MMM to power Robostix
 

The MMM having a switching PSU for the BEC makes it attractive for robotics projects. So long as your brain can take in 5.8v, and not draw too much current, the MMM should do the trick.

A nice little AVR board that is fairly inexpensive and complete is the Robostix by Gumstix (I own the Roboaudiostix.) This board can operate independently of a Gumstix computer or provide a realtime brain for motor control while communicating (and powering) a gumstix computer.

The Robostix and the Roboaudiostix take a voltage between 5.5v and 9v (with voltage above 5.5v being spent in heat.) The MMM outputs clean power at around 5.8v which is darn close to ideal.

The Roboaudiostix, unmodified, will allow the MMM to power other servo's connected to the Roboaudiostix but the board, its self, will require another power source.

On the Roboaudiostix, the PWM (servo) headers negative pins have a connection to ground shared with the BOARD power, and the positive pins are connected to pin one on V_MOTOR. Next V_MOTOR is V0 which is connected to BOARD pin one.

V_MOTOR is there simply to allow for easy installation of an additional power source to keep BOARD power and SERVO power separate. It does not power a motor.

Simply jumpering V0 and V_MOTOR pin one will allow the MMM to power the Roboaudiostix, thus allowing everything to be powered by the MMM.

I cannot speak of the plain Robostix. I have read that the plain Robostix already has V_MOTOR pin one jumpered to BOARD power pin one. If this is true, the plain Robostix requires no modification. Finding out would simply require an ohm meter.

Enjoy.

Sounds good, but I would suggest adding something to the BEC if you plan on using a switching one. Switching BECs can fail and if they do, will send full battery voltage to the brain.

Depending on the optimum brain voltage, main supply voltage, and brain current draw, you could either use a linear regulator running off the battery (but the current would have to be around 100mA tops for ~15-20v main supply).

Or, attach a low dropout type set to 5v at the output of the MMM BEC. The BEC would be doing the majority of the stepping down, but the LDO linear reg would serve as a protection device if the BEC went bad. To keep current to a minimum, you could just feed the brains through this, and leave the MMM's BEC as-is to power other things.

good to know, I'll look into that

On a positive note, the battery source I am using with the robostix is 7.4, which is within the limits of the robostix. So, if the voltage did run away, I'd only have to worry about other devices on the PWM headers, but it's good to know I should think twice about plugging in a 6S LiPO ;-)

I have a 1800kv motor in my Slash so I can fly at 6S, and putt putt at robo speeds on 2S.

Sounds pretty cool!:smile:

The V3 MMM has a new sacrifical TVS on the BEC output from what Patrick's said, which is designed to kill the BEC output if it ever rises above 7.5V...so there shouldn't be a risk of a full-voltage runaway if the BEC does fail. At least, that's what I gather to be the case, so CoreyFro's Robostix should be safe. :wink:

Yeah, hopefully that will work, but a linear reg at the output is just a little extra insurance. Depending on the cost of the circuit it is powering, it might be worth it. In corey's case, 7.4v is not a whole lot extra voltage, so even a linear reg by itself might work fine. Linear regulators have built-in thermal and overcurrent protection. It is really hard to blow one unless you are really really trying - and the only way I know of is to overvolt the input. Even then, the worst that would happen is the output would be 0v.

OTOH, using a pass-transistor to boost the current output of a linear reg can cause problems during a failure since they do not contain protection circuits...

Ah, fair enough. I guess it's always wise to have a bit of insurance. :smile: