LM337-N: The Output voltage is dropped and not stabilized

Almog Azulay

Part Number: LM337-N
Other Parts Discussed in Thread: LM337, LM317

hello,

I work with this component in its normal configuration (Figure 1), and according to the resistor ratio, I should be at the output -49Volt.

when I work with the regulator stand-alone it works fine(figure 1)- I measure with DMM -49.1oltage at the output of the regulator.

figure 1:

As soon as I connected resistor and diode between to voltage (-54V, -49V) at another place in the PCB the voltage output regulator drop to ~-52V(Figure 2).

Do you have an idea why? I simulated this circuit and this situation doesn't happen. (do you have another spice model for LM337 besides the one that appears on the site)

figure 2:

regards,

Almog

over 3 years ago

0 JCHK over 3 years ago

TI__Mastermind 43630 points

HI Almong,

The first thing I see is that the device is rated for a maximum VIN-VOUT differential of -40V. What are 1GND and 2GND? What is the load between GND and the -49V rail? Adding the diode and resistor in parallel with the LM337 will help to ensure that the LM337 is not exceeding its maximum VIN-VOUT differential but as drawn, this is an additional leakage path. Depending on the load resistance, this may be enough to reduce the voltage as you are seeing.

Here is an article showing how to float the LM317: https://electronics.stackexchange.com/questions/188734/floating-lm317-used-at-high-voltage-protected-by-two-zener-diodes

For this to work, the diode above labeled 1D would need to be flipped around.

As for another model, we currently only have the once that is one the website.

I hope this helps you to find the root cause.

0 Almog Azulay over 3 years ago in reply to JCHK

Prodigy 50 points

I am sorry but i do mistake in the figure gnd1=gnd2.

The vout-vin differential is 5v.

But when i add parallel the resistor and diode the vout regulator dropt down.(at simulation the vout is stable).

0 JCHK over 3 years ago in reply to Almog Azulay

TI__Mastermind 43630 points

Hi Almong,

Ok, that is a bit clearer. In your simulation, what is the load and corresponding load current?

Also, are you simulating with the -54 powering up or is it steady state at the time of the simulation?

0 Almog Azulay over 3 years ago in reply to JCHK

Prodigy 50 points

The -54 V is power-up when the simulation starts and after that the voltage at a steady state.

I measured on parallel resistor (40ohm) 100mA (as it should be in a PCB), and the VOut regulator stays stable.

at real PCB Vout it is dropped and No power consumption.

0 JCHK over 3 years ago in reply to Almog Azulay

TI__Mastermind 43630 points

HI Almong,

I went ahead and put together a simulation and realized you need to change the 5k resistor to 4.7k, I varied the load in the simulation and it seems to reliably regulated -49V.

This checks out with equation 1 from the datasheet:

LM337.TSC.zip

Selecting R2 as 5k gives a VOUT of -52V,

I hope this helps.

Power management

Power management forum

LM337-N: The Output voltage is dropped and not stabilized