LM5050-1

Robert Steinbauer

Other Parts Discussed in Thread: LM5050-1

Dear all.

We are using two LM5050-1 to or two different supplies.

The VS-pin is connected to the out pin via a 100Ohm.

When i disconnect one of the inputs, i can still measure a voltage of about half the voltage on the other input(ored output).

Is this a correct behaviour of the chip or is there anything wrong?

This is a problem for me because I want to measure the input voltage on both inputs with a >300kOhm resistor divider. And for me this behaviour looks like there is an supply conntected.

Best regards,

Robert Steinbauer

over 8 years ago

0 Dinesh Kumar6 over 8 years ago

TI__Intellectual 2980 points

Hello Robert,
Could you please share the schematic with us to understand the problem.

Regards
Dinesh

0 Robert Steinbauer over 8 years ago in reply to Dinesh Kumar6

Prodigy 185 points

Hi.

See attached the schematic.

Regards,

Robert

0 Allan Herriman6 over 8 years ago in reply to Robert Steinbauer

Prodigy 40 points

Hi,

I have a board that's basically the same as Robert Steinbauer's schematic, but without D1 and D2.

I infer that about 78uA is flowing out the LM5050-1 IN pin, under the following conditions:

VOUT = VS = +12V
VGND = VOFF = 0V
VIN = 1.625V
GATE = open
Temperature about 25 degree C.

This is causing similar problems to Robert's.

What can I do about this? (Apart from obvious things like adding a "bleed" resistor between the input and ground.)

N.B. an earlier iteration of my design used an LTC4355 instead of a pair of LM5050-1. The LTC4355 did NOT have this problem - when I disconnected an input, the voltage on that input dropped to zero.

Thanks,
Allan

0 Brian Daugherty over 8 years ago in reply to Allan Herriman6

TI__Expert 5940 points

Allan,

We have measured similar on the bench as you observed and reviewed in detail internally and implemented as shown in the attached with several customers. The leakage path occurs when voltage is present on Vs with IN left open or GND. It occurs through the reverse comparator and is non-damaging to the device so there is no reliability concern. Other than a bleeder, recommended resolution is either removal of bias from Vs or adding a transistor on the gnd per the attached.

Brian Daugherty

LM5050-1 Vout Leakage E2E.pdf

0 Allan Herriman6 over 8 years ago in reply to Brian Daugherty

Prodigy 40 points

Thanks Brian,

That was very informative.

Would you consider adding this "Iin(reverse)" parameter to the datasheet?

It would only need to be a "typical" number, not guaranteed or tested. (We don't want to upset the test staff or increase costs. :)

Regards,

Allan

0 Brian Daugherty over 8 years ago in reply to Allan Herriman6

TI__Expert 5940 points

Allan,

We are updating our data sheets over the next several qtrs with enhanced formating and detail, The LM5050 update will include this notation and app circuit.

Thanks,

Brian

0 Alex Triano over 8 years ago in reply to Allan Herriman6

TI__Genius 12380 points

Hi Allan,

I am working on the LM5050 datasheet and have this information included in the draft. It will be online by the end of this year (probably by the end of November).

Thanks!
Alex

Power management

Power management forum

LM5050-1