PDA

View Full Version : Microcontroller Board Heating Up



swabygw
05-16-2011, 06:14 AM
I have two P60 gearmotors (RS-550) connected to the .NET Serializer to each motor terminal. However, it seems to be overheating my board, particularly the two, main motor components. They'll work for about a 20-30 seconds, and then the motors start sputtering (I think I can smell a small amount of burn from the motors) and the LED lights are sputtering on the board, too. It seems almost as if they motors are struggling to turn from the power being delivered or back-current is occuring.

Fortunately, I have a fuse in place the protects the board, but keeps blowing within seconds of receiving power. I know it's not the board, because I've used the board successfully with smaller gearmotors. So, I've tested the gearmotors on an external power supply and they seem to draw about 3-4 amps (12v each). Any ideas? Under what conditions would the motor components heat up and/or the LED lights sputter?

Thanks, in advance.

tician
05-16-2011, 11:31 AM
From the trossen page (http://www.trossenrobotics.com/store/p/5152-RS2-550-Motor-12V.aspx), it appears the RS-550 pulls 1.2A at no load, but can pull ~85A at stall. They also most definitely need noise dampening capacitors across their terminals. What is the current capacity of the motor drivers on the serializer? What is the current capacity of the power supply you are using?

iBot
05-16-2011, 11:45 AM
Don't those drivers have current limit around 4A ? With a gearbox I would guess that the steady load of the motor is around 3 or 4A as you suggest.
At startup the motor will pull more around 20 to 30A likely putting the driver into current limit. To get the current down to the 3 or 4A the motor has to spin fast enough for the reverse EMF to reduce the current which may never happen. If you measure the DC resistance of the motor you will better understand the startup current.
Also check if the current limit on the driver is set right and if it is based on measuring current or on chip temperature.
Capacitors will help noise, but not provide enough current to overcome this. I think these motors will need a beefier driver.

swabygw
05-16-2011, 12:04 PM
I spoke with Banebot and they said the same thing...use smaller motors or a different motor driver. It looks like the motors are drawing much more current than the motor driver limit, so it's heating up the card and so forth. I want to continue using the Serializer, but can I daisy-chain beefier motor drivers to the Serializer (to maintain power control) and somehow "trigger" them from the Serializer? Any suggestion on motor drivers to use? (btw, thanks for the responses...electronics is my weakness)

swabygw
05-18-2011, 08:23 AM
Any recommendations for motor drivers that interface well with the Serializer? (the higher the amps rating, the better)

tician
05-18-2011, 10:13 PM
Once upon a time I planned to build a small two wheeled robot with a pair of RS-540 (or RS-550?) directly driving ~1.8" wheels. I could not find (m)any drivers capable of the +80A required at stall, so the plan was to use two H-Bridges made from discrete components. This (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=IXTP180N085T-ND) was the N-channel MOSFET (low side) I settled on and I think this (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=SPP80P06PIN-ND) was the P-channel MOSFET (high side) (seems its max current is too low for the RS-550). Each half bridge was to be a P-channel and an N-channel bolted to this (http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=HS380-ND) heatsink. The heatsink would not only provide thermal dissipation, but also serve as a common electrical connection to a motor lead (+Vbattery->P-channel->motor lead/heatsink->N-channel->GND). Control could be accomplished with two pins per motor/H-bridge (PWM would be a nice addition on the pins, but not a necessity).

The robot was meant to be a small piece of aluminium for mounting the two RS-550 and wheels, with four heatsink/half-bridges and an ATmega168 installed in a breadboard on top of the aluminium sheet and two casters and a lipo battery strapped underneath it all. I was thinking either Sharp IR distance sensors or just contact switches around its perimeter for wall following at rather high speeds.

swabygw
05-19-2011, 10:25 AM
Yep, the amps required at stall are high, so I think your method makes sense. I was hoping to find something off-the-shelf. I'm considering a Sabertooth (like, the 2x25 or 2x50) motor controller.

swabygw
05-19-2011, 10:35 AM
Actually, I'd like to start using diodes in my project, but am unsure of their placement and sizing. From my research, it looks like I should probably be using Schottky diodes between the controller board and the motors (guessing on the positive side). The motors peak efficiency current is about 10-12 amps but the stall current is 85 amps(!) and battery power is 12v. I was hoping my local Radio Shack would have those diodes, but not exactly sure if the types they have are Schottky and/or are the right size.