Land Raider

More H Bridge work

2006-03-20T10:37:00.000-08:00

March 20th '06

Today I made the double NOT gate circuit for the motor controller.

For the first time I tried one of those blank predrilled PCB's from radio shack. What a total pain in the butt. It looked nice though and was less expensive and more sturdy than the acid etching stuff I've done before.

Here are the plans and the finished board.

Here it hooked up. It seemed to work for a bit but something went horrible wrong and fried the L298 H bridge chip. *cry* What a huge waste of time, I ordered a handful of NOT gate IC's that afternoon, along with a replacement L298 chip. The not gates are about 60cents a piece for a 8 gate chip, so why fight for hours building my own?

Thank goodness for the suction iron....

Per a request here are the notes I made for the H bridge and NOT gate stuff.
http://www.pasqualy.com/landraider/stepperplans.jpg

There may be errors on that page so make sure you double check anything I wrote down before you use it yourself.

Replacing the RDFR with a dual H bridge

2006-03-17T10:35:00.000-08:00

March 17th '06

Alyssa and I worked on the robot today. I'm taking out the underused and overpriced RDFR speed controller and replacing it with a simple dual H bridge.

Alyssa says, "That's a price savings of about 300 bucks!"

The unit will mount in the front bay like the old one did, but forward of the motors.

Its is much smaller as you can see.

"Dad! Those motors draw 980 milliAmps each! You should add a heat sink!"

Right, here is the board with heat sink mounted on the L298 chip.

Fortunatly the motors have little fans on the back that kick out air in the direction of where the H bridge will be mounted, and there is an air vent just underneath of where the heatsink will stand.

Here it is mounted.

Here is the SV203 board driving the inputs on the H bridge, Alyssa was pretty close with her current draw estimate. The L298 can handle 2 Amps per motor.

Now here comes the tricky part.
The L298 has 6 inputs.
Left motor Forward
Left motor Backward
Left motor Disable
Right motor Forward
Right motor Backward
Right motor Disable

My SV203 controller only has 8 inputs total and I need 2 of them for the gun's pan tilt and another one for the trigger device on the gun. Plus I'd like to have another two for a second camera's pan tilt.

So I had to use some electronics wizardry(joke) to minimize the number of outputs used up by the H bridge (the RDFR used only two outputs Left/Right channel and Forward/Back channel all with automatic mixing).
So what I did was get it down to three outputs like this:

Left Motor Direction
Right Motor Direction
Left and Right Disable

For the left/right motor direction I simply used one line going to forward, and then did a logic inversion of that line same going to backward. So if forward is on, backward is off, if forward is off then backward is on.
Then I connected both sides disable to a single line, so either both are on, or both are off.

You can see me testing this with a NOT gate (or inverter) built from a resistor/transistor circuit built on the bread board. It worked great so all I have to do is build a board for it and stick on the robot.

Replacement motherboard is in

2005-11-22T13:53:00.000-08:00

The replacement motherboard came in for my tank today.

I added lots of switches, buttons and LED's while I waited for it over the past few weeks.

Thank goodness it all worked!

The red and yellow bottom LED's are for motherboard power LED and HDD LED.
The big red switch at the top right is for master power.
Below it is a green LED and switch for motor control power.
Below that is an amber LED and switch for servo controller power.

To the right is a power(red) and reset (black) buttons for the motherboard, and between them is a 12V DC jack for when I dont want to run off the battery.

Replacement drive wheel

2005-11-13T01:33:00.000-08:00

21st Century toys was kind enough to send me a free replacement drive wheel for tank. The original was cracked and tended to slip off of the brass drive nut and then the nut would just spin in place inside the plastic housing. The corners had worn away so bad that it was becoming a big problem.

You can see the small crack here.

This is the brass nut that drives the drive sprocket. The drive sprocket is held on only by the tension of the plastic around the shaft and the tendancy of the tracks to keep it in place.

Installing switches, buttons, and indicator lights.

2005-10-25T10:39:00.000-07:00

While I was waiting for a replacement motherboard, I decided to work on adding some switches and what not to the back of the tank.

First off I was cleaning up my wiring and making things nice and neat when I started to consider where I was gonna put all the switches etc. I realized that my mounting board was going to have to be modified or else I could never fit all the switches in, and even if I did, I could never get the board out without pulling out every LED and switch I installed.

So I had to yank out the mounting board and modify it.

I removed a notch that would end up being the place for the mobo LED's and removed the two 'wings' that filled the gap toward the back of the tank. I had to take those off because they would have sat under the switches and stuff and made it impossible to remove the board without first removing the switches.

Let me tell you, this was a HUGE pain in the ***. I wish I had accounted for it before, because I had to take all kinds of crap apart to get it back out and then modify it.

Now that I had that taken care of, on to the switches etc.

Master Power Switch
When it was off, the power line to the batteries and 12V jack was broken, however it did not break the connection of the 12V jack and batteries, so the batteries would still charge off the jack when the main power switch was off.
When this switch is off nothing has power.

Motor Controller Power Switch
This broke the power to the motor controller in the front of the tank. This would allow me to work on the tank and not have to worry about the thing running away on me.

Servo Controller Power Switch
This one is a little tricky. The Servo controller doesn't run off of 12V. I't is a 5-6V system. Instead of adding its own DC-DC voltage regulator, I just ran it off of the power supply that runs the motherboard. If you don't know it, PC power supplys give 5V(red lines) and 12V(yellow lines).
So the switch is simple in that it just breaks power to the servo controller from the motherboard power supply, but tricky in that the motherboard supply has to be on to give power to the servo controller. Not that big of a deal really.

Here is an image of the switches installed.

