The Walking Box

walking box

[Ryan Walker] had recently constructed a 16 R/C servo controller board and needed a platform to test it with. He wired 6 cheap TS-53 servos from Tower Hobbies to the bottom of an empty box. He’s got push buttons on the top for direction control. It uses a PIC18F452 for processing and the board was routed on an LPKF circuit board mill. There are videos on his site, but try using these cached links first: walking on a table and walking on carpet.

20 thoughts on “The Walking Box

  1. >>>Anyway, don’t know the reason for using the 8 fets or whatever they are, since PIC could drive the servos directly

    Yea, like driving motors right off io pins on a pic that source like 10ma makes sense…

  2. Actually I used regulators to provide 5V power to the servos, the pic can only safely source ~20ma per channel. The servos can pull upwards of 400ma if you stall them. The whole board could be run strait from a 5V receiver pack and bypass the regulators, but 7.2V packs are generally cheaper and have way more capacity.

  3. Ryan,

    how was that PCB made? It looks more like it was was a copper-cladding that was engraved, to form a circuit, rather than doing it with light-sensitive methods. Most interested in the former method.


  4. owen,

    I used an LPKF circuit board milling machine. It uses a small .8mm rotary cutter to remove the copper between traces. It also drills the holes with changeable drill bits. There are contour router bits to cut the perimeter of the board from the copper sheet. It is a slick setup but the copper is prone to corrosion and looks terrible once there are finger prints on it. This could be solved with a protective spray of some sort, but it is generally used more in a prototyping environment than long term use.

  5. [Yea, like driving motors right off io pins on a pic that source like 10ma makes sense…]

    Hobby servos have built-in motor controllers. They’re driven with PWM output from the PIC’s IO pins.

    I’d guess that the power transistors are to provide speed control in addition to the standard position control.

  6. @6: The PIC only supplies the servos with position information using PWM at a digital input. The PIC does not drive the motor directly – the servo itself handles the high current stuff.

    My question is why have a regulator for each servo? Why not just use level shifting circuitry for the PWM input on each servo? (A couple of resistors and a transistor, maybe less?). That way you can still use the full 7.2V pack voltage. I wouldn’t be surprised if the servo even worked correctly with just the 5V PWM signal from the PIC without any voltage translation.

  7. flinty,

    When building this circuit i took the “better safe than sorry” approach. I’m not sure of what the logic circuitry can handle inside the servo, so running a voltage regulator will save the trouble of burning up all the servos if a much higher voltage is applied as the source voltage. They may function fine at high voltages, but i figured regulators were less than a buck and servos start at $10 each. maybe i’m just paranoid ;)

  8. You are right, I once applied 9.6V to a microservo, there was just a small spark inside and did not work any more:) I had to replace a capacitor and it started to work again. Not sure how the capacitor could have caused it, though.

  9. Cool. I see your point. I was thinking that you might get more torque/speed out of the servo when using a higher voltage, however I now remember that in my model aeroplane days 4 NiCd cells at 4.8V gave the servos plenty of juice.
    Now i’m wondering how I got away with 7 NiCd cells (8.4V) in my RC car. It must have been right on the limit of blowing up, or the receiver/servos had voltage regulation built-in.

  10. timbo,

    I was told the C30 system purchased by the school cost $20,000 for the hardware setup (requires a vaccuum system and some regulators as well as the mill). The one at my work is much bigger and in the neighborhood of $60,000.

  11. flinty,

    you are correct, modern R/c cars with electronic speed controls contain “battery eliminator circuitry”. hooking the speed control (which is in turn hooked to the ~7.2v pack) to the receiver provides ~5.5v at around 1000ma to power the servos and radio receiver. in the case of older mechanical speed control setups, a separate receiver pack 4.8-6v is needed in addition to the 7.2v or 8.4v pack to supply the radio receiver and servos.

  12. I would love to get servo running of my 18F458 PIC! Do you have a any source code? Our PCB milling machine cost us 50 000$ cdn and NEVER worked good. We order our pcb in Calgary and I live in Quebec.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.