Intro

My PVI-4.2 solar inverter recently stopped working, so I decided to open it up, see what failed, and if possible fix it.

After a serious safety warning, I’ll walk through the teardown, explain what each part does, and then dig into what went wrong. This isn’t a repair guide or a recommendation to attempt something similar, but if you’re technically inclined and curious about how grid-tie inverters work, it should be useful.

Safety Warning (IMPORTANT)

Grid-tie inverters contain lethal voltages in many areas. Both high-voltage AC and DC are present on multiple internal rails, and disconnecting the unit does not automatically make it safe.

Large internal capacitors can retain 300–400 VDC for minutes after shutdown. There’s more than enough energy in those to kill you. These voltages are found on the MPPT input, the bulk DC bus, and the inverter H-bridge, and they’re spread across multiple PCBs.

If you haven’t worked with power electronics before, or you haven’t developed the instinct to spot something dangerous, this is not the project to start on.

This post documents what I found, how the inverter is structured, and what failed. It’s not how to repair your own unit. Read at your own risk!

Teardown Overview

The PVI-4.2 is built into an all-metal enclosure with a prominent front-facing heatsink to which the MPPT (maximum power point tracking) and inverter MOSFETs are clamped.

Access is through the rear cover, which is secured with screws. Once removed, a second set of screws allows the main PCB mounting plate to swing open on a maintenance-friendly hinge but take care to disconnect the ribbon cable before opening it fully or you risk damaging the connectors.

Internally, there are four distinct PCBs:

• Main Power Handling Board
  A dense, multi-function board that includes:

  • Input surge protection (AC and DC)
  • AC/DC filtering
  • Voltage and current sensing
  • Grid isolation relays
  • A DC-DC switch-mode power supply (SMPS)
  • A small CPU
  • External interface headers

• Power Conversion Board
  Handles the heavy lifting:

  • MPPT boost conversion
  • Bulk DC storage
  • Inverter H-bridge MOSFETs
  • Massive output chokes

• CPU/Logic Board
  Contains:

  • Multiple CPUs (one for UI, others for MPPT/inverter control)
  • 12 V to 5 V regulation
  • Ribbon connections to the power and UI boards

• Display/UI Board
  A front panel module with:

  • Multi-character LCD
  • Indicator LEDs
  • Buttons
  • Piezo buzzer

System Overview

Before diving into individual boards, here’s a high-level diagram showing how everything connects together:

Each solar input (IN1/IN2) is independently boosted via its own MPPT stage, feeding a shared bulk DC rail. That rail is then inverted into grid-synchronised AC via a full-bridge inverter, filtered, metered, and relayed to the grid. Control logic is split across dedicated CPUs and powered by an onboard switch-mode power supply (SMPS) which is fed from the solar DC input.

Main Power Handling Board

This is the central nervous system of the inverter. It handles power conditioning, metering, protection, and communication with the outside world.

• Input surge protection on both AC and DC sides
• Input filtering to reduce conducted EMI
• DC buffer caps feeding the boost stage
• Grid connection relays with anti-islanding relay. (The anti-islanding relay may be optional for some markets.)
• AC and DC current metering circuits
• AC EMI filtering
• Internal SMPS generating power for logic and measurement
• USB interface for service connections to a PC

Power Conversion Board

This is where the high-voltage work happens. The power board handles all energy conversion, from regulated DC boost to AC grid output, and carries some serious current. It’s mounted directly to the enclosure via thermal clamps to dissipate heat from the switching devices.

Key components:

• MPPT Inductors and MOSFETs
  Each solar input feeds its own MPPT boost stage. These consist of:
  • A high-current inductor (wound with thick wire)
  • A switching MOSFET, a 47NG0C3, driven by the control CPU
  • Boost output feeds through a diode (ISL9K1560G3) into the bulk capacitor rail

• Bulk DC Capacitors
  Four large electrolytic capacitors (1200 µF, 315 V) serve as the main energy reservoir. These appear to form two pairs in series contributing to a common bulk rail. These store energy between MPPT boost and inverter draw. Across each is a 120 kΩ bleed/balancing resistor which gives a time constant of about 140 seconds.