Next, I needed to install indicator lights for the Motor Controller, and Servo Controller switches. The Master Power Switch has its own built in light.

The two LED's had to be connected correctly with the switch and then resistors added. Unlike lamps, LED's don't have built in resistors, and you must be mindful of polarity.

I put the resistors inline with the wire just behind the LED.

Then I covered them up with some heatshrink.

Next I drilled holes and mounted some LED holders. The plastic was way to thick for them to catch right so I had to modify them and glue them in place.

Here they are mounted etc.

Now as mentioned before, the Servo Controller has no power unless the motherboard is on. Since I had no motherboard, how did I power up the power supply to test the light?
Like this..

To power an atx power supply simply short the green wire on the motherboard connector to ground.

Motherboard Power Button
This is just a momentary switch that turns on the motherboard. It connects to the built in pins of the motherboard itself.

Motherboard Reset Button
Same as above, just a momentary switch that connects to the reset pins of the motherboard.

These two switches were very straight forward, just drilled the hole and mounted them. Again, I yanked the wires from old computer hookups and wired them to the switches I got at Radio Shack. The green button will be replaced with a red one.

Power on and HDD Activity
These were just straight forward LED's that I actually yanked from an old computer someplace (I have a huge box full). They connected to the pins made for them on the motherboard. I didn't have to add resistors here because that was taken care of on the motherboard.

Here is an image from the inside view where LED sockets mounted. This is what the gap in the mounting board was made for.

Here they are mounted inside and outside views.

C1 = You are screwed.

2005-10-15T12:23:00.000-07:00

FYI a POST code for Award BIOS of "C1" means that there is a memory error.

So if you know your memory is good and that its seated correctly, then you know your board is bad.

:(

Thankfully the guys at mini-Box (buy stuff from them!) answered my email prompty with instructions on how to get a return. They answered email on a saturday too!

I was sooo looking forward to having this thing up and running this weekend.

Internal tank mods -or- I am a dremel Ninja

2005-10-15T09:05:00.000-07:00

First thing I did was remove the fan and heatsink from the RDFR speed controller. It really wasnt needed now since the motors were so small and it was big, noisy, and power hungry.

Next thing I had to do was remove the old battery compartment that was attached to the upper hull. It normally would have rested right on top of where I was planning on mounting the motherboard. Since I wasn't going to use it for anything, I just lopped it off.

I did make a small mistake here. The upper hull gained considerable strength from the battery compartment. I cut too close to the top of the hull and removed all of that strength. I should have left about an inch or two below the top of the hull. The cuts are highlited in red to make it easier to see.

Next I measured and cut the rear compartment riser to fit the motherboard. There was a ledge at the very rear of the inside of the hull that I wanted the moatherboard's mount to rest on. I planned on using a piece of plywood to rest across these parts so I would have something for the motherboard to mount to. The ledge is highlighted with yellow and the cut is highlighted in red.

Here is the first piece of plywood that I cut for the board mount. It worked ok untill I realized I had made it too small from front to back to fit the motherboard. I had to toss it, but I realized that I would be much better off making a single plywood piece that covered the entire inside of the hull. I would cut out spaces for the batteries etc.

This time I measured things a little more carefully.

It fit quite well on the first cut, I had to do some sanding on the sides where the batteries went so that they fit a little better. Not to mention tons of dremel work to get the bottom of the bay smooth. It had pegs etc from the old electronic's mounts.

Here is what it looks like with the batteries installed. Perfect fit. The four dots you see on the plywood are the locations for the motherboard risers.

At this point I did a ton of measuring and marking and cut out the slot for the motherboards rear panel. Let me tell you, this was a huge pain in the ass. I had to measure it and get it right to about 1/32 of an inch. To top it off I had no good way of cutting the stuff. You just can't get a good straight edge with the dremel, and my jig saw couldnt fit back there. I ended up taking my jig saw's foot off and doing most of it with that, but I cut my hole small and used a file to widen it as needed.

All the hard work paid off though. It fit pretty well in the end.

View from the back. Awesome.

Here it is with the board mounted and the motherboard risers mounted on the board. The moment of truth approaches, will it fit?

Thankfully it fit just right. I wasn't kidding about the dremel ninja skills.

Case modders, eat your heart out!

Here is a shot of all it fitting together inside the bay. The markings you see between the battery are where I screwed a wooden block under the plywood board to bridge the gap between the plywood and the bottom of the hull.

Next I mounted the RDFR in the front compartment. After I spent time making it look nice and pretty, I realized I needed to put it behind the motors or else the control wires wouldn't reach the SV203. Also, with that front part open, it will make a nice little spot for a camera or grabber arm or something.

Meanwhile Alyssa was being helpful by fixing an old mobile phone she found in my bits box.

I also added a power jack in the back to make recharging the batteries easy without having to remove the cover etc. I just need to find a 12V power adapter that can push enough juice to power the thing and charge the batteries at the same time.

Here is the final layout, I just need to clean up the wiring.

(click image for full size version)

Parts came in today.

2005-10-10T07:03:00.000-07:00

Wooo!!!

Once again, big thanks to mini-Box for the great service and discounts.

Here is a good image to give you an idea of the scale of things. The ITX board is 6.5 by 6.5 inches. It should fit just about anywhere inside the tank bays. The problem with the tank is that the inside is sectioned off by inner walls. This gives the body some stength but gets in my way.

My fist step was to get the board up and running and do some tests to see how much current she pulls. I found an old 10Gig laptop hard drive, a floppy, CD-Rom, and a copy of good ole Windows 98. I figured there was no good reason to stress the little guy out with a copy of WindowsXP, especially since it only has 128M of Ram right now.

