Summary

Benq GL2460B (front marked with GL2440) monitor powering itself off after running for 5-10 seconds.

TOP267KG IC on the power supply board was overheating to > 118C and shutting itself off.

Replaced this IC with https://sg.element14.com/power-integrations/top267kg-tl/ac-dc-conv-flyback-40-to-150deg/dp/3126288 and monitor remained powered. In an open case, the IC temperature remained around 60C-61C

Background

I received a used monitor, seller claimed it was working but didn’t elaborate that it was only working for around 10 seconds before automatically turning off. Good thing it was for free $0.

Opening

Monitor was a bit tricky to open, the front bezel was clipped to the rear plastic piece. This was the main assembly method, the panel and electrics were sandwiched in between the front and rear plastic bodies. A few screws secured the rear plastic piece to the base of the monitor.

Investigation

There are only 3 boards in the monitor –

• Power supply.
• Display controller.
• Menu button board.

Power supply board contains a connector to the display controller board.

On this connector are 12V, 5V and some control lines that set the monitor panel power and brightness.

When the power fails, these voltages drop significantly (12V, 5V drops to around 2V).

Wondering if this is some kind of special low power state triggered by the power or other menu button being stuck on, I probed around the menu buttons.

All menu and power buttons were working as expected. Multiple buttons share a common line back to the display controller board. Each button is connected in series with a resistor, and connects to ground when pressed. When connected to the display driver, it forms a voltage divider. When a button is pressed a voltage divider is formed and the display controller can read the voltage. Multiple buttons pressed at the same time create different voltages so the controller can detect simultaneous button presses.

Series resistor values on 3 buttons are 3K, 13K, 333K.

Turning to the power supply board – it must be turning off due to some overcurrent, undervoltage, or temperature issue.

During the displays operating time, its panel brightness from what I could see looked as expected. The screen appeared evenly lit, with no observable regions or excessive bright nor dim. From this I assume that the current draw is within design spec.

Since the secondary voltages (12V, 5V) were present and stable while the backlight power was on, I assume that the secondary DC power circuit is OK, and turn my attention to the AC side of the power supply board.

The AC input gets rectified to around 334VDC and stored with a 100uF 450V cap. It looked OK (as did all caps on the power board), no leaks, no swelling observed. I didn’t go as far as verifying the capacitance matched the labelled value.

Only 1 control IC is in the primary circuit. It’s labelled as a TOP267KG.

A web search shows it’s available from Element14 https://sg.element14.com/power-integrations/top267kg-tl/ac-dc-conv-flyback-40-to-150deg/dp/3126288?st=Top267kg

It’s a PWM controller for use on power supplies (duh).

Investigating if this could be faulty and shutting down, I started with the thermal image camera.

This showed that the body temperature was reaching around 120C while the panel backlight was powered on. When the backlight power turned off (power supply shut down), this IC started cooling down.

A power remove/reconnect was needed to get the backlight to work again, as long as that IC had cooled to below around 50C.

Looking at the datasheet shows that thermal shut-down occurs at temperatures between 135C to 150C. This seems to match my observations.

It also has a thermal hysteresis of 75C, which also seems to match observations.

Replacing this IC was a little tricky. It is directly connected to a large copper area which acts as a heatsink, and an exposed pad on the underside of the IC that is connected to the board also.

Trying to desolder with the hot air gin caused the nearby copper to bubble.

I cut all the legs with a scalpel, allowing g the IC body to be removed. Then I was able to remove each of the still-soldered legs with a soldering iron.

Clean away the old solder, and install the new IC. Because the pad under the IC isn’t connected to anything, I put extra solder on the exposed available pins to better connect them to the copper plane heatsink. More solder mass should allow more heat to be transferred (that’s my theory, anyway).

Power on the repaired monitor, and it seems to work. The replaced IC holds steady at around 60-61C.

A thermal imaging camera here was extremely helpful. It lets the faulty IC stand out like a red warning beacon, demanding attention. Without the camera, I feel this repair would have taken me much longer.