Anyway as I started to use it after two months. It\'s was a situation without positive issue. I wouldn't be able to RMA it anyway. If I had schematics and part specifications, or only part specification, I could try to repair it. But in this case...
(11-30-2019, 10:15 AM)Nikolay_Po Wrote: [ -> ] (11-30-2019, 12:43 AM)diskers Wrote: [ -> ]PS. From the beginning this device made a slight buzzling sound when powered and up.
The buzzing is the sigh of problems in DC/DC voltage converter. Either defective or inappropriate (saturating) inductor or faulty feedback compensation circuitry. Possible causes of that in my opinion are (decreasing probability):
- Physical damage to DC/DC converter components, ferrite inductor or ceramic capacitor. Most probably the board was accidentally dropped and struck the floor while handling.
- Component damage at an assembly factory, while mounting and soldering.
- Incorrect values/part numbers of components (design mistake) or assembly mistake at a factory.
Usually, operating at a hundreds of kHz or at MHz frequencies DC supply components of PCB rarely producing an acoustic (hearable) noise with normal design. Mine boards still working without of a noise. Now with case it is evident that the noise from working SBC PCB is a sign of future failure (IMHO).
I agree with
... buzzing is a sign of a potential problem... either imminent or in the future. I would have been checking the inductors to see if any were cracked, or if pressure on any of them stopped it, and then given it a dose of epoxy to deaden the vibration or replaced it as the case may be.
If you're still interested
in the schematic, here it is from the
wiki information page for the rockpro64. Datasheets for all the major components are there also.
The fact that it powered down, and then stopped working actually sounds promising... sounds like something just isn't starting up again properly. The two LEDs(white/red) are both on the 3v3 rail as they are directly controlled by the RK3399, so unfortunately won't work unless the main CPU boots (and 3v3 present!), so don't tell you much more than that. Anyway, page 11 is a good place to start, and from the looks of it, if you check the RPi Header 5v and GND pins, that should always be present if the rockpro64 has power - looks to be the 'hardwired' system 5v, so check that first. If it's present, carry on to page 18, as it's probably one of the PMIC outputs that's gone.
Thanks, I will try to revive this RockPro. As you know Christmas and festivities season are around the corner and adding to all this work, it gives not much time for
other things. Anyway I will keep you informed how the situation is evolving.
Hi everyone,
I did not touched anything form last time, but I hope that next week I will be able to spend some time with the faulty device.
Glad to hear it... Fingers crossed!
(01-09-2020, 03:51 PM)diskers Wrote: [ -> ]I did not touched anything form last time, but I hope that next week I will be able to spend some time with the faulty device.
Just for record. Yesterday I started second server based on RockPro64. And I'm hearing audible noise. My first server is still working right in front of me and I'm not hearing any sound. But second server produces noticeable amount at 0.3m distance. At 0.5m distance the noise is not noticeable.
Tried to touch suspicious components. And it seems to me the nose is going not from inductors but from ceramic capacitors. C2005 or C2012, which is closer to power jack. And the noise is depending on CPU mode of operation.
Guessing the silent server Linux distributive keeps CPU cores on and at single speed. A distributive with more aggressive power saving which is changing CPU core speed dynamically may produce more noise due to larger change of supply currents. But this is just my thoughts, have not checked yet.
(03-09-2020, 12:45 PM)Nikolay_Po Wrote: [ -> ]Just for record. Yesterday I started second server based on RockPro64. And I'm hearing audible noise. My first server is still working right in front of me and I'm not hearing any sound. But second server produces noticeable amount at 0.3m distance. At 0.5m distance the noise is not noticeable.
Tried to touch suspicious components. And it seems to me the nose is going not from inductors but from ceramic capacitors. C2005 or C2012, which is closer to power jack. And the noise is depending on CPU mode of operation.
Guessing the silent server Linux distributive keeps CPU cores on and at single speed. A distributive with more aggressive power saving which is changing CPU core speed dynamically may produce more noise due to larger change of supply currents. But this is just my thoughts, have not checked yet.
Strange that capacitor make noise, normally should be inductor.
(03-10-2020, 12:28 PM)Nikolay_Po Wrote: [ -> ]Yeah, I thought the same before the usage off high capacitance ceramic capacitors become common.
Wow... thanks for that. So not only can these stupid ceramic caps fail due to heat stress, act like a short, and cause some catastrophic failures or smoke, but they can wine a log... great, just great!