My first problem was that it drew way more current than I was hoping for. I must confess that I didn't read the freaking specs like I should have and I was expecting about a 1Amp draw from the board. Once I strip it down to the board and the Hard Drive only, its pulling about 2.1Amps. Eventually I'll replace that Laptop HD with a Compact Flash card and that should save me about .5 Amps of draw but its still a hog. I wish that I had gotten the MII6000 version that is fanless and draws much less power.
I took a look at some specs later and noted that the MII10000 draws about 15Watts when idle, and the MII6000 draws only 11Watts. Maybe the extra processing power will be worth it, we'll see.

In light of this new info, I realized that one little 12V 7Ah battery wouldn't be enough. Not only would it empty itself too fast, it would probably run hot. Luckily I already had 2 batteries of the same type so I started figuring how I would fit all this crap inside the hull of the tank.

Here is the clean tank hull, the motors and gearboxes are the only internal devices from the orginal tank.

Here is how I plan on squeezing it all in there. The motherboard is in the rear with the faceplate mounted in the back of the tank. Batteries go in the large center bay of the hull on either side, this leaves me with a gap between them for the hard drive or the SV203 or whatever else. The RDFR speed controller goes up front next to the motors it runs.

(click for larger image)

Mobile computing to the extreme

2005-10-03T14:22:00.000-07:00

I got the whole thing mobile this morning using my old crappy laptop. The Epia board will basically replace the laptop when it comes in and hopefully be much easier to work with.

The control page is at http://www.pasqualy.com/landraider/temp/java1.html but it probably will not be on if you go look, I cant risk someone running around smashing into everything at my house etc. I really need a way to let some people watch, and let some people control, and have a que up system for those who can control.

First thing I did was mount the pan/tilt camera, I didn't use the guncam because I couldnt fit it and the laptop on the tank at the same time, the mini-ITX board should solve that problem. Notice the RDFR speed controller under the camera. (the tanks original controlling hardware was ripped out a few days ago)

Next I wired up the SV203 to the laptop and its 6V battery, the RDFR speed controller uses 12V, so eventually I'm gonna just use a 12V batt and convert it to whatever else I need.

Here is the final result, and a screen of the web interface.

More to come, hopefully I can get some help with the web interface so that it wont suck so bad.
As soon as the mini_ITX board gets here, I'll give out user/pass access and leave it running all day long.

- Cy

LandRaider motherboard purchased

2005-10-03T09:03:00.000-07:00

I just went on a shopping spree at mini-box.com

Thier prices were excellent, and the response to my questions were super fast.

I purchased a VIA EPIA MII 10000 a 200W DC to DC converter to power it and some other goodies. I got a Compact Flash to IDE converter that will be a sweet power saver, but I don't have a CF card big enough to hold an OS yet, that will come later.
I was going to get one of the TC versions of the Epia boards that come with a built in DC-DC converter but they were under featured it seemed.

This board has a built in CF and PC card slot, and everything else you would expect to see in a full featured motherboard (USB, Vid, LAN, Sound, FireWire, etc) it only has one Ram slot and one PCI expansion slot, but I probably wont even need that. I'll use the PC card slot for the 802.11 card.

It also has TV out and LCD out options that I might use one day, but in general I wont need video unless I'm working on the board outside the LandRaider.

My wife will kill me if she finds out I just bought all this stuff.....

Control/Data flow scheme

2005-10-03T02:33:00.000-07:00

I could do this a zillion ways, but this is how I've settled on it.

Basically, users connect over the net to my house. My router forwards the correct traffic over to the land raider via WiFi. The land raider meanwhile is sending data (mainly images) back the same way.
What I still need to decide on is if I want to do this with some browser based method (like what I have currently setup with Serial Host) or if I want to write my own application that runs on the remote computer. Maybe I can find an easy way to do both.

Getting the images/video from point to A to B is going to be the hardest thing, right now I'm just using some crap webcam software that updates every 1/2 second.

Serial Host up and running

2005-10-03T00:02:00.000-07:00

Worked on getting serialhost up and running.

current page is
http://www.pasqualy.com/landraider/temp/java2.html

It will probably be off most of the time to avoid having people screw things up in my office etc.

I had the change the ip address, and strip most of the old crap from the buttons. The left/right turret movement is now via a hacked servo controller wired into the original turret motor. This means I can't give it absolute positions anymore, only relative movements.
I could get absolute values by attaching the servo's pot to the turrent somehow, but that would leave me with only 255 movement positions and over a 300^ arc, that kinda sucks.

Here is what the setup looks like right now, its all very temorary.

Here is a closeup of the workings. You can click the image for a larger version.

New Blog

2005-10-02T20:25:00.000-07:00

This blog will be for landraider updates etc.
My personal Blog is at www.pasqualy.com/blog/

Any posts that show up under this one (previous date), are just reposts from the old webpage that I should be taking down/cleaning up soon.

www.pasqualy.com/landraider/

- Cy

Project Update 05-31-04

2004-05-31T02:51:00.000-07:00

Here is an update after weeks of slacking.
Sorry for the delay.

Here is the video version:
http://www.pasqualy.com/landraider/...31-04update.wmv

Last time I posted any kind of update I think I was installing a cooling fan for the speed controller. Well the fan worked great but I managed to get some wires crossed when putting the controller back on the bot and fried it. Thankfully the guys at vantec had mercy on me and fixed it for $70.

Chain Drive v2.0
I also have gotten a second 12V battery for the bot and at 24V total with the two batts. it has some pretty awesome power. However more power just showed how terrable my chain drive was. So I set about making some changes in the way things worked. I came up with what basically amounts to a double sprocket system. I don't know why I just didn't do it this way from the get go. It has several advantages. Gearing up/down doesnt require that I replace all 4 sprockets for each spindle. Just the ones on either the drive motor or drive spindle or both if needed. Chain tension is not nearly the problem it used to be either.

