LM338: Negative Voltage Regulator of 3A or greater

Hi Mike,

We don't have a higher current negative LDO at the moment, but you should be able to leverage this application idea from the LM317 datasheet to parallel devices for more current:

Do you think you can make this work?

Hi Mike,

John is correct, there is not a negative LDO with 5A current limit.
The LM137, LM337, LM120, LM237, UA79, LM79, LM320 and LM120 all provide 1.5A of current on a negative rail.

You would need to float the positive LDO if you were to use it for a negative rail.
Most designers choose not to go this route for this amount of current.

John's technique above gives you all of the features in an LDO with the additional current.
If you want something simpler at the expense of some protection features at the higher current range, you can adapt the following circuit from the LM317 datasheet for a negative rail.  The 22 ohm resistor multiplied by the output current, sets the threshold where the darlington pair turns on, providing additional current to the output.  You will not have thermal protection of the BJT's or over current protection, but in many applications that is not needed.  If you do need those features then you will want to consider the approach John provided.

Thanks,

- Stephen

Hi Mike,

This circuit could certainly be adapted for a negative rail.
I'm pretty sure the main difference is changing the BJT's for their alternative counterpart.
So an NPN becomes a PNP, and a PNP becomes an NPN.

I've been studying these older application circuits in an effort to modernize them, and I have noticed that they still used the original part numbers from when they were designed in the 1970's or 1980's.  So if you choose to use these application circuits, you may wish to locate modern SMD equivalents, which will save you significant money and board space.

Thanks,

- Stephen

Hi Mike,

Using Stephens recommendation or mine really depends on your comfort level, personally I think the multiple regulators are better as they share current and have inherent protection in the event of a load short. Also, they help spread the heat. For this reason, multiple LM317's might be better than a single LM338. The external transistor solution may need some more thought to ensure the circuit is robust. 

I am not sure why a sharing application idea was not added to the LM337 datasheet, this came after the LM317

Hi Mike,

On the surface, this configuration seems to work.  I would definitely recommend getting EVM's of both devices to see if the control scheme works in practice. 

I think more time fine-tuning is definitely in order to ensure this works properly in the application. 

Hi Mike,

If you don't mind can you send me a note with the application requiring 3A or more on the negative rail?

We are looking at making a next-gen higher current negative LDO. 

Thanks