Peter Wallers Robot Ramblings

[peterwaller]

I turned the suction on then wound the blade up to full speed.
I then turned the spinner off and when it had wound down to a reasonable speed I turned off the suction.
I then wound up the spinner until it jumped so the suction had been off for a few seconds so I think the air pressure would have equalised by then.

[peterwaller]

Not to sure what to do next on the full body spinner so I have gone off on a tangent and produced a new Succulant with a standard spinner.



Chassis is 3D printed and in trying to save weight I have made the front end a little flimsy but even with the cover weight with a 350Ah pack is 139 gms.
To compensate for this I have made a 0.5 mm Polycarb cover that strengthens and stiffens the front end and protects the electronics from the blade.
Here is a close up showing how much the chassis flexes when the suction takes hold.
http://youtu.be/Fx73cvNYNAs
Here is the spinner running without vacuum it appears not to pull down on the rougher surface of the arena so I need to make some adjustments.
http://youtu.be/lX9wP7tKou0

[haz]

Peter that is brilliant, looks very mean and very efficiant. you really have got the spinners down to an art now
that is going to be tough to beat, wouldn't surprise me if it won the AWS. i like the wheels on that, they look grippy.

[Andrew_Hibberd]

I watched the video before reading your post Pete, the blade goes anti-clockwise and as the robot lifts off the ground it also goes anti clockwise which isn't torque induced. It looks like the disc is pushing down, or flexing as it is spinning and scraping the ground. That little bit causes the vibration which makes the disc dig in and the explosion. Whats really odd is with suction it doesn't happen, which could be flex in the chassis??? I think its a similar issue with the huge blade I had in anticipation. No idea how to solve it though...

[peterwaller]

With the inconclusive results from my tyre testing ie the Gecko tape not doing as well as suggested by the perfomance of Void in actual battle I have decided to hedge my bets.
The diamond cricket bat rubber has been lost and a variety of tyres are now fitted to the team.
Combatant and Defiant are back to sticky blue rubber.
Variant and Fleabite are Gecko Tape.
Succulant is soft rubber.
Deflea is still bits of rubber band stuck to the tracks.
Alsoran is HP paper feed roller rubber.

[peterwaller]

I have for some time been looking at ways to improve my chances of winning battles and it seems to me, and it probably comes as no suprise to you, that I need to improve my driving. After every event I promise myself I will practice before the next one and every time I never quite get round to it. Even the new Succulent was only tested to see if it griped and spun OK but I didn't even try the weapon against anything solid. Certainly built in mixing and rate swithches have helped so I am thinking of going one further and introducing a bit of autonomous operation into my robots. Because we have no real standardisation of arena, like that we have in mini sumo full automation would be next to impossible but some degree would certainly be possible. I am thinking of adding a mode where the robot once close to the opposition would keep turning to face the other robot using either infared or visible light senders and receivers. This mode would be manually started and stopped using the 4th channel to avoid locking on to walls etc and probably involve an LED on the robot to inform the driver when the opposition had been detected. This will involve a bit larger processor with more I/O but most of what I need is already built into my minisumo boards so it shouldn't be to big a a step. I might start by converting an existing robot possibly Dominant as it is one of the more suitable that is not part of my fighting lineup.

[BeligerAnt]

Nice to see you pushing the frontiers again Peter

Will we have to introduce rules about laser-guided weapons?

[joey_picus]

Technically Plunderbird were first in 1997..."I want to know what a fixed beam laser guided spike is" "Well, it's a bit of serious business, because we're serious people..."

I like the idea of this a lot though, I'm imagining that using a microswitch or other pressure sensor you could even make the weapons autonomous too (although on a serious note, would something like a mousetrap which springs automatically on contact be counted as an autonomous weapon and be bound by the transmitter rules?)

[peterwaller]

Slight problem in that both Dominant and Defiant bodies are so thin that the motors touch both the bottom and top so very difficult to connect up sensors and electronics without added lots of wiring.
Have just designed a deeper body version that has positions for opto units at either end and ordered it from Shapeways.

[peterwaller]

Finally started to make some progress.
The chassis is made and the circuit design is finished (hopfully correctly).
I have just etched the main processor / drive board plus the front and rear sensor boards.
Here seen below laid in the chassis to gether with 4 100:1 HP motors and wheels.