Here is a concept image of the new design. Notice the Motor sprocket drives the 1st sprocket on the spindle. The second sprocket on the spindle in turn drives the sprocket on the other spindle for that side of the vehicle. So its still 4 wheel drive.

Because there had to be two sets of chains side by side, I had to move the motors further in. This required a bit of wood work to make room for them.

You can get an idea of how things are going to work from this picture. The motor sprocket is 10Teeth. It drives a 24 tooth sprocket on the spindle. Thus its geared down 2.4:1. There is a second 16T sprocket on the spindle that drives a 16T sprocket on the other spindle for that side.

Now because my town is small and crappy, I couldn't find any keyed shaft. So I had to dremel my own keyway. It was annoying but setscrews alone just wont cut it for the amount of power we are dealing with.

If you don't know, keyway is like a channel running along the length of a rod. The sprocket also has a Notch of the same size inside of its hole. These notches are half as deep as they are wide so you can fit a little piece of keystock inside them and it fits halfway in the shaft and halfway in the sprocket. Then you tighten down on this with a set screw and it assures that your sprocket holds fast to the rod. I didn't bother to keyway the hole shaft, just the places where I needed it. Here is a closeup of the sprocket, keyway, keystock, and setscrew hole.

These two images show the final layout of the new chain drive. You can see that I also removed the old PVC roller on the tension bar and replaced with a small sprocket. This will also change with the final version. Toejam sent me some sweet tension products from his company that I plan to make use of, thanks!

Here she is in all her glory, with the speed controller and two batteries. I gave here a test run at this point and let me tell you she kicks ASS. Un freeking stopable. Compared to the previous version she now has almost 5 times as much torque. You can drive it up to a wall and it will just crawl up the wall till it flips. I can STAND on the forward edge and it will still crawl up the wall with me on it!!!!! Check out the pimping lighted power switch too.

Adding a Top Deck and Device Mounts
As stated some time ago, I had already realized that I didn't have enough real estate on the top deck to put everything I needed. So I decided to add a second level. I figured the camera would do well to hang in between the bottom floor and the top deck . This keeps it from getting smashed to pieces.

First I put down some 2x4 blocks for the top deck to sit on.

Then I slapped on the top deck.

Next I cut a hole out for the speed controller fan to pull air through. I also mounted the gun base.

Here is the view from the under side.

Next I started working on the camera mount. At this point my fabrication skills have grown by leaps and bounds and I have much better tools backing me up. So the mount came out much better than the old way of just sticking two servos tegether with tape and ticky tacky!
I used old pieces of sheet metal from broken power supplies and old computer cases. Thats why you see "BAD" marked on some parts. Its how I know a part has been tested and isn't any good.

I'll spare you guys the details and just give you the final product.

Here it is mounted under the top deck. It turned out pretty sweet.

Video:
http://www.pasqualy.com/landraider/...31-04update.wmv

Project Update 04-06-04

2004-04-06T02:55:00.000-07:00

Speed Controller Modding.

I got another 12V bat. from Surplus Center to juice the bot up to 24V. I started getting worried about burning up the speed controller at this point. I did a little math and realized that I was fine unless the robot stalled itself at full power somehow for for than 5-10 seconds or so. In case you didn't know, electric motors draw more and more power as they meet more resistance to movement. If you lock down a motor so that it cannot move it will draw its internal electrical resistance is at its lowest and thus its current draw is at its highest.
To help get rid of some of the heat that the speed controller might build up during lots of heavy use, I decided to add a heat sink and fan. It is probably not really needed but I was bored and it just looks cool :)

Another thing that prompted me to do this was that I noticed the controller was in pretty bad shape. Its my brothers and its expensive so I dont want to screw it up. Its rated for 30Amps cont. or 60A for 5 seconds. Right now my motors are pulling about 6 or so at no load but when you lock them down they probably pull about 24Amps! At 24 Volts that would be 48 Amps if locked down. Well under the 60 max but well above the 30Amp cont. level.

I took a shot of the crap leaking out of the capacitors in the speed controller. There was even a spider nest in there.
(Note: Later on I learned that this isn't leaking caps. Vantec said it was a glue they use. *shrug*)

I hooked up the Land Raider and played with it a bit and felt the heat on the controller box, it wasn't much and the metal box seemed to be doing pretty well at sinking it. However I figured I'd stick my heatsink/fan on there anyway because it would look cool and better safe than sorry etc. The power mosfets were all resting against a heat pad that was stuck on the top of the metal enclosure. I put some heat sink grease on them and it should make a big difference because not all of them were touching the pad very well.

I found an old processor heatsink that fit pretty well. And by coincidence its made by Vantec, the same name but different company that makes the speed controller. I polished up the case top and sink bottom with some sand paper and drilled holes in the sink to match the controller box.

I greased the heck out of this as well and then mounted the sink.

Then after finding new screws that were long enough I bolted it down and put the fans back on. At 12V they draw about .2 amps, I think I'm gonna either just run them off the 6V battery that I'll be using for the servo controller or at least put a resistor to slow them down a little because they really are overkill. Just the sink and grease alone probably made a huge difference.

Project Update 03-28-04

2004-03-28T02:56:00.000-08:00

Goodies in the mail!

Got this stuff in the mail from surpluscenter.com. With these and other recently procured goodies, I started working on the wooden chassis mockup.

