PDA

View Full Version : Beginner build



Carl Gustav
12-12-2017, 07:51 AM
Hello all, I've been thinking about building a robot to compete for a number of years now. At first I wanted a cam driven hydraulic actuated robot like the sprawl robot, but then I was informed of the No cam driven rule. Now I'm thinking of building the mini kame from thingiverse.com as a base and worry about designing and building the top half later on. After I start to get the hang of this robotics stuff. But I do want to explore more than just learning to make a quadruped walk.

First of all I would like to buy a set of target plates for a "light" class mech so that I can see what dimensions I have to work around. Which 5.8 GHz video stuff should I buy? Is there a standard setup that most people seem to be using or anything?

I have a few questions though, if anyone is brave enough to try to help.
Do you think these servos are strong enough to use later on after I add the top half of the robot? (Turnigy TGY-306G-HV)

Also, does it matter which xbee I use? I was going to get the basic series 1 but I don't know if that one interferes with other xbees or something, or maybe if you guys have had better luck with the pro?

I haven't bought anything yet except a picaxe to try to learn. I'm going to buy xbees and arduino unos to learn Rx tx, and I think the 5.8 video should be simple enough to setup and use, no? I think I can adapt the kame into a mechwarfare competitor, I'm very uncertain about how to setup a xbee controller with joysticks and buttons though, and I'm not sure if the arduino uno is the best board to use to adapt the kame's servos into a wirelessly controlled robot. A HUD would be nice but i have no idea how to do that. Also unsure on how the wiring for the gun would work. I know alot about airsoft as a longtime airsoft player, so I was thinking of using a plastic version 2 gearbox, and upgrading it to a dual sector gear setup, to double the rate of fire, and half the fps. I'm not sure if an airsoft mosfet would interface with an arduino but I know MOSFETs are better than regular brass trigger contacts on airsoft guns. (The brass always wears out/gets bent out of place) I can adapt the innards of an angel custom aeg drum mag into a smaller package, for the feeding of the bbs.

Here's the pages for the kame robot I've been referencing

https://github.com/JavierIH/miniKame/wiki/BOM
https://www.thingiverse.com/thing:1265766/#remixes
https://m.youtube.com/results?q=qwerty4&sm=3

I like it's speed and the weird gaits it has. Maybe I could adapt the code from its original gaits into arduino controlled walking, I hope.

I would love to make this all by myself but it seems I'm having a hard time finding resources on how to do alot of this stuff. Btw, mechwarfare needs way cooler buildings, and it needs like a propper announcer, and like, a video series. Im a modeller, I could make super realistic buildings and roads. I could help make a YouTube video series too. It'd be great to make mechwarfare as popular and as cool as I think it deserves to be.

Carl Gustav
12-12-2017, 08:05 AM
I just thought of something else that would be cool too. What if there was a way to make a gas blowback airsoft gun mount to a mech? It would be possible to make essentially the lower half of an airsoft Glock in the shape of a robot arm\turret. I can get custom hpa high cap mags, that would take care of the feeding. Use a small 88 gram compressed air tank, regulated to like 90 psi. And just add the slide\upper half of an actual airsoft Glock to the bottom half you've 3d printed.
I saw a 3d printed bullpup chassis for an airsoft glock out in the field before so I know it's possible. It would be so rad for your mech's guns to have recoil. Imo. There are also shotgun shells you could set up to be used like a launcher. Cam 870 shells or goblin shells. You fill them up with green gas or propane just like how you would on a traditional gas blowback magazine, and it releases the gas and fire the bbs when the pin on the back of the shell gets pushed in. You would just need a housing to slide the shells into, and actuatorsacting like a firing pins. You wouldn't really need any kind of barrel.

There needs to be far more videos done on this btw. I hate watching the same videos with no commentary over and over. This has so much potential. If there were different maps essentially, with different layouts other than the basic grid the mwf competition uses, it could be way more fun. To watch atleast, and I would assume more fun to pilot aswell. A desert or a snow map would be amazing to watch, but even a different city map with a more asysmetricical layout, and some low buildings, would really add more dynamics to the battles.

jwatte
12-12-2017, 09:49 PM
mechanics

Xbee series 1 is what most of us use.

I think you'll have a hard time getting enough space on the Kame for all the necessary equipment.

