I've recently purchased my first BP6 board - it was described by seller as "rock solid, 100% OK". Unfortunately, first look at board's components revealed the truth: some of caps were bulging, and two of them showed signs of leaking as well:
Well, I fired it up with 2x433 Cellis, flashed BIOS to RV version and run Slackware on it. I've installed lm_sensors and, prepared for worst, made some burn-in tests. To my surprise, system was really rock-solid. Running 2x cpuburn-in as well as kernel compilation for hours didn't result in any crash...
The only strange thing was sensors readings. All voltages - even supplied from PSU - were fluctuating in range of several tenths of volts. So was temp readings - changes of about 2-3 degs from reading to reading.Furthermore, MB temperature in stress was above 60 deg C! But, apart from that, there were no symptoms of instability.
After few days of testing I decided to change capacitors. I bought a set of caps from my local store and - again, prepared for worst
- made some soldering. I've replaced all 24 high-capacity caps and did "EC10 fix". Result of my work is:
I soldered caps on the edge of board somewhat above the PCB - they can then be inclined to allow fitting of bigger heatsink (Spire WhisperRock IV):
Again, I did some burn-in tests.
Voltage and temperature readings are now stable. MB temp is not exceeding 45 deg C in stress - huge difference! Vtt is quite stable, BUT when system is stressed and both CPU temps exceed about 45 degrees, it begins to fall down even to 1,2V. It needs to be fixed...
Of course, system can run under 100% without any crash for several hours. I'm planning to do some 24/7 burn-in tests - for now, I can only run it for 10 hours a day...
And, a few words about caps removed from board.
I've measured it's capacitances using RLC meter. Results are:
- out of 12 pieces of 1000uF/10V caps (Jackcon), 4 were bulging significantly - they did not show ANY capacitance at all! Next 3 had capacitance in range of 200-300uF. Rest had nearly-nominal 800-900uF.
- all 12 pieces of 1500uF/6,3V (Tayeh) was not bulged. ALL of them showed nearly nominal capacitances 1300-1400uF! Perhaps that's why this board was so stable before recapping...
Poor condition of 1000uF caps resulted in overheating of one inductor (L18) on board - and, as it is placed near RT3 temp sensor, it resulted in very high MB temp readings.
That's all for now, if you have any questions - feel free to ask