Wooden Mockup Construction

If you want to save some time I suggest you just watch the vids. They pretty much cover everything here plus some more.

http://www.cp-tel.net/pasqualy/land...es/DTtest01.wmv ~7 megs

http://www.cp-tel.net/pasqualy/land...es/DTtest02.wmv ~4 megs

First off I had to do a ton of work on the wheels, spindles, rims, and hubs. The original hubs were designed to spin on the shaft. Im my case they needed to locked to the shaft. To do this, I drilled a hole though the hub and shaft and pinned it with a bolt.

Now that I had done this, the rims had to be modified so that the hubs would fit though with the bolt. Trying to stick the bolt in once the hub was in its socket was just out o f the question. This was a huge pain in the rear.

Here you can see the beginning stages of the body and spindles. Getting all this stuff to line up was pretty hard but I'm pretty satisfied with the way the work came out. I had to do very little adjusting.

Here you can get an idea of how things are gonna work with the motors and chain.

I mounted down the motors between wooden blocks to keep them from shifting.

This white bar you see is PVC pipe slipped over a 1/2in rod. It was supposed to provide chain tension but it didn't work so well.
One side ate into the PVC and the other side was a little too slack.

You can see how the chain winds its way between the sprockets etc. It goes over the motor sprocket and then under the PVC tension bar. I found out later that this doesnt make it bite nearly enough teeth for what I'm doing. But more on that later.

Here are some pics of the final deal. I aired up the tires and found I had two flats. I patched one and had to purchase a tube for another.

Here is a good view of how things are working in there.

Here are two pics to help you understand the size of this thing. Its freeking huge!!!
Remember, those are 14 inch tires!

I placed some of the other possible devices on the top deck just to get an idea of how much space I had. I realized that I just didn't have room for everything. So I decided to build a second deck. That would come later though. Looks pretty neat with all these goodies on it!

Here is the aftermath in my store. No worries though, it cleaned up ok. :)

Problems Problems...
When it was all said and done it worked pretty well but for two major problems.
1. Chain popping.
The chain just wasn't gripping enough on the motor sprockets and the tension bar sucked. This will definetly have to have a major change done to it.
2. Skid turning requires lots of torque!
When you think about it, it makes sense, skid steering requires that your tires move off center to the direction that they want to turn. Rolling a wheel is easy. Pushing one sideways is not. The wider the wheelbase (left to right) the easier the turn because there off angle of movement is smaller. The Shorter the wheelbase (front to back) the easier the turn as well. I have about 5 inches or so in between wheel edges (front to back) that I can lose to help reduce torque needs but this won't be enough.
I can't go any wider either, she is just slim enough to fit though a doorway. So I have to add more power and also probably gear down the drive train.

If you want to know more about the problems of skid steering click here.

We also decided to try out only having a two wheel drive and caster setup. However this looked really lame and wasn't nearly as strong offroad as the 4wheel skid steering design was. Also it was really hard to drive this way so we gave up on that idea pretty quick. More on that and other solutions in the next update.

Here are the videos.
They Actually go into better detail than this page does and they include video of the caster test etc.

http://www.cp-tel.net/pasqualy/land...es/DTtest01.wmv ~7 megs

http://www.cp-tel.net/pasqualy/land...es/DTtest02.wmv ~4 megs

Project Update 03-13-04

2004-03-13T02:58:00.000-08:00

ghhh, so much to update on. I'm not sure where to start. I'll try not to forget anything.

Even More Chassis Planning

First of all, I've done a little more design work on the chassis and drive train etc. I'm pretty much dead set on the motor in the middle design. I made a little video of the whole thing coming together and it makes things a little more clear.
drivetrain03-04.wmv about 1 meg

As you can see, I modeled the pillow blocks and flange bearings and made the tires a bit better than they were before. Its helping me decide on the actual demions of the frame itself. Right now its looking like the frame will be about 16-18 inches wide, with the tires that makes it about 30 inches max. Still pretty big.

Weapons Systems Work

Last update, I had just gotten my little BB gun. Will its been about two weeks and now my office is covered with little green plastic BB's. This thing is a BLAST!!. Big props to Z over at SomethingAwful.com for suggesting it to me. As I mentioned before, firing it with the SV203 board was just a simple matter of bypassing the trigger by soldering leads right onto the motor of the gun itself.

After that, I connected it to the 5V lines of an old AT Power supply from a junk computer I had laying around. It worked like a charm. Then I put a little reed relay between the gun and the power, the SV203 board controlled the reed relay via its on/off output capabilty. Perfect for the relay because it likes 20milliamps and the board puts out 25.
What wasn't so perfect was that the relay was only rated for 1 Amp throughput. I don't know how much the gun draws, but it was powered by AA batteries and those push 1.8 Amps. The relay tended to 'stick' in the on position alot. I fixed that by putting a power transister behind the relay. SV203 turns on the relay, relay turns on the transistor, gun goes shooty.

Mounting the thing on servos was a bigger pain in the rear. Our first attempt was just really really ugly, and crappy. But it worked. We screwed down one servo on the right side of the gun for verticle movement, chopped up an old power supply case to get a piece of angle iron and used that to mount one servo to another for horizontal movement. Then we just taped the hell out of that servo on a pole so the gun would have some clearance to move. It was really bad for the servos to be holding that weight, it hummed, it wobbled, it was ugly, but wooooo it was a blast!

Now, we still had problems with how to aim it and shoot it, and how to let other people over the net aim it and shoot it.
I tried a few pieces of software and 'Active Webcam' was the first one I got to work and work well. Though I'm not sure its the best for what I need. It is basically acting like a web server and a streaming camera. It serves out a single web page that contains the camera applet. However you can edit that web page to include what you want on it. That is where Serial Host comes into play.