• Inverter H-Bridge IGBTs
  Four large IGBTs, configured as a full bridge to generate high-frequency AC from the bulk rail. These are controlled by PWMs from the CPU delivered through interlocked gate drivers (2ED020I12) that switch in sync with the grid and push current through the output chokes.

• Inverter Output Chokes
  Two large inductors filter the PWM signal from the inverter stage into a clean, sinusoidal AC waveform. These are directly wired to the grid-side PCB.

Control signals come in via ribbon from the CPU board. Gate drive timing and fault feedback all pass through this interface.

CPU/Control Board

The logic board is the brains of the system. It handles:

• MPPT PWM generation
• Inverter switching logic
• Safety interlocks and relay control
• User interface and external communication

Breakdown of major areas:

• UI Microcontroller
  An ATmega128 CPU manages the front-panel display, status LEDs, buttons, and buzzer. Completely isolated from the power control logic.

• Inverter and MPPT Controllers
  Two additional DSP CPUs from Texas Instruments TMS320 series, handle the core switching logic:

  • MPPT tracking and boost control
  • Inverter synchronisation and modulation
  • Relay sequencing and grid detection logic

• Power Regulation
  A small onboard linear regulator converts 12 V from the SMPS into 5 V for the logic domain.

• Interconnects

  • Ribbon to the UI/display board
  • Edge connectors to the main power board (relays, sensing, SMPS)
  • Header to the MOSFET board (PWM outputs, sensing, gating)

Notably, these CPUs have externally accessable programming or debug headers suggesting in-field firmware upgrades or diagnostics are possible (but undocumented).

UI Board

The front-panel board acts as the interface to the user / installer.

• A segmented alphanumeric LCD for status display
• Several indicator LEDs
• A few buttons for controlling the UI
• A small piezo buzzer

It connects directly to the CPU board via a short ribbon cable.

Failure Analysis

In my case, the inverter stopped booting altogether and showed a blank display.

After a quick inspection, I noticed light thermal damage on one of the grid connection relays. When I flipped the board over, the real cause became obvious: a large blob of solder had melted off the relay’s PCB pad and fallen down the board. Judging by the trail, it had briefly shorted part of the 12 V rail from the SMPS before cooling off and fusing to a trace.

It’s worth noting that these relays are switching what should be two synchronised AC waveforms. In theory, this means no arcing, just clean zero-cross switching. That’s the theory.

In practice? They melted themselves off the board.

The relays are rated AC-7a, which is for relatively benign resistive loads. For a grid-connected inverter that might be switching under non-ideal conditions or slight voltage mismatches, that rating is optimistic at best.

After identifying the damaged relay, I desoldered and removed the four AC-side relays and cleaned up the PCB. That gave me the chance to inspect the rest of the system, in particular the SMPS, which supplies 12 V logic power to the CPU and control sections. That’s where I found the next issue.

A crowbar diode across one of the 12 V rails had failed short. Most likely it sacrificed itself when the molten relay solder briefly bridged the mains and logic domains.

To test the SMPS without the diode, I applied 100 V DC to the PV input using a makeshift battery stack of five Makita tool batteries in series (don’t try this at home).

With that, the SMPS started up and produced the expected 12 V rails. Logic power is now confirmed working.

However, after a few minutes the root cause of the failure quickly became aparent as one of the IGBTs in the H-bridge started to let out magic smoke. My worry in fixing it is that a MOSFET failure often cascades to the gate drivers which are critical in an inverter. As this inverter is 13/14 years old, it’s likely it has reached the end of its natural life.

I will put a new inverter into service because I want that to be reliable; but this one may become a hacker project in which case, watch out for a part II!

Final Thoughts

This wasn’t a planned project, but digging into this inverter gave me a much deeper understanding of how modern grid-tie systems work (and when they don’t!).

Hopefully this teardown helps someone else who’s staring at a blank inverter panel and wondering whether to start unscrewing things.

If you spot any mistakes, have insights, or just want to share your own inverter horror stories, I’d love to hear them.