Project Background: Building a 2kW induction heater for a hobby foundry, but my parallel MOSFET setup keeps failing miserably. Here’s the nightmare: 4x IRF540Ns in parallel, blowing fuses within 5 minutes of operation. Current probe shows wild imbalances—some devices pulling 3x more current than others.

1. The Setup (Barebones Breakdown)

• Topology: Half-bridge at 85kHz, driven by Arduino Nano + IR2110
• Power stage:
  ✅ 4x IRF540N (100V/33A, Rds(on) 44mΩ @ 25°C)
  ✅ 0.1Ω source resistors (for current sensing, but readings are wonky)
  ✅ Cheap Chinese heatsinks with thermal paste (maybe the problem?)
• Test gear:
  📊 Rigol DS1054Z scope (watching Vgs/Vds waveforms)
  🧰 Amprobe AM-510 clamp meter (questionable accuracy?)

2. Weird Symptoms I Can’t Explain

2.1 Current Hogging Mystery

• At 1.5kW load:
  ⚡ Device #1: 18A, Device #2: 5A, Device #3: 12A, Device #4: 7A
  🥵 Device #1 hits 110°C in 2 minutes (datasheet max 150°C, but yikes!)
• Hypothesis: Vth spread? Measured values: 2.9V, 3.2V, 3.1V, 3.4V (using this DIY Vth tester)

2.2 Layout Chaos (Pics Attached!)

• My PCB is a mess:
  ✖️ Gate traces vary from 2cm to 4cm (oops, rookie mistake)
  ✖️ Source connections star-shaped, but with different via counts
  ✔️ Did use 2oz copper for power planes (proud of that at least)
• Need your eyes: Check out my Fritzing layout (link in comments)—where’s the worst bottleneck?

2.3 Thermal Runaway Loop

• What I see happening:
  1. One MOSFET gets hotter → Rds(on) increases
  2. But current shifts to others, which then overheat
  3. Rinse, repeat, smoke...
• Question: Should I add active cooling (like a fan) or fix the root issue first?

3. Failed Fixes (Embarrassing But True)

✅ Tried matching Vth within 0.1V (better, but still imbalanced)
✅ Added 100nF ceramic caps across D-S (no change in waveforms)
✅ Swapped to 0.05Ω source resistors (made current sensing worse)
✅ Even tried hand-matching Rds(on) with a multimeter (waste of time?)

4. Desperate for Hacker-Style Solutions!

• Specific asks for the community:
  1. Best DIY way to measure switching times (no expensive gear)?
  2. Creative ideas for balancing current without complex circuits?
  3. Cheap MOSFETs that parallel better (thinking of upgrading to IRLB3034s)?
  4. How to model this in LTspice (attached my crappy attempt)?
• Bonus challenge: First person to find the layout mistake gets a 3D printed heatsink from my failed project!

5. Why This Matters (Beyond My Burned Fingers)

Hoping to share the fix on Instructables once sorted—want to help other hobbyists avoid this pain. Tag a hacker who loves power electronics problems!

Join the debug party! Drop your wildest theory or practical tip in the comments. Let’s turn this failure into a success story!

#HackadayHelp #PowerElectronics #MOSFETFail #DIYEngineering #InductionHeater