Serial Host is a sweet little tool. Basically it converts incoming http requests to serial commands. "Hunh what?" you ask? Well its simple. I make a link on a web page, or email etc. When you click that link the request goes over to Serial Host and it then passes the command over to the serial port on your computer.
For example, if serial host was running at on my PC and I gave you the following link to use:

http://12.14.254.181:82/Pontech?SV203=MoveAbs(1,100)

Serial Host would take recieve that url request and then send the command over the serial port on my computer for the SV203 board to move Servo 1 to position 100.

If you still dont see how this would be useful, then just take a look at this:

Above is the final product of a toy gun, SV203 Servo Controller, servos, lots of electric tape, a webcam and software, and Serial Host.
There is my friend in Texas, shooting cups off of computers in my office in Louisiana.

I can't say enough good things about serial host. Steve McFarlin was very helpful when I had email questions as well. As soon as I have the money set aside I'll be sure to buy this piece of software because its just so darn useful.

Here is a video of the first version of our gun mount in action.
WeaponSystestFeb04.wmv about 7 megs

Trash vs Treasure

Over the weekend I visted my parents house and picked up a ton of goodies. I got some scrap lumber from the new house they are building, I'll be using it for a wooden mockup as soon as I have all my sprockets etc. I also found 3 or 4 sheets of 3/4in lexan or plexiglass. VERY useful stuff. It would have cost me probably $150-200 a sheet, but my parents were just going to toss it. I picked up some chain, a few sprockets, some old motors that are probably ruined, some old bearings and I got a chance to weld my sprockets to my hubs.

I had a chance to actually put together a quick mockup of one of the tires on a chain drive. I used a C clamp to hold the motor down, put the sprocket on, and ran a chain to the one wheel I had with a sprocket built on. The gear ratio was probably 3:1 so it was 3 times slower and stronger than what I will have at 1:1. I couldn't seem to stop the wheel using both hands. It was powered with a 12V car battery so I'm pretty pleased. These motors definatly have some umf to them.

I got my first donation to the cause as well. Tballz, a WWIIOL player sent me a box full of 'junk' that his company was going to throw away. Some fans, speakers, hardware, etc. The coolest was the handful of solor powered digital thermometers. Big thanks Tballz.

Weapon Systems II

I decided to redo the gun mount with my newfound materials. I picked up a router attchent for my dremel tool at Walmart along with a cutout bit. It works real well but you have to be very careful because once it gets hot its very easy to cut out too much. The Lexan sheets are a pleasure to drill too.

The base of the mount was made from an old power supply case (useful things to keep around!). We mounted a servo on the bottom and mounted a Lexan disk to the servo horn. 4 plastic 'bumpers' keep the disk from wobbeling too much.

Mounted on the disk are two upright pieces of Lexan. We used old computer expansion card slot covers as L brackets to mount them. Scrap is a wonderful thing.

I added a rod to the left side of the gun and kept the servo on the right side. The rod onthe left takes pressure off the servo and makes things more stable. However its not perfect, I may change it to a gear system so that the servo isnt directly connected to the gun, I'd also be able to limit the elevation with this method.

We tool the webcam out of its housing and used hot glue to stick it to the upper front of the gun. A much much better system. We had to move it later when we realized we put the wrong size arm on the servo horn and it kept slipping. So we glued the camera to a plastic tube that fits inside the gun and is screwed into place. That way we dont have to remelt the glue to get the gun apart in the future.

I also removed the handle, trigger, and clip from the gun. All of these were just spare plastic getting in the way. The clip was always there for show anyway, the BB's are kept in a hopper on top of the gun.

After we were all done, we turned on the web cam and the Serial Host server and let the guys over at WWIIOL and SA have at it. We had a blast letting them shoot old coke cans etc. We got hit a few times ourselves.
Here is a movie of the gun in action. At one point I had a pice of Lexan up as a shield and the sneaky boogers shot my fingers!
WeaponSystestIIMarch04.wmv

Project Update 02-28-04

2004-02-28T03:05:00.000-08:00

Last update I mentioned the little electronic BB gun that I got. I took the gun apart, figured out how it worked and then decided on a way for my SV203 board to fire it directly and bypass the trigger.

Here is an image of the gun itself.

I took it apart and here is what it looks like on the inside.

Basically, the trigger pushed an on/off switch for the little motor that was geared way down. The final gear had only about 80% of its teeth. It pulled back a piston of sort. Just as the piston got to its rear position a BB falls into the chamber and the gear runs out of teeth and allows the piston to fly forward and spit the BB out.
All I have to do is bypass the trigger and connect directly to the motor itself with power. The SV203 can act as an on/off switch but it wont push enough amps to make the motor turn. I plugged it directly into my 800milliamp power supply with no luck. Next I tried the 5V contacts on a PC power supply and that worked like a charm. So I'll have to hook up some kind of power transistor to allow the SV203 to turn on the high power needed from another source.

The whole firing mech is a bit more than I thought it would be and I cant rip it out and still have it work unless I make my own housing. The current housing is going to be a pain to mount to some kind of servo aiming system.

More Chassis Planning

Ever since I got the wheels and understood a little better about how I would mount the axles etc. Before I was unsure exactly what parts would all go well together etc. Finally, after seeing the wheels I decided on having a live axle. That is, the axle would not sit still with the tire spinning around it, it would actually drive the wheel.
The Motor spins the sprocket, the sprocket drive the chain, the chain spins the sprocket on the axle the axle spins the hub with the tire on it...

I found some pillow blocks and flange bearings that I liked at the roboto marketplace and I modeled them in 3DS Max and added them to my current chassis models.

