BPI-R4 SFP+ VCC became too high

After about 1 year of use the VCC on the SFP+ slots is now too high. It should be 3.30V but module tells me it’s 3.50V and triggered the ALARM today. Before I was wondering already for weeks now why upload became slow and XGSPON ISP sometimes disconnects me – now I know why!

So on schematics I can see that 3.3V is generated by MP8759GD stepdown part. In the datasheet of that part I read “1% Reference Voltage”, but in reality is 6% deviation by now. I’m not so much into electronics, but something must be wrong here.

I’m using the original BPI case with fan, so temperature is not the problem. Also strong and reliable 12V 5A power supply should to supply the R4, and no WiFi or M.2 module installed, just XGSPON SFP+ and DAC cable to a switch.

@simon Please provide me with instructions how to check and repair the board.

sfp1-isa-0000
Adapter: ISA adapter
VCC:           3.50 V  (crit min =  +0.00 V, min =  +0.00 V)  ALARM (MAX)
                       (max =  +0.00 V, crit max =  +0.00 V)
temperature:  +57.8°C  (low  =  +0.0°C, high =  +0.0°C)
                       (crit low =  +0.0°C, crit =  +0.0°C)
TX_power:      3.60 mW (max =   0.00 W, min =   0.00 W)
                       (lcrit =   0.00 W, crit =   0.00 W)
RX_power:      6.00 uW (max =   0.00 W, min =   0.00 W)
                       (lcrit =   0.00 W, crit =   0.00 W)
bias:         19.00 mA (crit min =  +0.00 A, min =  +0.00 A)
                       (max =  +0.00 A, crit max =  +0.00 A)
2 Likes

Yes, the 3.3V of the SFP is fixedly generated by the UP13 (MP8759GD). Do you have a way to measure this voltage? For example, with a voltmeter.

If the actual measurement confirms 3.5V, then there is a hardware problem; otherwise, the SFP module’s detection voltage is incorrect.

Is there a convenient point to measure the voltage? I cannot measure it inside the SFP cage as I lack a SFP breakout cable, unfortunately. Maybe a nearby and more exposed via or testpoint?

I also couldn’t find any corresponding test points. It should be possible to determine the power supply of the SFP by combining the device bit diagram for testing this point.

Yeah, I saw that in the diagram, but tracing it down on the physical board is hard…

I’m not at home to take a closer look and measure them, but I have a theory based on the type and size of the parts:

Sorry for mixing up SFP1 (schematics) and SFP2 (on the PCB), but the power-supply structures look nearly the same. :slightly_smiling_face:

  • If that’s correct, it should be possible to measure between R103 (or R113) and GND.
  • If my theory is wrong, sorry about the circumstances
1 Like