy with them and they are ready to go onto the ship and be plugged into the boards when they arrive:

Now, for the part that took me about 4 hours to do -the engine lights. ***WARNING - if you are an Electrical engineer/Electrician - then you may want to look away now as the next few images are not pretty.

But they work!!

They will be split into 3 groups of 4 and will operate in a circular motion as you will see soon. I won't lie - it was very challenging for me to get 12 LED's each with their own resistor into a circle only 2cm in diameter. But now that they're done, I can relax and move onto other things.

I have done other stuff but nothing worth taking pictures of just yet.

Thanks for looking

MH

[Madmonk]

Yep, as an elecrical engineer, that is ugly!

With all those bare wires just be careful of shorts. Next time you buy any resistors try looking for 1/8W instead of 1/4W, they can be used for single LEDs up to 5V, much smaller.

The Wave Gun sequencing is proving a bit of a pig as the tri-colour LED isn't a true white when all 3 are lit, my other solution of individual LEDs was even less successful so went back to the tri-colour and came up with something I think works.

Will put a video up later.

Cheers,

Warren

[Madhatter]

Yes, it's certainly not going to win any aesthetics awards for sure but I really struggled to get them all in together in such a small space. I ran them for about an hour last night to ensure they didn't short. Mind you, if you know of a better way to wire them all up I'm all ears The resistors came with the LEDs. I have a pack of the really small ones but I didn't know if they would be the right ones or not. These were requested from the specs you gave me for the boards. I probably could have done it differently by having small sections of wire off each individual LED and join them all up but I wasn't sure at the time if that would work properly. I might actually try that but I'll need to buy more LEDs first.

[Madmonk]

Got the video uploaded, this is back using the tri-colour LED (red, blue & green). Fading up the red, then the blue, off for 0.5 sec then all 3 colours on for 6 seconds for the white blast.

Cheers,

Warren

[Will Vale]

That looks cool but maybe a little slow? Some shimmer (e.g. a triangle or sine wave) composed over the run-up would keep the same timing and make it feel more alive.

e.g.

// Your function to provide LED brightness
byte led_level(unsigned long t);

// 16 bit multiply and normalise to 8 bit
byte mul_norm(byte a, byte b, byte shift)
{
	return ((word)a * ((word)b + 1)) >> shift;
}

// Invert top half of sawtooth wave to get a triangle
byte sawtooth_to_triangle(byte sawtooth)
{
	return (sawtooth > 127 ? 255-sawtooth : sawtooth) << 1;
}

loop()
{
    // Load time
    unsigned long t = millis();

    // Mask to get a sawtooth wave of period ~1/4s, and convert to triangle
    byte triangle = sawtooth_to_triangle(t & 255);

    // Optionally smooth here

    // Only modify top 25% of brightness
    byte scale = 192 + triangle >> 2;

    // Apply to smooth LED level you've already computed
    byte final_level = mul_norm(led_level(t), scale, 8);

    // Write PWM pin here
}

You can also use mul_norm to scale the different RGB LED outputs slightly in order to get a better temperature for white.

HTH,

Will

[Madmonk]

Hi Will,

I won't for one minute pretend I understand what you have suggested! This is the code I used, any suggestions to improve it will be gratefully accepted. It is in the set up area as I only want it to run once when power is applied.

The sin function was pinched from another piece of code I found.

// pin assignments
const int Green = 2;
const int Red = 0;   // PWM
const int Blue = 1;    // PWM




void setup ()
  {
  pinMode (Red, OUTPUT);
  pinMode (Blue, OUTPUT);
  pinMode (Green, OUTPUT);
  
  float in, out;
    for (in = 4.712; in < 7.854; in = in + 0.001)
  {
   out = sin(in) * 127.5 + 127.5;
   analogWrite(Red,out);
   delay(1);}
   
   for (in = 4.712; in < 7.854; in = in + 0.001)
  {
   out = sin(in) * 127.5 + 127.5;
   analogWrite(Blue,out);
   delay(1);}
   
   delay(500);
   
   digitalWrite(Red, LOW);
   digitalWrite(Blue, LOW);
   delay(100);
   digitalWrite(Green, HIGH);
   digitalWrite(Red, HIGH);
   digitalWrite(Blue, HIGH);
   delay(6000);
   digitalWrite(Green, LOW);
   digitalWrite(Red, LOW);
   digitalWrite(Blue, LOW);
  
  
  }  // end of setup
        


void loop ()
  {
    
   
  // other stuff here like testing switches
  }  // end of loop

I'm using an ATtiny85 chip so there are only 2 PWM pins available hence only fading the red and blue legs.