Here is my first chassis design and my current favorite with the new bearing mounts modeled.

I've pretty much decided that the center motor design is going to be the one I go with. I also wanted to see if I could get the width of the vehile down by offsetting the motors so that they could overlap in the center. I made the following design. Its pretty drastic but it shows how the idea would work. I'm not really sure how big/small I need the whole thing to be but this 3D model is probably smaller than it will have to be in real life. Hopefully soon I will be able to make a final decision on the dimensions and then I'll just make sure I can fit everything in. There has to be room for batteries and the computer, and whatever goodies I need

If you have any suggestions or advice now is the time to give it. I'm pretty much set on the above design, all thats left is the actual distances between the tires and motors etc.

Project Update 02-25-04

2004-02-25T03:06:00.000-08:00

I got this little guy in the mail this week. It was only $33 bucks after a carton of 2000 BB's and shipping. The gun itself was only $12 or so. I picked it up from http://www.airsoftatlanta.com/electric.htm.
I was going to rig it up to my servo playground but ummm it shoots really hard and fast and I'm afraid someone will shoot my eye out and make a mess of my office. We shall see though.

I got these this weekend off of an old go cart that my dad had gotten for me and my friends and his grandkids a long time ago.
They are 14-15 inch or so, a little bigger than what I was originally looking for but they will do fine for my robot. I was really pleased when I saw them because I was afraid that the tires would be two sizes, smaller in the front and larger in the back. All together these things will probably save me about $120 in new tires and rims etc. I may have to replace some of the tubes though and they are rusted up pretty bad. Also, the hubs are not made for a live axle as my current design will use. So I'll have to replace them or modify them. I may just make my own.
The go-cart was driven by a single chain and the tire that it was attached to was pretty bald compared to the others, I may replace it later but its not so bad that it wont work at all.

Here are some sprockets I picked up as well to fit on my motors.
I showed my brother, Lee, my two current ideas for the chassis etc. He suggested I go with the version with the motors in the center of the two side wheels instead of directly connected to one of the wheels. This would allow us to gear up/down the system. I decided on #35 chain and got some advice on how to mount the axles etc.
Armed with this new info I made a few new designs and modified my old ones etc.
I'll get into that with the next update as well as some modifications I made to the little BB gun.

Initial Project Update 3

2004-01-03T03:09:00.000-08:00

Last I had told you, I had written a little application that let me control the servos via jstick instead of command line or remote desktop stuff. Some of you have used to software to move the camera around etc. At this point I figured it was time to get serious and move on from my brothers to building my own chassis etc.

I started doing lots of research and looking for parts etc.
www.northerntools.com
www.wmberg.com
www.robotcombat.com/store.html
www.lynxmotion.com
were some of the great places I found for parts etc.

I decided to buy my own motors and not try and steal the ones from my brothers robot. I went to ebay and got a sweet deal on a pair of used wheelchair motors. After shipping it cost me $75 bucks or so.

They are pretty sweet. They have built in gears and seem pretty strong but I havent really been able to test them as I dont have anything to fit on the shaft.
Here's a pic.

Now that I had my own motors, I started some serious design on the chassis. The earlier rendering was just a concept type thing. Now I know what I was dealing with and could start some serious design work.

The speed controller that I had drove things like a tank. Basically both wheels on one side are powered in tandem. To turn one side goes forward, the other backwards. I figured since I had 2 motors and not 4 I would use a chain drive to move both wheels with one motor. Here is my first design.

As you can see, a single motor drives the two wheels on a chain, the sprocket above the motor is a tension/idler sprocket that would be ajustable to make sure the chain is tight.

I modeled the motors more accuratly to the real thing using drawings from the manufacturers website. The wheels are 12.5 inches if you need to get a sense of scale. The body is 10x30x36in I think. It would about 4.5 inches of ground clearance.

My second design was to have the two motors directly connected to the wheels. This made the things a little simpler but the downside is that I wont be able to gear up/down the drive system by changing my sprockets out. This is the design im favoring at the moment.

Here is a closeup of how I would have to use two bearings (green things) to keep weight pressure off the motor shaft. I'd have to use some kind of coupling as well. (red thing)

Top View

Heres a little animation I made.
http://www.cp-tel.net/pasqualy/land...comotion001.mov

So this is where I stand.
Once I decide on the final design of the chassis I have to find the money to buy the parts and material. Then I have to find the time to head down to the farm and get one of my brothers to help me weld it. I don't weld well enough to do it myself.
Money is a big issue here. Any way I look at it, its going to cost me at least $200 bucks in parts including wheels, bearings, chian, sprockets, etc. Then I have to buy the aluminium for the frame. Then I have to buy batteries. Probably another $200 there. Though I have a few 6V's that I can rig up that I got from various broken UPS's.
I'm also looking at a mini itx board from via. They have one that has a built in 12V power supply hookup that will work perfect for me. Another $150 bucks or so. I have a notebook harddrive for it and adapter already lined up as well as a pci 802.11b card and antenna.

I'm also looking at incorporating something like this from lynxmotion.
http://www.lynxmotion.com/Category.aspx?CategoryID=25

I may build my own however and make it a little more robust.

Anyway, I'll post further updates as they come. As you can see above I just ordered a little electric BB shooter and I plan to mount that on the current battle bot for fun.
Let me know what you think about my chassis designs. Im very open to ideas right now. Personally I think the damn thing is too wide.

Initial Project Update 2

2004-01-02T03:11:00.000-08:00

After getting the prototype robot to work and prove that my idea was sound, I had to work on refining the method. Specifically I needed something other than the crappy bandwith hogging processor melting remote desktop method. I figured I'd have to learn to program well enough to write my own little client server application that would take joystick and other inputs from one PC and then send it over TCP/IP to the robot's computer and then dumpded down onto the SV203 board. Also the robot PC would have to take inputs such as sound, video, etc and send them back to the controlling PC so that I could see what I'm shooting at, where I'm going, etc etc.

