Gene’s Simulator Projects

(Read the prior blog entry to this for details on the projector stands)

After completing the projector stands it was time to finally use the damn things!

Because of the flexibility of modern games and the popularity of “triple head gaming”, new games like Need for Speed: Shift support the resolutions that can be found using a TripleHead2Go.  In my case that’s 3072×768 because I’m using three 1024×768 projectors.

My friend Dave is a nut for this game and I finally understand why. It’s insanely fun to play on a 180 degree wraparound display.

Here’s how it was set up:

The projectors are configured in a “cross fire” setup.  This means that the projector on the left is handling the arc on the right, and the right side projector is handing the arc on the left.

The projectors are set up to give a roughly 6″ overlap to each other in order for the screen to be blended properly.

I use a program called Nthusim that will allow me to distort and align the projected image in order to make it seamlessly conform to the curved screen.  It works VERY well.

In order to get Nthusim set up, you need to set it up in configure mode and start adjusting things:

Those white lines are where the alignment points are.  At the ends and intersections of each lines is a point that you can drag with the mouse in order to make the image conform to the screen.  Unfortunately this only works in programs that use DirectX and OpenGL, but not the 2D desktop.  Makes life interesting when you’re just working with the desktop.

The pictures are very blurry because I took them in a totally dark room with my DSLR.  Because I didn’t have a tripod handy, I had to hold as still as I could while the shutter was open – in those and some following images, the shutter open time was a full second or more!

Here’s three more pics that show Need For Speed: Shift in it’s widescreen gaming glory.  My friend Dave is at the controls. (he drives better than I do)

Here’s a short video showing it in action.  Unfortunately my Flip Video camera has VERY poor low-light performance.

Here’s what the car “sim” looks like:

Thanks for reading!

As of yesterday, (14Aug10) the new projector stands are finished and they turned out great!

I got this idea after the last setup I’d done last month – it turns out it works just as well as I’d hoped.

The projector stands were built from 3/4″ plywood and 2×4′s for the legs.  The support box is designed to hold a 3.5″ PVC pipe that acts as the actual support for the projector.  This allows the projector to rotate left and right very easily.  The height of the projector can be adjusted using a “step” pin that the PVC can rest on.  The series of holes you’ll see in the sides of the support box are for those pins.

Here’s what the interior of the projector stand looks like:

The column is right around 42″ inches high or so.  The box is built using “splines”.  Basically you cut your box sides at a 45 degree angle and then leaving the blade at 45 degrees, you flip the part around and cut a channel along the length of the 45 degree cut.  By doing this you can insert a “spline” or long thin length of material into the slot that will act as a reinforcement to the joint.  It results in a very strong box when you’re done.  I used 1/8″ hardboard strips as the spline material as it’s nearly the same thickness as the blade kerf (the width of the cut made by the blade) on my table saw’s current blade.

The inserts you see are there to support the PVC pipe.

The top of the box is capped with a 3/8″ thick sheet of Baltic Birch.  With a 45 degree chamfer on it, it looks very nice:

Mounted on top of the PVC pipe is a “yoke” that the projector itself will ride in.  The center line of the yoke matches the center line of the stand in order to maximize the stability of the setup.  The mounting points on the projector are located at it’s lengthwise balance point.

Here’s the three stands prior to mounting the projectors:

The yokes have an odd shape in order to ensure that even at an extreme down-angle, the VGA connector on the projector will still be easily accessible.

Here’s the projectors mounted up – this is as “low” as they’ll go.  The PVC is currently resting on the crossbeam that’s formed by the 2×4 legs.

That’s all for now.  Thanks for reading!

The 16 channel output board I’ve been working on is now finally completed!

This is a render of the final version of the board – this version has been sent off to a PCB house for manufacture (blank boards only):

This board will provide 16 channels of output, up to 500mA per channel at up to 30VDC.  The board costs about $20 in parts to make.  I’ll likely sell any spare blank boards I have for $10 each.  I don’t offer them as an assembled device or as a complete kit.

The link below is for a zip file that contains everything you need to build your own Gazoutta16 board, including the schematics, board layout and the firmware source code.  It’s based on the Arduino software stack (http://www.arduino.cc)

gazoutta16_project

If you have any questions, you’re welcome to contact me at the address given in the readme.txt file in the archive file listed above.

Thanks!

Short video:

The screen has a diameter of 15 feet with a usable circumference of about 30 feet.  Like its predecessor, the new screen is 78″ tall.

The glue still needs to dry before I do any finish work on it.  I’m going to either hit the local fabric store for some white muslin, or Home Depot for a couple of more drop cloths.  I haven’t decided yet.  Eventually, I’ll pony up for commercial grade screen material, but that will have to wait for funds to be available.

Here’s what the last panel for the projector looks like after being installed into the plywood frame.

