My Power Supply Saga

[alfa88]

Am I unclear on a concept or what? I'm trying to put together an adjustable 1-15V power supply. I tried a lm317 with a Voltage follower. Didn't work out so I'm going with a LM338. The most I can get out of it is 2A. I made sure the bottleneck wasn't the transformer to bridge section by loading just that part down. My current meter was able to exceed 6A and the transformer was still cool. The bridge got a bit warm. As a load I'm using 1 Ohm, 10W resistors connected in series and I'm picking off taps as I test. I realize that the resistors are under rated but I'm only leaving them connected long enough to get readings. I have the regulator on a nice big heat sink and it has never gotten beyond warm. What gives? This regulator is supposed to give me 5A.

 

[jackorocko]

As a load I'm using 1 Ohm, 10W resistors connected in series

How many?

Looking at your datasheet, it states that
Icl = Vin - Vout < or = to 10V 5A min
Icl = Vin - Vout = 30V 1A

 

[alfa88]

I have a string of 10 resistors for a total of 10 Ohms. 3 Ohms is as low as I go.

 

[(*steve*)]

Read your input voltage and output voltage across a known resistance (tell us that too) and let's see what's going on.

Also the input voltage when there is no load.

 

[gorgon]

What's your input voltage with and without load?
What's your output voltage with and without load? (over the load)
How do you measure the current in the load?
Where do the voltage drop?

TOK

 

[alfa88]

No Load:
In - 24-25Vdc
Out - 22.4 to 1.5Vdc

10 Ohm load:
In - 21.7-23.6Vdc
Out - [email protected] to [email protected]

5 Ohm load:
In - 21.4-23Vdc
Out - [email protected] to [email protected]

3 Ohm load:
In - 21.6-23Vdc
Out - [email protected] to [email protected]

 

[(*steve*)]

It would probably help if you wired up the LM338 correctly.

Looks like a simple mistake

 

[alfa88]

Well I finally broke the 4Amp barrier. and no it wasn't a miswire. I had opted for the TO-3 package and went bargain hunting on EBay. It seems that I got a bad batch. 4 baddies in a row. Fortunately I bought some TO-220 types at the same time. They're not giving me the full 5A but at least now I have a bit of dignity intact.

 

[(*steve*)]

So you're using a different circuit than the one posted above?

 

[alfa88]

Uh?! The bad part was a LM338K(TO-3), the good part was a LM338T(TO220). The application is the same. I like the heat dissipation of the TO-3 and besides, my heat sink is drilled for a TO-3.. The seller refunded me so I'll shop around again.

 

[davenn]

Alfa !

compare your circuit to the one steve linked to

you will see your circuit is incorrect!!
you need to sort out the problems, the reason you have killed LM338's probably has much more to do with your incorrect cct than with faulty reg chips

Dave

 

[BobK]

I think his circuit would try to make ground = 1.25V, which means it would just open up the pass transistor fully.

Bob

 

[alfa88]

Are you referring to R1? In all cases I'm using a 150Ohm from Output to Adjust. I corrected my simplified drawing to reflect...

 

[(*steve*)]

OK, that clears things up considerably

edit: My guess is that you have some LM317's (or similar) rebadged as LM338's

Having said that, I've not used LM338's and I don't know if they vary their maximum current based on input/output voltage differential. I've also not deliberately overloaded a 317 to see what it does in high current regimes, so take my comment with a grain of salt.