Cheers,

Warren

[Madhatter]

bloody hell Will - I so wish I understood ANY of that! I hope Warren does

BTW Warren - I think it looked really great! Will is right about the speed though - maybe a little quicker on the transition but not too much. The shimmering sounds quite cool too!

I re-did the engine rig and it looks a lot prettier now.

Isn't it great that the more times you do something the better you become at it. Instead of 4 hours total to make this one, it took me 2 from start to finish.

MH

[Madmonk]

That looks better Si, on mine I had the LEDs on the outside of the clear part, means a bit more carving up of the kit parts to fit it but there is a little more room!

The other altenative would have been to use SMD LEDs but not sure if you would get the same brightness as 3mm.

As for Will's hieroglyphics, not a clue!

Speeding up the transitions is no problem, that I can do!

Cheers,

Warren

[Madhatter]

I just wish I could do any of it! I like the 0.5 second gap between the Red blue green to white. Do you reckon you can get a shimmer effect like Will mentioned?

[Madmonk]

If Will can explain it a bit more I hope so!

[Will Vale]

OK so it looks like my 8-bit math is a bit too much and you'd be happier with floating point (that's the stuff that lets you use decimals.) That's probably OK since you're doing so little that even though the chip has to emulate the floating point it won't hurt you that much performance-wise.

So...

Try changing this:

float in, out;
for (in = 4.712; in < 7.854; in = in + 0.001)
{
   out = sin(in) * 127.5 + 127.5;
   analogWrite(Red,out);
   delay(1);
}

Which gives you this:

http://fooplot.com/plot/otfh3wiuoj

To this:

float in, out;
for (in = 4.712; in < 7.854; in = in + 0.001)
{
   float t = sin(in) * 0.85 + sin(in * 8 * 3.141) * 0.15;
   out = t * 127.5 + 127.5;
   analogWrite(Red,out);
   delay(1);
}

To get this:

http://fooplot.com/plot/2mcsd7xul5

The idea being to layer up different frequencies of sin(x) to get a more interesting waveform.

FWIW your use of sin for smoothing like that is pretty inefficient and makes it hard to work with the numbers because you have to use a specific range of inputs. You could try this instead:

// Cubic smoothing
float smoothstep(float t)
{
    return 3*t*t-2*t*t*t;
}

setup()
{
   for (float t = 0.0f; t < 1.0f; t += 0.01f)
   {
        analogWrite(RED, smoothstep(t) * 255);
        delay(1);
   }
}

Which gives:

http://fooplot.com/plot/fi3nkax3d2

I would also look at getting your timings from the system timer (millis or micros) rather than using delay() timing to get the result - it makes it easier to build multiple separate pieces and compose them together.

HTH,

Will

[Madmonk]

Thanks for that Will, I'll try it this evening and see what it produces.

My programming knowledge is quite limited as you can probably tell, enough to to do what I want with the lighting kits.

Will look into the millisecond function as there are a couple of kits I would like to flicker an engine while running a change machine to sequence lights but at present the delay function stops it from flickering.

Cheers,

Warren

[Madmonk]

Tried out Will's mods on the programme and it does look much better, thanks Will.

Intended posting the video to Youtube at work this morning but its blocked in this office for some reason so will post up this evening when I get home.

Cheers,

Warren

[Will Vale]

Yay! I did a helpful thing

You can fiddle the constants to adjust the mix of frequencies, and timings too.

W

[Madmonk]

This is the Will Vale modified version of the wave gun, the whole function lasts about 45 seconds.

Cheers,

Warren

[Madhatter]

Yeah that's really awesome! Is there anyway to intensify the white light at the end? I know, I'm very picky...