help with voltage with lm338t

[whomeno]

ok Guys
I'm not good at math equations.
I have 25v I need to get down to 16.5 volts
I'm using lm338 with a 120R form output to adjustment and 1.5k from adjustment to ground.
What do I need to change to get 16.5v.
Thanks

[whomeno]

ok so i see that is right but still getting 24.95
went to 120R and a 1K and still get 24.95 no bridges so chip must be bad

[PRR]

Short Vref to ground. Do you get 1.25V?

[whomeno]

PRR
Here is my layout. I got the lm338 from ebay, could it be possable they are not lm338? mislabeled lm317
ebay listing
https://www.ebay.com/itm/5pc-LM338T-P-LM338-5A-Voltage-Regulator-1-2-32V-Output-Replaces-LM317-USA/223160382831?ssPageName=STRK%3AMEBIDX%3AIT&_trksid=p2057872.m2749.l2649

[antonis]

Paul told you the way to verify regulator good working condition..

There is (or should be) a 1.25V voltage difference between OUT & ADJ pins..
With 120R resistor set between those 2 pins, you set the current flowing from ADJ pin to GND.. (negleting error current from itself..)
With 1k5 resistor set between ADJ & GND, you raise the voltage to ADJ pin from zero volts to a level set by the ratio of 1.25 times (1k5/120R) so you "elevate" OUT voltage to 1.25 times (1 + 1k5/120).

First of all, verify proper regulator pin-out and then short ADJ pin to GND..
(by-pass 1k5 resistor..)
You should measure 1.25V between OUT & GND..!! 

[whomeno]

First of all, verify proper regulator pin-out and then short ADJ pin to GND..
(by-pass 1k5 resistor..)
You should measure 1.25V between OUT & GND..!!

So I did this, Smoked the regulator.

[antonis]

I'm afraid you didn't exactly what I told you..

You might shorted OUT to GND..
(although it shouldn't smoke 'cause maximum IN-OUT voltage difference was not exceeded..)

[whomeno]

from looking at the front of it pin 1 is on the left side.
I am getting 25 volt on all pins before shorting out


pin 1- adjustment
pin 2- v o
pin 3- v i

[antonis]

>Included on the chip are current limit, thermal overload protection and safe area protection. All overload protection
circuitry remains fully functional even if the adjustment terminal is accidentally disconnected.
Since the regulator is "floating'' and sees only the input-to-output differential voltage, supplies of several hundred volts can be regulated as long as the maximum input to output differential is not exceeded, i.e., avoid short-circuiting the output.<

Copied from datasheet of relevant regulator..

[whomeno]

do not know what happened but I am sure I shorted pin 1 adjustment to ground.
But why would I have 25 v on all pins with 120R and 1.5K installed

[whomeno]

this is the lm338 I am using

[antonis]

Your layout seems OK to me..
(at least, for the power supply area..)

If your joints are OK too, then you have a defective regulator..

[whomeno]

this was the second one I tried.
do you have a good source for them
was running amp at 12v with out regulator and wanted to put more voltage into it,but not this much

[antonis]

this is the lm338 I am using

I can't say anything for a "ghost" Chinese company..
(neither an official side, nor any datasheet..)

[antonis]

Can't say for sure if you got counterfeit items or it's just a wrong pin-out issue..
(for the last case, it could be destroyed in case of reverse polarity connection..)

[whomeno]

it seems all datasheets for lm338 are all the same pinout

[anotherjim]

If you have a large capacitor on the regulated output, there is a concern that it can have enough charge to damage the chip if the input is shorted to ground. See the datasheet applications for the protection diode fix for this.
LM338 appears to be a higher power version of the LM317. I'd suspect that in the TO220 package, it can't actually handle that extra power without a heatsink.

[whomeno]

My version is based of of this one at guitarfx layouts
Tiny giant amp,under the amp section. Just moved ic's so I could bolt to enclosure for heat sink. It is using a 470uf  on the regulated output.
also lm338 is not grounded at heat sink
is there a way to test lm338 without putting it on the board?

[whomeno]

Just ordered 2 lm338 from mousers.

[antonis]

is there a way to test lm338 without putting it on the board?

The way Paul suggested you..
("some" DC voltage on Input, grounded Adj, open Out - you should measure 1.25V..)