L293 and clamp diodes do I need them for small motors?

[William At MyBlueRoom]

I've found many L293 schematics don't use clamping diodes. Are these completely nessary for small DC motors?

 

[Sceadwian]

Verifiy what chip you're actually using. There are two versions. L293 and L293D, the D version includes flyback AKA clamping, AKA freewheeling diodes.

 

[William At MyBlueRoom]

It's just the 16pin L293, the motors are very small 3-6VDC toy motors. 10K rpm sort of things.

 

[Nigel Goodwin]

I would suggest adding the diodes, they don't cost much - and it gives an added layer of safety.

 

[William At MyBlueRoom]

Going to order the L293D with built in clamp diodes. Once again the PCB space is limited and I don't want to use SOIC chips on a kit.

I've also got to move up to the 16F886 or 18F2525 on the Moth, as they have TWO PWM channels (the 16F690 has only one PWM but it can drive an H-Bridge directly)

 

[phalanx]

Have you considered using the 18F2431? It may be a little advanced for true beginners but it gives you the ability to control 4 half-bridges in hardware with independent duty cycles on each. This would allow for precision stepper motor control. It also has an advanced motion feedback module which allows for various forms of feedback including: quadrature encoders, velocity feedback, or A/D synchronization for other forms of feedback.

Besides all that, it has the standard 18F series features like a 2 channel CCP module, SPI, I2C, 8X8 multiplier, internal oscillator, and power managed modes.

 

[William At MyBlueRoom]

I'll look at that pic, I switched to th 18f2525 as a decent chip, but the 18f2431 might be tempting. I would love to give the hall sensors something better than rb4,5...

 

[ikalogic]

No matter how small is the motor, you have to add freewheeling diodes.

Note that small motors can be very dangerous to your cricuit when used without clamp diodes, because they have very low internal friction,so they reach very high speeds (high speed means high back e.m.f.)

another thing.. a 10 rpm motor? sounds like a geared DC motor (?)

 

[Hero999]

Actually I'm not too sure that's true, the back EMF due to rotation won't be any higher than the input voltage, in fact it'll be much lower. The only back EMF given from a motor is inductive, similar to that of a relay.

 

[philba]

A hard stop can do it. worse is a high speed reversal. use the diodes, it's just good engineering.

 

[Hero999]

Of course, I wasn't advocating not using diodes. I was just explaining the the back EMF created by rotation is different to that caused by the self inductance of the coils.

 

[ikalogic]

i have worked a lot with all kind of DC motors, and once judged a tiny 0.10A motor didn't need diodes, well, the result that the transistor could'nt turn it off, the fan would allways stay on once truned on.

Also even very small back-EMFs can cause trouble, because some driving device (usually a transistor) is not designed to block any back emf, wether it is a tiny induced current or huge one.

 

[William At MyBlueRoom]

I'll play it safe and use the L293D which has internal clamp diodes.

 

[Sceadwian]

Make sure those internal diodes meet the needs of your inductive load. A voltage spike will still occur if the inductance or rise/fall times are too sharp.

 

[William At MyBlueRoom]

It's for the Tamiya Twin Gearbox motors.