TCL 50P615 TV VGH/VGL Voltage Oscillation Issue on Integrated Mainboard – Causes and Fixes

User question

PLACA TV TCL 50P615 VGL E VGH OSCILANDO DE 30V VGH PARA 10V MENOS VGL COM O DISPLAY CONECTADO E COM O DISPLAY DESCONECTADO DA PLACA ESSA TV É UMA PLACA SÓ FONTE PRINCIPAL E T.COM TUDO EM UMA SÓ PLACA EU ACHO QUE NÃO DISPLAY

Artificial Intelligence Response

Direct answer to the question

• In a healthy TCL 50P615 the Gate-drive rails must be ≈ +26 … +32 V (VGH) and –6 … –10 V (VGL) absolutely steady.
• If these rails swing between +30 V and +10 V (or disappear) even with the display unplugged, the fault is inside the integrated PSU/T-Con board: the DC-DC (boost / charge-pump) that generates VGH / VGL is losing regulation.
• Most frequent causes: shorted ceramic capacitor on VGH/VGL, dried-out electrolytic on the boost output, open feedback resistor, or a defective multi-output DC-DC driver IC (often a MAX17xx, RT8xxx, MP26xx, TPS65xxx, etc.).
• If VGH/VGL become perfectly stable as soon as the panel FFC is removed, the LCD panel (gate-driver COF/COG) is shorted; otherwise repair the board.

Detailed problem analysis

1. Functional overview
   • The integrated mainboard contains: PFC & 12 V SMPS → local buck regulators → a TFT-specific DC-DC that produces AVDD (~15–18 V), VGH (~30 V), VGL (-7 … -10 V) and VCOM (~5–7 V).
   • The boost for VGH is usually a step-up converter followed by a charge-pump inverter for VGL. Regulation is closed via a resistor divider to the IC’s FB pin. Any over-current, over-temperature or FB error forces the IC to hiccup → you see the “oscillation”.

2. Interpretation of the user measurement
   • “Oscilando de 30 V para 10 V menos VGL” = the boost starts, VGH rises to ≈ 30 V, current surges, OCP trips, converter shuts down, voltage decays to ≈ 10 V (and VGL collapses toward 0 V); cycle repeats at ~1–10 Hz so a multimeter shows fluctuating readings.
   • You already proved it happens both with and without the panel. That rules out a panel short and points to the board itself.

3. Typical board-level culprits
   • Output filter electrolytics (4.7–22 µF/50 V) dried out → huge ripple, regulator unstable.
   • SMD MLCC on VGH or VGL shorted intermittently → over-current hiccup.
   • Schottky diode (boost) or bootstrapping diode leaking.
   • Inductor or its solder joint cracked.
   • Feedback network: divider open / drifted; IC never sees correct voltage.
   • DC-DC controller IC damaged (ESD, lightning, reverse current from panel).

4. How to isolate the failing component
   a) Power the board on the bench with the panel cable removed.
   b) Scope VIN (12 V), SW node, VGH, VGL. Confirm hiccup frequency; if VIN dips the primary SMPS is also involved.
   c) With power off, use a DMM in resistance mode between VGH and GND and VGL and GND.
   – Ω < 1 M for VGH suggests a shorted MLCC.
   – Ω ≈ kΩ-range is normal (feedback divider).
   d) Inject 5 V through a 1 kΩ resistor into the FB pin: if the IC stops oscillating the FB path is open.
   e) ESR-check electrolytics; replace any > 0.5 Ω.
   f) If suspicion remains on the IC, swap it or “dead-bug” an external boost module (+28 V / –8 V) to prove that the panel and rest of the TV are OK.

Current information and trends

• Single-board “Power + Main + T-Con” is now the norm in entry-level 4 K sets; VGH/VGL failures are a recognized weak point because cost-down forces the use of minimal-spec ceramic caps and no heat-sink on the converter IC.
• Service forums (Elektroda, BadCaps, CN tech videos) report the TCL 50P615 commonly failing in the RT6933 / TPS65132 area. Kits with ready-made ±30 V generators are sold in Asia for quick repair when OEM parts are unavailable.
• Industry trend: moving to COF driver ICs that accept a single 15 V rail, eliminating VGH/VGL from the mainboard in future designs.

Supporting explanations and details

• TFT gate needs ≈ +25 V to charge the pixel capacitor; when VGH sags to +10 V the TFT never fully turns ON, all pixels leak → uniform grey/no-image.
• The converter works like any boost:
  \[ V{GH}=V{IN}\frac{1}{1-D} \] where \(D\) is duty cycle. If load current \(I\) rises sharply ( short ), converter hits its current limit \(I_{LIM}\), turns OFF, VGH collapses, controller restarts → oscillation.

Ethical and legal aspects

• Live-chassis sets: lethal mains on the primary side; follow IEC 60950 / IEC 62368 safety, isolate the board with an isolation transformer.
• Warranty: device less than 24 months old may still fall under consumer-law warranty; opening the set voids it.

Practical guidelines (step-by-step)

1. Remove AC, discharge primary caps.
2. Visual: bulged electrolytics, burnt IC, cracked inductor.
3. Measure resistance VGH-GND, VGL-GND, AVDD-GND. Short < 30 Ω → suspect MLCC.
4. Replace the VGH output cap (often 10 µF/50 V) and the charge-pump caps (0.1 µF–1 µF/50 V). Cost < 1 USD.
5. If still oscillating, replace the DC-DC IC. Use hot-air at 350 °C, clean pads, apply new IC.
6. Reassemble, check VGH 28–32 V, VGL –7 … –10 V, AVDD 15–18 V.
7. Only then reconnect the panel and perform a 2-hour burn-in to be sure no further trip occurs.

Potential challenges

• Multi-layer board; lifted pads easy. Use Kapton and preheat.
• Unobtainable IC: fit a small daughter-board with TPS61175 for +30 V and LM2596-inverter for –8 V, feed the original rails via jump-wires.

Possible disclaimers or additional notes

• If the panel was damaged by the over/under-voltage event, artefacts (vertical lines) may remain even after board repair.
• Some boards read the panel’s NVM via I²C; if you run the board long without the panel, Android TV may log EDID errors but they are harmless.

Suggestions for further research

• Compare failure modes across TCL, Hisense, Vestel integrated boards; publish statistical MTBF of the gate-drive section.
• Investigate use of polymer caps or two-phase boost to lower ESR stress.
• Evaluate next-gen panel drivers that eliminate separate VGH/VGL.

Brief summary

Unstable VGH (30 → 10 V) and VGL on a TCL 50P615 that persist even with the LCD unplugged indicate a defective VGH/VGL generator on the all-in-one mainboard. Focus on the boost / charge-pump section: replace high-ESR electrolytics, check for MLCC shorts, verify feedback network, and if necessary swap the DC-DC controller IC. Once the rails are rock-solid at +30 V and –8 V the picture will return—confirming the panel itself is not the root cause.

Disclaimer: The responses provided by artificial intelligence (language model) may be inaccurate and misleading. Elektroda is not responsible for the accuracy, reliability, or completeness of the presented information. All responses should be verified by the user.