LM337 drop in replacement

[conmanxx]

Hello,

We use the LM337EMP/NOPB to provide a negative supply line for an analog circuit.
This is the SMT SOT23 version.
Used to produce a -12V from a -15V supply.


Am finding them very difficult to source, those who do have them are charging crazy prices.

Is there a drop in replacement?
Or is there something that'll fit the design with very little mods

Thank you.

 

[barry]

Are you sure you don't mean LM337IMP? Are you sure you don’t mean SOT223?

 

[conmanxx]

Yes, sorry should be SOT-223.

LM337IMP/NOPB or LM337EMP/NOPB doesn't matter too much, both are very difficult to source

​

 

[BradtheRad]

Simple voltage regulator. Load connects to 0V ground.
Transistor is PNP type because a negative supply is involved.
Adjust resistor value to suit maximum expected current load.

Or, is your 15v supply stable and unchanging? Then you might make do with dropping 3V via 5 series diodes.

 

[Easy peasy]

the above is a pretty good solution, if you don't need absolute 12.00 volts, also likely cheaper in volume ....

 

[barry]

Simple voltage regulator. Load connects to 0V ground.
Transistor is PNP type because a negative supply is involved.
Adjust resistor value to suit maximum expected current load.

Or, is your 15v supply stable and unchanging? Then you might make do with dropping 3V via 5 series diodes.

View attachment 178073

1) Not a drop-in replacement.
2) Not very little mods.

 

[KlausST]

Hi,

a 79L12 in SOT-89 comes into my mind. But also not compatible. No idea about availability and price.

Klaus

 

[danadakk]

C

Simple voltage regulator. Load connects to 0V ground.
Transistor is PNP type because a negative supply is involved.
Adjust resistor value to suit maximum expected current load.

Or, is your 15v supply stable and unchanging? Then you might make do with dropping 3V via 5 series diodes.

View attachment 178073

Curiosity got the best of me, looked at regulation and power :

Regards, Dana.

 

[KlausST]

Hi,

what worries me: is the ridiculous small diode current in the nanoamperes range.

usully zener diodes don´t work good in this range. Maybe there fore the poor voltage regulation.

It would be helpful if the OP gives
* expected input voltage range
* expected output current range
* expected output voltage range

Klaus

 

[danadakk]

Very odd as you point out, not until PNP approaches sat, where beta falls off a cliff,
does Iz pickup.

Regards, Dana.

 

[KlausST]

Hi,

I had a look at the 1N4744 datasheet. It says it´s knee voltage is 15V.
And due to the voltage drop on R2 (caused by base current) the zener will never see 15V ... thus never will show significant current.

Brad used a 12.7V zener.

Klaus

 

[danadakk]

Here is 13V Zener confirms your observation.

As a regulator for 10% load change not too bad. Of course thats not T and V considered.


Regards, Dana.

 

[Easy peasy]

it's easy enough to use a Darlington and set the zener current to 1mA, and use a 1mA ( rated Vz ) type zener, 5% tolerance will be fine .... at absolute no load the voltage will creep up a bit ( ~ 1V ) but most systems can handle that sort of thing.

--- Updated Aug 21, 2022 ---

We assume the load will be in the 0 - 75mA region as the original part was SOT-223 which has dissipation limitations from 15 to 12 V, i.e. 3W / amp or 3mW / mA, so 50mA load => 150mW dissiaption, which is getting up there for a 1" square pad on a 1 Oz board ....