I tried changing both power stage and undervolt while measuring the power consumption, it is strange I didn't notice any change, not even 10w.
The measurement was taken directly on the wall, no software involved.

Anyone else experienced this ? Is undervolting really works ?

Running latest image of SMOS.
My cards are XFX RX580 GTX 4GB, XFX RX470 8GB ( red fans ), Sapphire RX470 Nitro+.

It's most probably a blown VRM due to overheating. Those mosfets fail very often.
If the fan desn't spin at all then this is most probably the case.

You can check with a multimeter if you have a short over +/- of the power connector, if you don't
then check if you have a short after the fuses, if you do, visually inspect the VRM mosfets, most often
the failed one is exploded or cracked. Replace the VRM mosfet ( and fuses ) or remove it, most of the card work
well with one missing VRM. Only it will consume a little more power and it may burn the other mosfets in time
due to overloading ( depend on cards )...

If you're not lucky, the other problem could be the GPU failing.. in this case you need a new GPU chip.. probably around 22.. 25usd for the RX570 chip.

Hi!

I got a S9 control board that seems to have the circuit for fan speed feedback broken, the board hangs at "Checking fans!".
After measuring and trying to find the problem I concluded that the fan PWM and fan Feedback pins goes directly the CPU ( ZYNQ ), having
those GPIO I/O failed it seems that the only solution is to replace the CPU. Do anyone else had problems with the fan feedback circuit ?

Can someone provide the correct values for the two resistors and two capacitors near the fan connector ?
From my measurements I got: R146 - 10kOhm, R148 shorted ( seems wrong, this resistor deals with speed feedback ), C200 about 2.7pf, C198 about 120pf.
It is very difficult to trace the path to the CPU since it is a multilayer board...

The board has 6 fan outputs, on my board measuring measuring from R146 to ground I got 10k, all the other outputs ( even those not populated ) are shorted to ground.
When powered I get 3.3 volts on R146, that seems to be correct since it is the PWM control of the FAN, also got 2v over the fan speed feedback.

Another helpful method would be to bypass the fan protection but it seems to be compiled into the Kernel, if anyone have a method to
do this I am really interested.

Thanks!

You have incorrect shares. In miner screen press "s" and look at incorrect shares. If you see a lot of those you overclocker too much or you have a bad bios.
I also have RX580 8GB cards and with 1120/2120, I get a steady 29.2mh/s without any incorrect shares.
You can try: 1120/2120/950 undevolt/powerstage 5

Sometimes simpleminer reboots itself, could'd find a reason because I cannot find any logs, after it comes back on sometimes one graphic card is not initialized:

[    7.332060] amdgpu 0000:06:00.0: runtime IRQ mapping not provided by arch
[    7.332270] amdgpu 0000:06:00.0: enabling device (0000 -> 0003)
[   15.728039] [drm:atom_op_jump [amdgpu]] *ERROR* atombios stuck in loop for more than 5secs aborting
[   15.728062] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing E18C (len 428, WS 8, PS 0) @ 0xE2D7
[   15.728080] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing D3DC (len 256, WS 4, PS 4) @ 0xD4DB
[   15.728097] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing B304 (len 290, WS 12, PS 4) @ 0xB3D7
[   15.728114] [drm:amdgpu_atom_execute_table_locked [amdgpu]] *ERROR* atombios stuck executing ADA4 (len 144, WS 0, PS @ 0xAE1B
[   15.728116] amdgpu 0000:06:00.0: gpu post error!
[   15.728118] amdgpu 0000:06:00.0: Fatal error during GPU init
[   15.728120] [drm] amdgpu: finishing device.
[   15.765939] amdgpu: probe of 0000:06:00.0 failed with error -22


01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
01:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
02:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
04:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
04:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
05:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
06:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
06:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
07:00.0 USB controller: ASMedia Technology Inc. Device 2142
08:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
08:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
09:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev e7)
09:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0


System is composed out of 6 x RX470 and one RX580, both 8GB cards.

I'am not a SMOS dev but I believe it should not impact the system in any way since he only makes some rest api calls to the website backend.

Yes, I did that too and it seems to be ok now.
It's strange, all the card are identical and I believe from the same lot but two of them don't overclock that much.
Now I have 4 with:
1100 core/2100 mem/900 undervolt
2 with:
1050/2050/900

Seems to be stable at least for now.

Hi,

I'am facing a problem and don't know how to approach it correctly.
I'am running simplemining with 6 x XFX RX470 8GB Hynix, there are two thing happening:

1. After about 10 minutes of mining, hashrate of GPU3 goes down from 28mh/s to about 22.. 20mh/s and stays there, after some time eventually it will crash with
the following message in dmesg:
[ 1084.587016] amdgpu 0000:07:00.0: GPU fault detected: 147 0x05708801
[ 1084.587016] amdgpu 0000:07:00.0: VM_CONTEXT1_PROTECTION_FAULT_ADDR 0x00E0983C
[ 1084.587016] amdgpu 0000:07:00.0: VM_CONTEXT1_PROTECTION_FAULT_STATUS 0x06088002

Lots of those messages.

2. Sometimes all GPU's hashrate drops to about 10mh/s and the CPU usage goes to 100% ( claymore process ), again, the same error appears into dmesg (the gpu fault).

Now, I did shut off GPU3 and identify the card that cooled down, the problem is that the dmesg errors are for PCI ID 03 not for PCI ID 04 ( card that drops in hashrate ), here is my lspci:

01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
01:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
02:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
02:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
03:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
04:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
04:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
05:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
05:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0
06:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 67df (rev cf)
06:00.1 Audio device: Advanced Micro Devices, Inc. [AMD/ATI] Device aaf0


Now, I believe claymore see GPU in order, right ? I mean, GPU0 is 01:00, GPU2 is 02:00, etc... is this correct ?
Because the card that is dropping in hashrate is GPU3 but the error in dmesg are for PCI ID: 03:00, which should be for GPU2 in claymore ( GPU0, 1, 2-- 3rd card ).

So, what card is faulty ? The one that drops in hashrate or the one that id refereed by dmesg error ?

Has anyone else encountered this type of problem ?

Thanks!