Power supply for robotic arm

[tuxerman]

My buddies and I are working on this manually controlled robot, which has an arm-like mechanism with five servo motors, which can pick blocks and place them at other places. It moves with the help of DC geared motors, while using servos for its limb-movements. I tested the servos and found that they take in a max of around 600 or 700 mA while retaining their position despite me trying to move the shaft with ample force.
I have a lead acid battery source, and need to power these motors with 5V. Guess 12V is what I'll need in the lead acid department, but I am confused over the type of regulator to use. A linear reg with say lm317 and pass transistors would result in an insanely large drop, thereby requiring a massive heat sink. Should I go for a switched mode power circuit? Buck converters?
Pl suggest suitable ways to drive around 6 to 8 A from the battery, supplying at 5V. Can normal buck circuits supply that much current?

 

[ronv]

Have you thought about using a 6 volt battery and adding a diode to drop a little and protect your circuits from a reversed battery at the same time? May take a couple of diodes as a 6 volt battery is really 6.36 volts full, 5.3 empty. If that won't work because of the voltage range of the battery the buck is probably the best thing.

 

[tuxerman]

Thanks, but I'm concerned about the current supplying capability. What modifications should i make to the buck so that it can supply around 7 or 8 amps? And I'm assuming, with the buck, i neednt use a large heatsink.. am I right?

 

[OutToLunch]

a single channel buck regulator can easily supply up to 30A (sometimes more with proper layout, component choice and airflow) so 8A is no problem. I would suggest a synchronous buck to get rid of the freewheel diode drop loss. There are a ton of controller ICs out there - many with application circuits/evaluation boards already designed that would either fit your needs or require only minor modifications (intersil, ti, semtech, etc). There are also off the shelf converters that you can buy as well (vicor, murata, etc).

 

[tuxerman]

Can I use a 555 to generate a square wave of appropriate duty cycle, and feed it to the gate of a power MOSFET? Suppose I use the MOSFET of sufficiently large current rating, the motors would get sufficient current, wouldn't they? Also, would around 50kHz suffice to provide the necessary smoothness of operation (after a shunt capacitor filter, that is) to the motors, etc? Do reply, folks..

 

[OutToLunch]

for a motor supply, i would strongly suggest a system that incorporates feedback control. when the motor kicks in, it will draw a lot of current and the duty cycle will have to increase in order to ramp up the current in the inductor. a 555 just throwing out a constant duty cycle will result in a sag in the output voltage and the proper amount of current will not be delivered. Not a very robust system - you want to be working on the control of the arm - not the effects of the power supply.

 

[crutschow]

Thanks, but I'm concerned about the current supplying capability.

Are you concerned about the current capability of a 6V battery? 6V batteries generally have a much higher current capacity then a similar physical size 12V battery.

 

[tuxerman]

OKies, so this would ultimately require a full fledged DC SMPS or what? Feeling lost now, could someone post a few links as to how to implement this feedback with a discrete circuit?

 

[OutToLunch]

it's not that hard to design - the controller ICs already have a lot of it worked out internally. board layout is likely the thing that would kill your supply. just read a few datasheets like this one...
https://www.electro-tech-online.com/custompdfs/2010/08/fn9255.pdf

 

[ronv]

Here is a 5 amp one. (5 X .7 = 3.5).

12V / 24V to 5V, Step-down, 5A, 25W - $14.50 : Current Logic, Expert of DC-DC Converter