WEBENCH® Tools/LM5085: Pgate Issues with 20V Vin

Brett Vernor

Part Number: LM5085

Tool/software: WEBENCH® Design Tools

I designed the following circuit using Webench its a 60VDC to 5VDC buck converter and I'm looking to run it with about a 1A to 1.5A load, I built a prototype of the board using the suggested components (SQ2337ES P-ch Mosfet), but I am having an issues with voltages over 20V, originally I assumes that the LM5085 Pgate pin was regulated, but it seems to measure a voltage that is just under the Vin voltage, so in this case when I exceed the 20V gate limit the circuit shuts off. Any suggestions on how to resolve this issue, would I need to add a voltage divider, is there a better way to limit the max voltage seen at the gate of the P-ch Mosfet? I would realyy appreciate any help or advise. Thank you

over 7 years ago

0 Samuel Jaffe over 7 years ago

TI__Mastermind 25590 points

Brett,

Can you attach some scope shots of what you're seeing. PGATE threshold should be measured from Gate to Source (Gate to VIN).

-Sam

0 Brett Vernor over 7 years ago in reply to Samuel Jaffe

Prodigy 100 points

I'm running off a multimeter at the moment, I was about to go back and add to my post I posted the wrong voltages, I am now reading from pgate to Vin, as Vin approaches 20V Vgate drops, its 1.5V at 20.1V and if I climb a little higher it starts to cut off, at 20.9V it cuts off and drops to about 0.5V, which is of course below the mosfet threshold. The buck converter circuit work fine at anywhere from 6-19V and gives a steady 5V out, but I have this cut-off issue at 20V and above Vin.

0 Samuel Jaffe over 7 years ago in reply to Brett Vernor

TI__Mastermind 25590 points

Brett,

The multimeter will give a pseudo-average voltage reading. PGATE is a switching waveform. So you will need an oscilloscope to see the actual voltages and waveforms from PGATE to VIN. Do you have access to an oscilloscope?

-Sam

0 Brett Vernor over 7 years ago in reply to Samuel Jaffe

Prodigy 100 points

Mine just failed, Tektronics with the bad U880 IC, I am ordering a new one but I wont have it until next week, I understand that this is switching at a high frequency and the multimeter doesn't do much for this situation but it is what I have on hand. I was hoping to see the behavior of the circuit with it, as far as I can tell at around 20Vin the PGate drops.

0 Samuel Jaffe over 7 years ago in reply to Brett Vernor

TI__Mastermind 25590 points

Brett,

This will be difficult to debug without an oscilloscope. This issue could be caused by stability, overvoltages, or something else, all of which need a scope in order to properly diagnose. But I understand your predicament. Let's try to get to the bottom of this with a multimeter (and if you can borrow a scope from another bench, please do so).

What's the part number on your diode? 1.85V seems high. This should be between 0.4V and 0.9V and it should be a Schottky.

Did you follow the layout guidelines from the datasheet? Please share your layout.

-Sam

0 Brett Vernor over 7 years ago in reply to Samuel Jaffe

Prodigy 100 points

STPS4S200B-TR

https://www.mouser.com/datasheet/2/389/stps4s200-956754.pdf

0 Brett Vernor over 7 years ago in reply to Brett Vernor

Prodigy 100 points

I have the eagle files for the schematic and pcb layout, do you want those?

0 Samuel Jaffe over 7 years ago in reply to Brett Vernor

TI__Mastermind 25590 points

Brett,

K, that's a good diode. The schematic threw me off.

A PDF or pictures would be preferred. I don't have eagle installed at the moment. I just need to see how everything's connected.

-Sam

0 Brett Vernor over 7 years ago in reply to Samuel Jaffe

Prodigy 100 points

Ignore R2 and R3 that is just a voltage divider to give feedback of the Vin voltage, not used in this situation.

0 Samuel Jaffe over 7 years ago in reply to Brett Vernor

TI__Mastermind 25590 points

Brett,

Thanks for the updated schematic. I need the layout as well. I need to know how each component is placed on the board and how they all connect.

-Sam

0 Brett Vernor over 7 years ago in reply to Samuel Jaffe

Prodigy 100 points

0 Samuel Jaffe over 7 years ago in reply to Brett Vernor

TI__Mastermind 25590 points

Brett,

You will probably have to redo your layout to get the circuit working as you expect. In general it's best to look at the EVM for guidelines, then look at the datasheet layout guidelines for specifics. Here are some of the big things I noticed on your layout which should be fixed:

There should be a large GND plane connected to the thermal pad and to the GND pin of the IC.
Connect high-current and high di/dt connections with thick traces (or polygons or planes).
Connect high di/dt networks with as small a loop as possible
- When SW goes high you get high di/dt from CIN+ --> RSEN --> Q1 --> L --> COUT+ --> COUT- --> CIN-. Make sure that loop is as small as you can make it.
- Same deal with the loop when SW goes low.
Move C3 close to the IC. That large loop with thin traces causes unnecessary inductance which can mess things up.
Connect high-current pads with a solid connection, not thermal reliefs.
And more. See the EVM and datasheet for directions.

-Sam

Power management

Power management forum

WEBENCH® Tools/LM5085: Pgate Issues with 20V Vin