Society of Robots - Robot Forum
Electronics => Electronics => Topic started by: sonictj on May 08, 2009, 02:41:55 PM
-
I'm working on a line following robot that has a PID controller for rpm control and another PID controller for following the line (inspired by the pololu 3pi robot). I've blown up two sn754410 from switching directions too quickly. When the first one blew I wasn't sure quick direction change was the issue but the second confirmed that assumption. I just learned last night that the sn754410 does not have kickback back protection diodes like the pin compatible l293d. The diodes in the sn754410 are for ESD protection (I don't know what that means).
When the motor drivers failed my code had the robot using differential drive in the forward direction to stay on the line. The robot would then spin quickly if its line sensors lost the line. The switching from forward drive to spinning too rapidly is what killed the drivers. I'm inclined to only use forward differential drive, but the problem there is the sn754410 does not have braking. This causes the robot to be not so responsive.
hardware:
motors 12-15V ~50mA no load guessing ~ 500 to 750mA stall, I didn't want to stall the motors to find out exactly
questions:
1) If I were to use a higher current h-bridge chip could I switch directions more rapidly without risking blowing anything?
2) Recommendations of any motor drivers that have braking? over current protection would be nice too.
-
Hi,
I just learned last night that the sn754410 does not have kickback back protection diodes like the pin compatible l293d. The diodes in the sn754410 are for ESD protection (I don't know what that means).
ESD = ElectroStatic Discharge
If you walk on eg. nylon carpet and then touches something like an earth connected radiator you are discharging the ElectroStatic charge that you have build up.
ESD is also the phenomenon when you rub a balloon against your clothe and it then clings to walls (or make your hair stand up)
In electronics, ESD can damage devices so that eg. a lead on a silicium chip is partly cut and perhaps days, months or years later, the lead breaks completely from overload in the affected spot.
The flyback protection that you need is to counter the inductive kick from the motor windings, as they deliver quite a high voltage the "wrong way" round and can pack quite a punch.
I'm inclined to only use forward differential drive, but the problem there is the sn754410 does not have braking. This causes the robot to be not so responsive.
hardware:
motors 12-15V ~50mA no load guessing ~ 500 to 750mA stall, I didn't want to stall the motors to find out exactly
questions:
1) If I were to use a higher current h-bridge chip could I switch directions more rapidly without risking blowing anything?
2) Recommendations of any motor drivers that have braking? over current protection would be nice too.
Just add the flyback diodes and then change your code to force a pause when going from one rotational direction to the opposite.
If you think you need more, you can put 15V..25V tranzorbs/transils in parallel with the flyback diodes but mounted as close to the motor terminals as possible (you allready put caps there, right?)
-
If you think you need more, you can put 15V..25V tranzorbs/transils in parallel with the flyback diodes but mounted as close to the motor terminals as possible (you allready put caps there, right?)
I put an electrolytic across the motor supply but not across the motors. Is this to dampen the response of the motors? How do I determine the correct value to use ? The motor caps should be ceramic right?
-
ceramic would be the right way to go, electrolytic are polarized and go boom when you change direction...
a .1uf or 1uf would do the job, preferably .1uf.
-
Hi,
[...] across the motors. Is this to dampen the response of the motors?
I'd rather say that it's to dampen the spikes that the motor lets out due to the the brushes (metal "fingers") slightly breaking contact with a low impedance motor winding every time it shifts from one winding to the next (the commutator cannot be perfect as it is mechanical) and breaking an inductive circuit results in an inductive kick.
In a brushless motor (BLDC), the commutation is purely electronic so it doesn't generate the spikes that a regular DC motor does.
How do I determine the correct value to use ? The motor caps should be ceramic right?
Usually 10nF to 100nF (0.01µF to 0.1µF) is adequate for small DC motors, the most important part is that it has to be ceramic caps, as they have the lowest ESR (Equivalent Series Resistance) and that is needed to dampen the spikes as much as possible. They have to be rated for at least 50V.
It may prove a bit hard to come by 100nF ceramic caps, but more than one of a smaller value could be used in parallel - this actually has the added benefit of lowering the overall ESR.