LM317 to LM338 circuit changes

[spacewaster]

Hi, everyone

I'm new here, and as you are about to see, my nic is pretty apt! I can put other people's designs together, but ask me to create and I will look very blankly back at you.

I've been using a LM317 circuit to get an 18V DC 1A input down to 9VDC. From what I can tell, it's a standard 317-type circuit, using a few caps between + and - "rails", a 2k4 resistor into ADJ and a 390R between ADJ and OUT to get it down to 9V.

The output needs have increased, and now the input is going to be 19V DC and 3.68A- an old laptop supply. I believe that would cause the 317 terminal distress.

Looking at the data sheets, the LM338 should cope admirably with some spare overhead.

Is it possible to reuse the principles of the 317's circuit, and how do I go about calculating the cap and resistor values?

Please be gentle...

SpaceWaster.

 

[Leftyretro]

Hi, everyone

I'm new here, and as you are about to see, my nic is pretty apt! I can put other people's designs together, but ask me to create and I will look very blankly back at you.

I've been using a LM317 circuit to get an 18V DC 1A input down to 9VDC. From what I can tell, it's a standard 317-type circuit, using a few caps between + and - "rails", a 2k4 resistor into ADJ and a 390R between ADJ and OUT to get it down to 9V.

The output needs have increased, and now the input is going to be 19V DC and 3.68A- an old laptop supply. I believe that would cause the 317 terminal distress.

Looking at the data sheets, the LM338 should cope admirably with some spare overhead.

Is it possible to reuse the principles of the 317's circuit, and how do I go about calculating the cap and resistor values?

Please be gentle...

SpaceWaster.

Well the fact that you can read other people's design and build should be enough to utilize the chip's manufacture's datasheet. It usually includes design application examples. Be sure to study the part about heat sink requirements for the power level you will be using.

Here is a link to a LM338 datasheet:

**broken link removed**

Good luck

Lefty

 

[spacewaster]

Thank you Lefty, I shall roll up my sleeves, scratch my head and probably be back with more silly questions.

Many thanks.

SpaceWaster

 

[dch222]

There should be no problem in going from 18V to 19V if the 3.68A you mention is the rating of the laptop supply. The output circuit will only take the same current as before.

If the output current has increased for some other reason you will just need to look at perhaps more heatsinking on the 317 or 338.

 

[spacewaster]

If the output current has increased for some other reason you will just need to look at perhaps more heatsinking on the 317 or 338.

Thank you for that information, dch222. I thought the 317 was only good for about 1A? That is why I started looking around. I may be wrong.

-SpaceWaster

 

[dch222]

Yes you are right the LM317 is only good for about 1.5A. I did not understand from your post that the 3.68A was the OUTPUT current.

 

[Hero999]

Is the LM338 actually capable of dissipating 36.8W?

I would consider using a switching regulator.

 

[audioguru]

The resistor between the ADJ pin and the OUT pin of an LM317, LM338 and LM350 should be 120 ohms or less. Then the other resistor should be calculated for the output voltage needed.

With the 390 ohm resistor the output voltage will rise when the regulator doesn't have a load for some of the ICs.

The first page of the datasheets show the more expensive LM117, LM138 and LM150 that can use a 240 ohm resistor.

 

[spacewaster]

Many thanks, everyone, and apologies to dch222. One of humanities biggest problems is starting the verbal discussion half way through the mental one. Why don't you know what I'm already thinking!

Just to clarify, where I'm coming from in simple terms is that I want to knock 19v down to 9v so I can run several 9V devices off the one "chain". I had thought that using a laptop power supply would have made more sense than a "wall wart" given the requirements of the laptop for a clean supply, and their availability these days.

All the best everyone! Off to start HST (Head Scratch Time).

-SpaceWaster

 

[Hero999]

Have a look at some of the Maxim chips, there are a few that will do what you want and it will probably work out cheaper as youd won't have to use a huge bulky heatsink.
**broken link removed**

 

[mneary]

You can get laptop power supplies at voltages other than 19V. I found a few that look good on https://www.mpja.com.

The 12V units would give you 3V overhead for your regulators, just make sure the description says regulated. A switching desktop supply is a plus, because it'll be smaller and lighter.

 

[spacewaster]

Cool! I just had one siting around that was 19V.

I'm moving around now exploring other possibilities so this had turned into a real learning experience. My brain hurts...

SpaceWaster

 

[Hero999]

You can also make a fairly cheap and cheerful switching regulator with the LM317.

See the datasheet for more detail.

 

[spacewaster]

Are you guys trying to break my brain even more???

Thank you everyone- I really appreciate your guidance on this one. Wow- I'm even learning things!

- SpaceWaster

 

[tab a]

The resistor between the ADJ pin and the OUT pin of an LM317, LM338 and LM350 should be 120 ohms or less.....

Is the 120 ohms value something you've calculated? I ask because in reading through several manufacturers' data sheets on the 317 I didn't see any reference other than something like "240 is often used".

Thanks.

 

[eng1]

Is the 120 ohms value something you've calculated? I ask because in reading through several manufacturers' data sheets on the 317 I didn't see any reference other than something like "240 is often used".

Thanks.

Worst case specifics show that the minimum load current to mantain regulation is 10 mA for the LM317. The voltage drop on that resistor is constant: 1.25 V (1.2 V w.c.).
1.2/120 = 10 mA.

 

[audioguru]

Is the 120 ohms value something you've calculated? I ask because in reading through several manufacturers' data sheets on the 317 I didn't see any reference other than something like "240 is often used".

Many people get away with using the wrong resistor. Each IC is different.

On the first page of the datasheet for the LM317 is a circuit with 240 ohms. The IC is the more expensive LM117, not the cheaper LM317.

The datasheet lists a max operating current of 5mA for the LM117 and 10mA for the LM317. This operating current flows through the resistor if there is no load and causes the output voltage to rise if the operating current causes a voltage higher than 1.25V in it.

Do the math: 10mA through 240 ohms = 2.4V. 10mA through 120 ohms = 1.2V.
Use an LM117 if you want to use 240 ohms.

 

[Hero999]

I wish National Semiconductor would change thier datasheet,it's probably confused designers the world over.

 

[Nigel Goodwin]

I wish National Semiconductor would change thier datasheet,it's probably confused designers the world over.

Only the incompetent and lazy ones!

 

[Hero999]

Well most people on the Internet must fit into that category; how many LM317 circuits can you see that have been designed properly?