The Turnigy servo is probably good enough? Maybe? However, the Dynamixel XL-320 is cheaper, lighter, and just as strong (if not stronger.) Might be slower perhaps?
But, also, the XL-320 works well with the OpenCM controller, and doesn't need a zillion little PWM output ports, as it uses the DXL bus. And you can read load/battery/status back from them.


something else

The airsoft gun must have an exit speed in the 160 fps range (the exact upper limit is in the rules -- I think it's somewhere below 200 fps.)
I'm also not 100% sure of compressed air cylinders -- will they, or their feeder tubes, explode if hit point blank by another robot's legs or BBs?
Flammable gas (like propane) is probably a bad idea in the vincinity of LiPo batteries.


There needs to be far more videos done on this btw.

I agree that the sport is under-marketed. The reasons have to do with people not having time -- what time we have, we want to spend on our bots :-)
Also, having a camera man inside the arena is not particularly safe for the camera, the man/woman, the mechs, or the arena.
The R-Team people are adding more stationary cameras to the arena this year, though, which I'm sure will help in capturing footage. Then anyone with a lot of time and some Premiere skills can do what they want with it :-)

Carl Gustav
12-12-2017, 10:47 PM
I just ordered some xbee series 1's, and an arduino uno r3. I'm going to look into the OpenCM controller and the xl-320 servos. Thanks for the tip.

I think dual sector gears in plastic version 2 gearboxes should be perfect for this. Or you could use it in semi auto, and use a cut m90 spring, and poke holes in the nozzle to lower the fps. But I do think having anything in the vicinity of lipo batteries is a bad idea though, lol. Even though I put my face right next to them while they're running quite often. I just won't leave them in my house or my car.:tongue:
I should contact the guys from the Tested youtube channel to do an interview with someone here for a video. Or maybe the guys from FliteTest would do it. It is fpv and radio controlled after all.

Do you (jwatte) know who I should contact to get 'light class' target plates made? I guess they would want me to build the mech first and then they'll make the light plates, should it require them. I need to know where to make space for them though. I think the rules for 2018 says that light class mechs only have 2 plates, one on either side, and not the front or rear. That cant be right, can it? How will the enemy score hits?

jwatte
12-13-2017, 12:01 AM
For the plates, the drawings are available in the RTeam git repository if I remember correctly.
You should check in the "rules for 2018" proposal thread, they might have links there.

giantflaw
12-13-2017, 12:50 PM
First of all I would like to buy a set of target plates for a "light" class mech so that I can see what dimensions I have to work around.

Also, does it matter which xbee I use? I was going to get the basic series 1 but I don't know if that one interferes with other xbees or something, or maybe if you guys have had better luck with the pro?

A HUD would be nice but i have no idea how to do that.

mechwarfare needs way cooler buildings, and it needs like a propper announcer, and like, a video series. Im a modeller, I could make super realistic buildings and roads. I could help make a YouTube video series too. It'd be great to make mechwarfare as popular and as cool as I think it deserves to be.

Light Mech has 4 half size plates, one on each side. Can be mounted vertically or horizontally. Loaners are available at Robogames. Dimensions are in the rules and they are mounted with Velcro

xbee 1mw works for at home but at Robogames it has been problematic so many people have changed to the pro for more power to overcome the Robogames radio interference. Also some have switched to the 900Mhz band. Both the pro and 900Mhz band worked great in 2017.

A few people have used HUDs in the past and have complained about them. They said they would not use them again. Not sure why.

We are going to have nicer buildings for 2018. The 2017 map layout was much improved over the previous years although it was symmetric with 8 buildings. For 2018 we have more variety in the size and number of buildings. As far as video we are adding half a dozen gopros to the arena for 2018 to get more video. I think everyone involved with this event would like a better presentation.