We tried a neat little program that forwarded incoming telnet data to a com port. It worked pretty neet but you had to type all the commands in by hand. Very slow.
We played around a bit more with the virtual desktop method. Here is my friend Ogrenutz looking around my office. We just stuck a web cam on top of two servos.

It was pretty freeking neat having someone control something in my office like that, kinda creepy. For fun we attached a pen to a servo and I let him knock coke cans off the desk.

Unfortunatly I taped over the video of this. The best I have is me doing some servo tests. Fun with lasers and cigarette smoke!
http://www.cp-tel.net/pasqualy/land...s/servotest.wmv ~2megs

My friend roXet and his wife came over to play with it so we wired it up and had a go. I had gotten my own laptop working so we were able to be a little more brave with it, including putting the smackdown on a stool in the office. Here is a video of us rigging it up and driving it around. We did find out that the range of 802.11b is pretty pathetic. (duh) so one of the big obsticals will be getting that range increased.
http://www.cp-tel.net/pasqualy/land...t_tech_test.wmv ~6megs

Playing around with the thing via remote desktop worked but it sucked pretty hard. So I started playing around with Visual Basic and cooked something up.
Most of my programming consisted of tearing bits of useful code from other progams and stuffing them into mine so that I had this frankenstine beast that was a monster but it worked.

Using Visual Basic, I first got a little telnet type prog working that let me send commands via a text box or a button to the servo. Then I got it to take commands from a joystick and send it to the servo.
Then I figured out how to make a simple client/server chat program.
Then I used that to take data from a jstick, send it over tcp/ip to another computer, and have that computer send it to the SV203 board. It worked really crappy at first because the board was getting flooded with commands because of the way the code was written, but man oh man when my pal in Kentucky moved his jstick and I saw the little webcame jerking around I was really excited.
About a month later or so, I finally got around to fixing things and got rid of the flooding problem. What I really need to do now is get video from the robot PC over to the controller PC. If any of you coders out there want to help me with this I would really appreciate it. I'm not much of a coder at all.

Anyway, another update soon. I'm almost caught up to where I stand currently. Just have to tell you about the motors and chassis design etc.

Initial Project Update 1

2004-01-01T03:12:00.000-08:00

Here is an early mockup I made in 3ds Max.

The idea is to have a PC mounted on a mobile platform that controls the platform's systems like motion, sensors, and gun turret, etc. That PC can then be controlled via wireless ethernet or cell phone dialup or whatever from another PC. Sort of like a mars rover.
In my back yard.
With guns...

The weapon system will be based around this thing from www.gatlingguns.com

I'll have to check with the ATF before I actually mount the thing I guess. Appently as long if it only fires one bullet per keypress I'm ok, otherwise its a machine gun

Right now I'm using my brother's battle bot as a prototype platform. He actually competed but got whupped in his first fight. So I borrowed the bot, got it working and then started chaning things to my liking.

I'm using a SV203 servo controller to interface between computer and the speed controller etc. http://www.pontech.com/products/sv200/

As you can see, it basically takes data (in the form of ascii commands) over a good ole serial port and controls servos, reads voltage, or acts like an on off switch.
The "controls servos' part is the important one. If you imagine any ole R/C car or whatever its basically got a device that takes input from a radio control and moves servos and speed controllers. Instead of doing it with a radio handset, this thing does it from a serial cable.
This allows me TONS of flexability. I can plug in the SV203 and control it with commands over hypertermal, or I can write my own software and do it with a joystick, or a gui of buttons and sliders. I can also set up client/server software so that I can send commands over tcp/ip from a remote pc to the pc that actually is connected to the SV203.
Then if I want to I can just replace the SV203 with any stander R/C reciever and use a radio control to dive it around.

Here it is connected to my laptop and a pair of servos. You can see the software that comes with the board on my screen.

Now that I've given you guys a brief intro to the project, here is a chronicle of what has gone on so far.

The first task was to get my brothers battle bot in working order. Its basically a box with 4 14.4 volt Drill motors in it and some kinda chainsaw thingy. It was the first robot he built but it turned out pretty nice. You can stand on the thing and drive yourself around and when the batteries are at full charge you cant stop it as long as it has traction. It was in pretty bad shape when I got it though. It has been impaled on the spike strips and some of the stuff inside was gutted when he was helping another team out with their robot.

I had to first figure out how it was supposed to work and then get it all wired up agian and make it work.
Here I am with the speed controller. Its a $300 dollar unit so I was really nervous playing with it.

I finally got it to work. Here you can see his pimping R/C controller. The reciever is that little black box above the speed controller (the blue metal box).

Once I actually got it working again (and crashing it around the office and destroying a few cardboard boxes)it was a simple matter of replacing the R/C controller with my SV203. At that point I could control the wheels with little sliders on the screen. However this was pretty lame because I was thethered to the darn thing with a serial cable.

So as a test we set up a quick wireless lan. Then I just remote desktoped onto the laptop so that I could control the robot from my desktop pc. Now this was a terribly terribly ineffecient way to do things. But it was just a test and we didn't have time to do anything different.

Here's how it looked, please don't laugh too hard

We were doing this with my roomates way expensive laptop so we had to make sure it was in a nice protective box with packing peanuts and everything. Here is a video of how it all worked out. Crude but it showed the idea was sound.
http://www.cp-tel.net/pasqualy/land...controltest.wmv (17megs)