Hello friends.

Here's my setup. I have a regulated switching power supply that provides 12VDC @ max of 10A. I connect the LM338 to it with 120 ref. resistor and 340 ohm Adj-Gnd pins of the LM. The output voltage is well regulated and stable (5V), however when I connect my 100 watt 100Ohm adjustable resistor and play around with the resistance (going all the way down to 1 ohm), I never get more than 1.5 amps of output without high voltage drop and never get beyond 2 Amps without voltage drop of whole 4 volts. When I short I get 2.1 amps @ 1.02 volts.

How come does the data sheet state that it can supply up to 7A peak when I can't even get descent 5. I have good heat sink on the LM and ok to touch at all times.

Please help. Thanks!!

 

Did you measure the input voltage and output voltage at the LM338 pins when you had the output heavily loaded?

 

According to the datasheet, you are doing it right. No excuse for the failure from what you've told. I can only ask for a schematic of your setup and recommend more measurements to be sure the 12V supply isn't collapsing or the capacitors around the LM338 are placed real close to avoid oscillation. PS, which pkg? TO-220 or TO-3?

 

Use a scope to check VIN for magnitude. Also check for oscillations on reg.

 

Hello, thanks for your replies. I've decided to make a video of the project - maybe you'll notice if I am doing anything wrong.
http://www.youtube.com/watch?v=d3oI6RVxLnk&feature=youtu.be

Thanks much!!!

 

Did you check the input voltage under the high current conditions?

 

Did you check the input voltage under the high current conditions?

Yes, the input drops only by 0.2 volts at high currents on the output side of the regulator and the ripple is very low (<.5volts)

This is a TO-3 package.

 

There isn't a capacitor within a foot of that 338. You should be scoping the output of the 338 to see if it is oscillating because you did not follow instructions about placing capacitors VERY CLOSE to the chip.

 

Well, I was going to attach the caps to the IC after the test, but it appears that it is an essential component here. I will try to adding the caps as soon as possible.

Thanks so much!!!

 

I have inquired about having a regulator chip designed that only needs local capacitors in the finished product, but not for casual breadboarding. The manufacturers did not respond well.

 

I have inquired about having a regulator chip designed that only needs local capacitors in the finished product, but not for casual breadboarding. The manufacturers did not respond well.

Question: I am thinking that maybe I am getting fake chips!!!

The LM338 that I have cost me only $1.10 on ebay and when I checked Jameco, they cost whooping 700% more: 6.95!!

Could these chips be fake?!!

Thanks!!

 

YES. Counterfeit chips have been a problem lately.

 

Could be. You might have a LM317 instead of LM338. That would give you about the maximum current you are seeing. Ebay is definitely caveat emptor.

 

Just a few more notes: I tested one more that I have from ebay: it turns out that the chip just died after drain of 1.92A for ~5sec with all the protection diodes. Interesting to add, the chip part number on top of the chip erases with just a simple finger rub for 8 seconds.

Thanks you guys for helping me with this, your help is really appreciated. I'll stop being cheap and order from Jameco.

Any other reputable sites I can order from?

Thanks much again!