Carl Gustav
12-14-2017, 08:19 AM
Ok cool, thanks for the tips. I really should try to get my 1mw xbees i just ordered exchanged with pros.
That is a little disheartening about the HUD, but I could see it making it harder to see what's going on, or maybe causing lag or video issues. I still really want to know how to make a HUD though. It would be great to make a HUD that displays your remaining points. Make the HUD flicker or warnings flash when the mechs near the end of its (health,hit?) points. If they could get everyone to have a overheating effect implemented on their robots that would definitely solve the "not moving and just shooting at eachother" problem. Make it so you can only fire full auto for about five seconds or less or else the guns stop working for say 9 seconds. Another dream of mine is to have my mech completely collapse when it's points reach 0 so the other pilot really gets to feel like he's taken down a mech.
I definitely liked the layout the best this year by the way. Some more alleys, and some buildings that are low enough to shoot over, and it would be really rad, maybe a tunnel you could try to fit through or a overpass you could walk up onto.
Wouldn't it be amazing if the mechs could walk so well that they could walk from rooftop to nearby rooftop, or if they could climb? I've got this grand dream in my head where people are controlling the mechs from hundreds of miles away in the their homes, and the matchs are like 5 v 5, and the mechs are climbing up onto the 1/18 scale buildings, flying around with ducted fan jump jets, shooting down the occasional attacking scale RC helicopters. I know this probably sounds very patronizing, so I'm sorry if an anyone feels like I'm downplaying all the work they've already put into their mechs and the arenas, and events in general, like I think its all been amazing. I'm just saying, like, here's what I'm going to try to do if I can, and maybe it gives someone else an idea they never thought of before. Imagine, if this got big enough and got maybe tons and tons of funding and employees on the clock working on it, it could get popular like video games is for general people. Then there would be tons of kids learning and thinking about robotics and electronics in general, it would certainly change the future in the way we'd like to see it. In the "more robots" way.

giantflaw
12-14-2017, 11:02 AM
To tell the truth, it has been very hard for people to actually build a basic walking robot with a gun on it that qualifies. One of the guys this year tried for many years in a row to qualify without succeeding. This year when he did qualify he had tears in his eyes. I would suggest putting in consistent long hours every week toward building a basic mech first with no frills. Just get it up and running so it can qualify. Test early and often. There is much learning in just doing that. Multiple competitors have built several robots to get where they are now after learning from the previous designs.

A finished inferior mech is better than a half built great mech.

7107

GhengisDhon
12-14-2017, 11:02 PM
Btw, mechwarfare needs way cooler buildings.

WHAT! you don't like my buildings?!

Seriously though, building a mech takes quite a bit of time and dedication. For a light mech, your best starting point might be a Metabot. We've had a few people in our R-Team Club build them and they are fairly straightforward to get up and running. You can use mostly Dynamixel components and control them through an Android device. Your biggest challenge for any of the light Mchs is going to be finding a small enough gun/weapon that they can carry and is effective. We can probably put together a part list for light weight metabot sans the gun.


https://youtu.be/IpluoEAx314

Here's the half size target plate dimensions for 2017. The 2018 plates will be similar.
7112

I'm working on plans for a full size Mech to pubicly release that should help new mech builders put together a starting Mech... but between work and life I haven't gotten as far as I'd like. All I've really managed to do is get my mech torn apart so I don't have a functional Mech atm.

GhengisDhon
12-14-2017, 11:07 PM
A finished inferior mech is better than a half built great mech.


For some reason gianflaw's quote made me think of this vid. Not sure if these are the inferior Robots or the hal built Robots.


https://www.youtube.com/watch?v=g0TaYhjpOfo

giantflaw
12-18-2017, 11:59 AM
T
A finished inferior mech is better than a half built great mech.


OK, probably too strong a statement. How about "A working mediocre mech is better than a non-working great mech". Several times at Robogames machines have placed in the top three places by just working the whole time and not breaking. Many know this too well by first hand experience. Our first time out years ago our club Mech RA was an inferior machine. It walked slow, shot slow (once/1.5 seconds). We did break our first fight but it was a draw. Our second fight we never fired our gun because we forgot to arm it. The other machine was self-scoring when it fired. Because we did not break the rest of the weekend we took 4th place against greater machines.

7116
The second machine we built, RA 2.0, was much better and on Saturday we beat Immortal for the first time. That night we made S/W changes to improve things. We introduced a bug that knocked us out of the running on Sunday. Did not even do as well as the first time we fought with RA.

Carl Gustav
12-26-2017, 05:57 AM
What are s/w changes? Size and weight? And I would like to have a complete set of backup parts for breakages giantflaw. Yeah it's definitely going to take alot of work and luck, I don't expect it to be easy, but I would like to get the basics down and make them seem simple to me atleast. I'm super impressed by all the mechs and the arenas by the way, way better than I can do, but I'm guna try to do better and win.
This should take me along time if I have to figure everything out from trial and error and posts and videos. I'm going to want to learn 3d modelling so I can make custom parts, more xbee stuff, ik, pan and tilt for fpv cameras that corresponds to head movment, more scale modeling, and the hardest part I bet will be getting everything setup for the ide, I hate trying to get libraries to work, getting the programming cables to work, and finding the right com port and things like that.

giantflaw
03-05-2018, 10:21 AM
s/w = software