I have a device that uses four 1.5V batteries. There are three voltages taken from these batteries: 6V, 3V and 0V. I'm trying to replace these batteries now with an existing transformer together with two LM317 doing some voltage regulation. I have built an tested my idea and it seems to work pretty well (if you don't count some occasional thermal shutdowns, I don't have a heat sink attached yet to my system). The load in my system is a DC motor. I understand what happens when the motor is connected to 3V and 0V pins nut my question is what happen when it is connected to 6V and 3V? Everything seems to work fine but is this breaking the 3V LM317 because current is going wrong way? Or more likely the current goes thru the resistors and I should check they don't burn?

There probably are much better ways to do this than using LM317, is there? What would you have done?

 

You are correct in worrying about the current going the "wrong way". An LM317 sources current and can only sink current through the voltage divider that sets the output voltage, and of course, sinking current would change the output voltage.

The situation in which you can sink current at the output of the LM317 is that in which a load already connected to the LM317's output (in this case, a load on 3V) already draws as much current as needs to be sinked (sank? sunk?) or more.

 

Why do you want to connect the motor between the 6V and 3V outputs?

 

Ok, I see. So when I connect my load between 6V and 3V the difference might actually be something else than 3V due this reverse current. I think I need to measure what happens to voltage in 3V pin. At least the LEDs seem to have constant amount of current in all situations, no flickering. If I remember correctly the minimum output voltage is 10mA so wouldn't my LED between 3V and 0V already satisfy this minimum current sinking? I think I measure currents also. Thank you for your answer DickCappels.

Why do you want to connect the motor between the 6V and 3V outputs?

One wire of the DC motor is always connected to 3V pin and the other is connected to 0V or 6V depending which way it should turn.

The system seems to work correctly as it is, but I was just worried I'm going to break something by misusing LM317s. If the voltage in 3V pin stays constant in all situations seems like this works and is safe, right? Well, at least after I install heat sinks and check that my resistors can handle the current going thru them

 

So, I finally had time to measure the currents and voltages while using my circuit. I was very happy to notice that voltages remained at the right level while I motors were running. While one DC motor was running the current drawn was somewhere near 300mA and when more motors were running the amount of current rose by that same 300mA per motor. As you can guess, without heat sink there was a thermal shutdown pretty quickly. Maybe my circuit works then, don't know

 

On IC1 you are showing 6v in and 6v out, is that correct? You shouldn't need it at all.. Plus a voltage regulator has to have some headroom, meaning the input voltage is higher than the output.

 

On IC1 you are showing 6v in and 6v out, is that correct? You shouldn't need it at all.. Plus a voltage regulator has to have some headroom, meaning the input voltage is higher than the output.

The input is >6V meaning that I want to be able to use any random transformer from my drawer that can provide more than 6V and enough current.