Here’s a pic of the LCD controller and the power supply:

Having all three projectors done is a great step to get completed.

Here’s the “Stack”

I still need to get the supports for those built, but that will have to wait until this coming Saturday.

I got six of the screen segment frames assembled last weekend.  This sucker is a lot bigger in real life than it was in AutoCAD…

Each segment is about 63″ wide as measured straight across.  Each 2×4 leg is 95″ long.  It’s BIG.

I used some canvas drop cloth as a screen so I could check out a few things.  Here’s what that looked like:

The stupid plasma MM2 is either acting up on me or the “deal” I got on the pots I used on the controls is coming home to bite me.  I’m not sure which…

Until next time!

I _finally_ won the LCD lottery on Tuesday!
(My wife referred to it as me finding the golden LCD in the chocolate
bar...)

This means that I'll be able to finish off the third projector on
Saturday!  I'm really looking forward to that.  After I've got that
finished I'm going to start creating legs for the screen segments.  I
still need to decide how I'm going to build the feet - they need to have
adjusters on them because of the poor pour quality of my shop floor.
I'll probably go the simple route and use 1/4-20 bolts in to t-nuts for
that.

With the exception of being able to change the board ID code function,
the Gazoutta 16 firmware is completed and tested.  I got the sockets I
needed for the bus transceivers so I should be able to fully test the
daisy-chain feature as well.

The firmware takes up less than 3k of the available 30k space on the AVR
so there is plenty of room for making the board do more - especially
since I've included pins for an I2C interface on the board itself.
There's an I2C interface library for Arduino that makes that easy to use.

Some time ago I removed the propeller pitch lever from my throttle quadrant and re-skinned it.

Here’s the end result:

The skin is .080″ ABS plastic that I heat-formed over the quadrant frame using a heat gun and a straight edge.  A hair dryer could be used as well.

Well the last post was “Projectors! Projectors! Projectors!”  This one was nearly called, “Projectors! Projectors, Proj..wait, what?”

When I was first planning for the projectors, I did some research on what kind of LCD panel would be both affordable and suitable.    This led me to a web forum run by Lumen Lab – they at one time were a great resource for parts on building a DIY projector – they’re now just a forum with a store that sells robot parts.  Go figure.  Anyway, the research I did on the forum led me to something I wasn’t aware of.  “FFC problems”.

It turns out that FFC stands for Flat, Flexible, Cable.  It turns out that many LCD manufacturers have a small (and I mean _SMALL_), thin cable that joins the circuit board along the top to a circuit board along the side.  This is fine when it’s being used for its intended purpose.  The problem is that in order to use the LCD in a projector, you’ve got to move both of those circuit boards out of the way so they don’t block the incoming light or the outgoing image.

This information led me to a forum thread that listed a large number of LCD displays that were known to not have this FFC problem.  (Thread is here:http://lumenlab.com/forums/index.php?showtopic=4)

Among the listed displays was the Dell 152FPC.  I figured since Dell is a pretty good brand of display I did an eBay search and scored four of them from the same seller.

Fast forward a few months – I get all the projector cases built and wired up.  The panels stripped down easily and I got them in place.  Upon cracking open the third LCD panel, I was met with something I didn’t expect.  It appears that Dell had changed their panel vendor, but hadn’t changed the model # of the display.  The new vendor used a panel that had the FFC cable linking the top & side boards.  I didn’t worry too much until I put that panel back together and cracked open the fourth.  You guessed it, same panel, same FFC issue.

After inventing some new curses to hurl at Dell, I went digging through the Lumen Labs list to find another suitable panel.  This time I picked the Gateway FPD1530 after not finding any Dell displays.  I found a vendor on eBay that had them and I grabbed three.  Can you guess what happened?  Yep.  All three had FFC issues.  Next up was the Mag Innovision LT565.  Yep, Gene loses at the LCD lottery again.  So now I’ve got six LCD panels ( well really only 5 – the Mag panel arrived inoperable from the seller – he’s unresponsive and getting a negative on his feedback list) that I really have no use for and an lonely projector case in need of an LCD.  After I recharge my LCD fund by selling off the un-needed panels, I’ll continue to play the LCD game until I get one I can use.

Here’s a couple of pictures that will illustrate the problem.  First up is what the FFC link looks like:

As you can see, the cable is VERY short.  It’s soldered to the side PCB so it’s not a matter of just finding a longer cable and replacing it.  The cable shown is a .5mm pitch, 20 pin cable.  This means that it fits 20 contacts into a 10mm wide cable.  There IS a solution to this – unfortunately the vendor is in Germany and the cable extenders they sell don’t match any of the cables in the displays I have.  Just figures, dunnit? *laughs*  (I have to laugh, otherwise I’d start screaming and throwing things.  Which I’ve been known to do….*ahem*)

Here’s what a non-FFC panel looks like:

As you can see, there’s no link between the two.  I have no idea how it works that way, but it’s the kind of magic I really like.  It just works.

