How to Reduce Battery Drain in Atmega32 Line Tracer with L297 L298 Stepper Motors?

Hi All,
I have a line tracer which is implemented using 2 stepper motors and a Atmega 32 micro.I am using Tsop sensors using 555 timer to detect the white line. The motors are driven using a L297 and L298(H bridge) pair. The problem with the machine is that its batteries are draining out too quickly. The capacity of batteries is 12V and 1.8A. I guess the batteries are quite strong. Can anyone suggest me a method on how to reduce the current consumption using tricks and hardware or software. I had also used PWM but thats not helping much. Any help is appreciated.

Can you include a screenshot or pdf of the schematic? Sounds like a fun little problem to tackle...

Hi,
It is a bit difficult to upload the schematic but it is very similar to what I have attached.

Vinit, I don't see any attachments? You should be able to upload an image or pdf via the form.

Joe,
Oops sorry dont know what is the problem wiht the attachment but the schematic is exactly similar to
http://www.roboticsguy.com/images/misc/l297-298%20schematic.png

In case the previous link is not posted properly pls see this one
http://www.dutchforce.com/~eforum/index.php?showtopic=35468&st=0&pid=313924&#e;ntry313924;

Thanks Vinit,

I've attached your image so we have it for reference .

Here's where I'd start. Connect a DMM to your main set of batteries and measure the supply current of various operating modes. Measure the current when the line-tracer is sitting "idle" and when it's moving (i.e. motors spinning).

You should try to isolate the power of various sections in your circuit (logic board, motor controller board, etc) and then attempt to reduce each section.

Once you provide us some baseline numbers for each section we can help you find ways to reduce the power.

Thanks,
I actually had done that but forgot to mention.
I measured the current when motors were spinning which was fluctuating between 120-150mA

Well that's certainly reasonable. Is this the total system power (measured at the batteries?)

Can you clarify your battery specification. Are you saying it's 12V with a 1.8 amp-hour rating?

If this is the case the batteries should last 12 hours with a 150mA load (1800/150).

Yes its total system power.
It is 1.8 amp-hour rating.
Just I would like to add one more thing. The current which I mentioned in previous case was when motors were running freely when the robot actually runs on ground the current is increasing to 1.0A-1.4A

Ok makes sense now, thanks.

So yeah looks like when loaded, your battery life is reduced by close to 10x. So you're only going to see between 1-2 hours of battery life. Is this what you're seeing now?

What's you're target goal for a battery life of the device?

Thats what my calculation turned out to be but the batteries are draining in hardly 15 minutes after full recharge

You could verify the integrity of the battery (independent from the circuit).

Disconnect the battery from your device and connect in a purely resistive fixed load. In this case connect a 12-ohm resistor with a power rating of greater than 12 watts. If you don't have a load that large you can back off accordingly but it's best to test the battery at your operating current for closest conditions.

Then measure the voltage of your battery every 5 minutes or so over the course of the next hour. Compare that against the battery specification datasheet and see if the discharge curve is similar.

Do you have the specs sheet for the battery (make, model)?

OK thanks I ll try that and let you know

It's not surprising that the system takes a lot more power when driving than when unloaded. At least we know all the power is going into the mechanical drive, and there isn't something unexpected going on in the rest of the circuit drawing more power than it's supposed to.

This also points out a issue with stepper motors. They are good for precise positioning, but are horribly inefficient in terms of mechanical power out to electrical power in. This is usually not helped by the fact that stepper motor drivers are generally designed to do fixed steps instead of continuous motion. Continuous motion is to them just a bunch of fixed steps at the maximum rate the motor is supposed to be able to handle.

Doing fixed steps means the mechanical momentum is lost between steps. Some of the energy is recovered thru the catch diodes, but this a very inefficient process. You basically spend the power to get the motor up to speed each step. At higher speeds it doesn't actually come to a stop, but there is still some of this going on.

Doing fixed steps also usually means driving each coil with a square wave at the step rate. This is fine for advancing between individual steps with good holding torque in between, but it is again inefficient for power transfer.

You can get a bit more efficiency out of a stepper motor by driving each winding with a sine. Some off the shelf chips do this when they feature "micro stepping". Otherwise you can use something like a Microchip dsPIC with motor control PWM outputs to synthesize the average sine drive from lots of much higher frequency pulses.

However, if efficiency matters then you don't want to use stepper motors in the first place. Some kind of servo motor with position feedback will be more appropriate.

Motor driving is a complex subject all on its own, but well worth delving into if you're a EE student.