Switch branches/tags
Nothing to show
Find file History
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
..
Failed to load latest commit information.
assets
.catalog_metadata
README.md

README.md

#354 SpiderBot

Boldport Club PissOff mutant offspring .. with spider legs and scuttling drive system.

SpiderBot

Notes

Somewhere in my travels - IIRC, at [Five Below](diy spiderbot tech kit - I picked up a cheap "spider" robot kit .. I think because I wanted to examine how they drive the legs.

But for something a little more interesting, I thought I'd combine this with LEAP#290 Boldport PissOff. The result: a motion alarm that not only barks at you, but scuttles back and forth.

I think I'm channelling Sid from Toy Story ..

SpiderBot Kit

It's a pretty cheap and basic kit, with a simple direct drive motor system that only does one thing. Good for about 15 minutes fun I guess!

the way they drive the legs with offset cams is pretty interesting, and impressive how they've managed to deliver it as a press-fit plastic kit.

kit_box

kit_parts

Assembled the "normal" way...

kit_assembled.jpg

Linking the PissOff

The PissOff has an active low "audio enable" signal that appears to be the most useful trigger to use to control the SpiderBot mechanism. It goes low (~0V) when the alarm is sounding. At other times it is at Vcc (~3V).

Unfortunately, there is no existing breakout for this signal, so it requires a bodge wire to tap the circuit. I attached at pin 1 of IC1 - the most accessible point.

Rewiring the SpiderBot drive system

I decided to use a simple oscillator to get the spider legs to go forward and reverse when triggered by the PissOff.

I'm using a 555 timer is configured for a fairly slow ~50% duty cycle of around 2.3s high / 2.3s low.

The difficulty turnout out to be getting it to control the motor. It is a terribly noisy motor in the SpiderBot! I tried a few other DC motors that I can drive quite effectively, but this one is totally obnoxious!

  • it totally messes with the 555 timer circuit
  • I've tried all manner of isolation:
    • motor is switch with relay (with flyback diode)
    • I've tried totally isolated electics (no common ground). It didn't help so the current incarnation does have a common ground.
    • hefty smoothing/bypass capacitors
    • importantly C7 capacitor across the relay switching FET was critical in getting anything decent

I refused to give up, though the results I have so far are not great. It does not have a balanced duty cycle:

  • proceeds forward (motor relay normally open) OK
  • but the reverse cycle (motor relay closed) is unstable - it gets tripped off very easily

Here's what happens to my nice uniform 555 square wave when the motor is attached an under load:

drive_wave_under_load

Motor Controls

I've drawn up the motor control system with EasyEDA - the schematic source is here

The Schematic

Construction

And here's the basic layout on to a protoboard "backpack" for the PissOff:

SpiderBot_protoboard_layout

SpiderBot_build

SpiderBot_rear_flank

SpiderBot_front_flank

SpiderBot_rear

Credits and References