Ok, that’s my sob story and I’m sticking to it.  On to more successful things….

One of the projects I’m working on is based around the stick gimbal that’s found in Mike Powell’s book, Building Recreational Flight Simulators (found here: http://www.mikesflightdeck.com).  The free chapter he offers covers this gimbal mechanism.

The gimbal base is made from a number of 3/8″ plywood and 3/4″ Pine parts.  Here’s how I cut the Pine parts:

The little handles are cam clamps made by a nice gent named David.  His website is called Welcome to The Sandbox.  These clamps use aluminum inserts that are installed just a bit off-center in order to work as a cam when the metal part is bolted tight to the table.  As you rotate it to the right, it applies more pressure to the surface you’re clamping against.  It’s perfect for the kind of job you see above.

Here are all the parts, ready for assembly.  I’ll get back to those…some day. *sigh*  (Yes, there are two sets of parts there…)

Part of the projector project was building a dedicated computer that could drive all those displays with a Matrox TripleHead2Go box.  My friend Dave spec’d out the machine and we went in on the build cost.  He wants to play Need for Speed:Shift on the projectors.  I can’t blame him.

Anyway, Dave got all the parts at Fry’s and brought them over one Saturday.  I was working on other things and didn’t pay a whole lot of attention until he brought out the heat sink.  Now I’ve built my fair share of machines but I’ve NEVER seen a heat sink like this.  It’s ABSURD.  See:

Never in my life have I seen a thing like this.  He tells me it’s a good idea to have if we over clock the machine.  I’m convinced he’s got Plutonium wedged under there and that heat sink is the only thing keeping us from a bright, hot death….  (I’d never seen a 1KW power supply either.  Holy crap. You can arc weld with the 5V rail in that thing…)

It’s safe to say that this machine is perfectly capable of getting the hell out of its own way.  It’s an Intel i7/960 @ 2.8GHZ with 8GB of 2GHZ DDR3 RAM in it.  This thing goes like a scalded cat.

I finally took the time to build the wiring harness that connects the left sub-panel to the Plasma-MM2.  I don’t care at all how MSFS deals with it, so I think I’m going to write a wedge application that feeds everything to MSFS and just tell it that there are no joysticks installed.

It’s pretty cool to be able to mash the Starter button and hear the engine crank & start though….or just crank if I have the mixture leaned all the way out.

Here’s some pictures of the finished projector – one of them anyway:

The above shows how the LCD is installed.  The fresnel lens you see is the output lens – the LCD is right behind it.  The fresnel is mounted on a 1/4″ dowel and it hangs freely in a pair of guides.  This allows some limited automatic keystone correction as the projector is tilted.

Here’s a picture of the “hot” end:

One lesson I learned VERY quickly – once warmed up, Mr. Metal Halide Lamp is NO LONGER YOUR FRIEND.  I need to make a sticker for the side that reads, “Do not look into lamp with remaining eye.”  It’s less painful to stare into the Sun.  Trust me, I checked.  It’s fun living in a purple-spotted world….really.

The gadget attached to the lamp is a concave mirror – its job is to aim the light coming out of the back “side” of the lamp forward.  This increases the light output by a noticeable amount.

The business end:

The output lens is  330mm “triplet”.  There are three lenses in the assembly.  The fresnel focuses the output from the LCD into this lens.  The way I have it set up right now gives me a 122″ focal length – this means that to be in focus, the projector needs to be 122″ away from the screen.  You have no idea what this did to my screen design. *facepalm*

The last two pics show the projector from the side, both open and closed.

The projector case is 30″ long and 17″ wide.  This is going to make for a challenging mount – the projectors need to be positioned right over my head.  This will result in two different mounts – one for the F-15 and one for the smaller cockpits.  Because of the focal length of the projectors and the amount of space each one takes up, my screen radius diameter is going to be 24 feet.

Yes, you read right.  Twenty-four FEET.  Holy crap…what on earth have I gotten myself into?  The screen size is such that it’ll work out great for the F-15, but it’s going to dwarf the poor 109F/X…  The screen for the moment is going to cover 180 degrees of arc and will be built from 8 curved sections.  I’ve got 16 of the 24 curved parts already cut for that.  I should have the rest done this coming Saturday.  I’ll write more about it when I have a few assembled so you can point & laugh at the fix I’ve gotten myself into.  The screen will eventually cover 225 degrees in order to allow the three projectors to show 75 degrees of view each.

