search

xinchejian / SwarmRobots

Xinchejian Swarm Robot project
4 stars 3 forks source link

L293D motor controller is "expensive" #3

Open spanner888 opened 11 years ago

spanner888 commented 11 years ago

From Edward - suggestion to look at using discrete transistors or FETs to reduce the costs.

spanner888 commented 11 years ago

The L293D has built in protection diodes - can we safely do without these or cheaply add the 8 (4 if only drive forwards) diodes as well?

nihaopaul commented 11 years ago

maybe look at a modern chip.. l293d = 0.6amp continuous current SN754410 is 1amp continuous current, both are pin identical, the SN754410 datasheet http://www.sparkfun.com/datasheets/IC/SN754410.pdf

lumi3005 commented 11 years ago

I am sure we can go without the protection diodes since these gear motors are using a worm gear if I rmember right and there is no way to soin them by hand to create a current backflow.

spanner888 commented 11 years ago

"we can go without the protection diodes" - are we reading the same datasheet? The one linked does show reverse protection diodes. http://www.sparkfun.com/datasheets/IC/SN754410.pdf

"current backflow" - is always going on with DC brush motors - such as those used on SwarmRobots - as the motor spins, the brushes and armature turn on/off different coils - hence the voltage spikes from the inductance of the coils and the need to protect.

On 28/11/12 14:55, Lutz wrote:

I am sure we can go without the protection diodes since these gear motors are using a worm gear if I rmember right and there is no way to soin them by hand to create a current backflow.

— Reply to this email directly or view it on GitHub https://github.com/xinchejian/SwarmRobots/issues/3#issuecomment-10792628.

lumi3005 commented 11 years ago

I was referring to Edwards suggestion using FET's or transistors.

Backflow diodes are mostly used to protect the circuit from the current which is created by spinning the motors manually or by breaking without power supply (rolling) With a worm gear there is no harmful current going back.

~Lutz

From: spanner888 notifications@github.com To: xinchejian/SwarmRobots SwarmRobots@noreply.github.com Cc: Lutz lumi3005@yahoo.com Sent: Wednesday, November 28, 2012 3:37 PM Subject: Re: [SwarmRobots] L293D motor controller is "expensive" (#3)

"we can go without the protection diodes" - are we reading the same datasheet? The one linked does show reverse protection diodes. http://www.sparkfun.com/datasheets/IC/SN754410.pdf

"current backflow" - is always going on with DC brush motors - such as those used on SwarmRobots - as the motor spins, the brushes and armature turn on/off different coils - hence the voltage spikes from the inductance of the coils and the need to protect.

On 28/11/12 14:55, Lutz wrote:

I am sure we can go without the protection diodes since these gear motors are using a worm gear if I rmember right and there is no way to soin them by hand to create a current backflow.

— Reply to this email directly or view it on GitHub https://github.com/xinchejian/SwarmRobots/issues/3#issuecomment-10792628.

— Reply to this email directly or view it on GitHub.

kirk1978 commented 11 years ago

here's my 2 cents thought about backward diode actually, backward diode are not just potecting the current generated by the free running motor. but also protecting the current generated by the energy stored in the inductor (the motor coil) what we are looking at is a large inductor RLC system. which when running, there are certain amount of energy stored in the L. and it will get released when the power got shut off. so a backward diode is always suggested. (same things go for mechanical relay) however, whether is a "must" is depends on how much the FET/BJT can handle the dangerous (to the transistor) occur when the FET/BJT junction is completely shut off. and the Drain-source resistance (or Collector/Emitter resistance in BJT) are very high. since Inductor is a "current" device, the small amount of current multiplied by the large resistance. the "I_R" could become a large instantaneous voltage, and have a possilbity to break the junction. (if I_R is alot bigger than the D/S or C/E breakdown voltage) this also depends on how fast the transistor can shut off, as well as how fast the Gate (or Base) get pull down (or UP)

another way to help is to add a AC cap, where it will damp the spiking current.

-Kirk

On Wed, Nov 28, 2012 at 4:20 PM, Lutz notifications@github.com wrote:

I was referring to Edwards suggestion using FET's or transistors.

Backflow diodes are mostly used to protect the circuit from the current which is created by spinning the motors manually or by breaking without power supply (rolling) With a worm gear there is no harmful current going back.

~Lutz

From: spanner888 notifications@github.com To: xinchejian/SwarmRobots SwarmRobots@noreply.github.com Cc: Lutz lumi3005@yahoo.com Sent: Wednesday, November 28, 2012 3:37 PM Subject: Re: [SwarmRobots] L293D motor controller is "expensive" (#3)

"we can go without the protection diodes" - are we reading the same datasheet? The one linked does show reverse protection diodes. http://www.sparkfun.com/datasheets/IC/SN754410.pdf

"current backflow" - is always going on with DC brush motors - such as those used on SwarmRobots - as the motor spins, the brushes and armature turn on/off different coils - hence the voltage spikes from the inductance of the coils and the need to protect.

On 28/11/12 14:55, Lutz wrote:

I am sure we can go without the protection diodes since these gear motors are using a worm gear if I rmember right and there is no way to soin them by hand to create a current backflow.

— Reply to this email directly or view it on GitHub < https://github.com/xinchejian/SwarmRobots/issues/3#issuecomment-10792628>.

— Reply to this email directly or view it on GitHub.

— Reply to this email directly or view it on GitHubhttps://github.com/xinchejian/SwarmRobots/issues/3#issuecomment-10794124.