Some time ago I started working on a little project – I’ve become totally disgusted with the state of cockpit interface electronics. Ever since Beta Innovations gave up the ghost, pickings have been damn slim.  What is left is either too complex, WAY too DIY or too expensive.  That annoyance gave rise to this:

This rats nest hides an ATMega328 microcontroller and a pair of ULN2803 darlington transistor arrays.  This gives me 16 channels of output that can handle up to 500mA per channel at up to 30VDC each.

The device was christened Gazoutta16.  (yes, there’s gonna be a Gazinta)  It’s based around what I learned by examining the other output boards I have as well as the schematic for a similar (PIC based) device in Building Recreational Flight Simulators.  This board is _stupid_ simple.  It uses a dirt common ATMega328 uC from Atmel for the brains – the software is written using the Arduino IDE.  After some fiddling around with the bread board above, I had four boards made:

The board communicates with the outside world using an FTDI Basic Breakout board that I bought from SparkFun for about $14.  This provides a generic USB serial port that will talk directly to the AVR at 250,000 baud.  Because of how I built it, you only need one of these to talk to a theoretically unlimited number of the Gazoutta16 boards.  They can be daisy-chained to one another.  The chip missing on the prototype is for a buffer chip that makes this work out.  This makes it handy for my F-15 project because just between the Master Caution panel and the BIT panel, 74 output channels are needed.

This board can also drive a number (9 I think) 7 segment displays just by uploading different firmware to it.  The board uses the same basic circuit as an Arduino Duemilanove board so you can update the firmware on it via the Arduino IDE instead of needing a special AVR programmer like the STK500 or similar.

I still haven’t decided whether or not I’m going to sell these things or not.  If I do, the firmware source code will be released under an open source license so folks can tweak it or make it better as they see fit.

I’m done for now – hopefully it won’t be so long until the next update.  Thanks for reading!

…but before that…

I replaced the Plasma V2 with a Plasma MM2 that I had.  Turns out that the P2 was the core of all my controllability problems.  Once I installed the MM2, it was a whole ‘nuther ballgame.  No more crazy spikes or sluggish control inputs.  It was pretty damn nice actually.

Here’s the board I put together in order to interface the MM2:

What kills me is that this device is no longer available.

Another change I made was to the return springs that were in the pitch axis mechanism.  The original springs weren’t up to the task with the amount of weight in the grip & interface box.  The stick would stay where ever I let go of it.  The springs I use are compression springs that are 2″ long and made from .072″ diameter wire.  I changed to a 2.50″ spring made of .080 wire.  This made a big difference for a small change.  The pitch axis now returns to center the way it should.  However, the feel is still a bit weak so I may change that to an even stiffer spring.

Ok, now back to the title of this entry…

I ordered three projector kits on the 31st of December.  After an agonizing wait they were finally shipped.  Here’s some of the parts:

The kit consists of optics (220mm & 330mm fresnel lenses and a 330mm “triplet” output lens), power panel, 400W Metal-Halide lamp (that the 12 year old in me has christened The Photon Phallus. *snickers*), Mogul base for the lamp, temperature controller for the fan, a fan and a 400W ballast for the lamp.  All of this neat stuff is available from DIY Projector Kits (http://www.diyprojectorkits.com).

Be warned, the plans (at least for the one I have, the S-15) are VERY, VERY, VERY vague.  If you get one and have questions, either see my build thread on the site (“geneb’s triple threat”) or email me directly.  My build thread will do what it can to rectify the captionless pictures that comprise some of the documentation.

On the 23rd, I spent the day doing maintenance work on the ShopBot, including adding a 1/32″ gasket to all the vacuum zones.  The gasket drastically improved the performance of the vacuum hold down system.  I then got all the parts cut for the projector cases.  I used 18mm (3/4″) Baltic Birch.

Here’s the first of two sheets that were cut:

The parts in the foreground with the long slot in the center is the projector top.  The slot is the air intake for the lamp cooling fan.  It draws air in-between the “cassette” that holds the 15.4″ LCD panel and the cassette that holds the collimating fresnel.  You’ll see pictures of these things in the coming days.

Here’s the stack of parts that resulted from yesterday’s run:

The next steps are to mark & drill the case parts for the pocket screws I’m going to use for assembly and make the LCD and fresnel cassettes….

To hold you over till next time, here’s a video I shot of the beta version of the avionics system I’ll be using in the cockpit:

The past couple of weeks has brought some good progress to the simulator hardware.  It’s still not flyable yet, but it’s a lot closer than it was in December!

The main wiring harness for the pitch & roll axes gimbal and the stick wiring was installed:

I used cable clamps to attach the harness to the underside of the cockpit.  To separate the side from the center, you only need to unplug the harness from the gimbal assembly and remove one cable clamp.

It turned out nicely and while I should have combed out the wires (makes a better harness) it’s great for the prototype wiring.

(continue reading…)