S17+ hashboard error 64 of 65 chips error

[ardeverb]

I have a problem with 2 s17+ hashboards having this error. I have medium knowledge and can repair 60% of all the s17+ i see my eyes on. This one drives me crazy because i tried everything what i know off.

I replaced  chip 64 and 65 3 times and chips 63 one time. RO signal is 1,8 volt  and the clk signal is ok with 0,9 volt. All normal values as i know.

I replaced the voltage regulator and little part around chip 65 as a last measure. No result. Anyone has a tip or advice what to do?

[BitMaxz]

What tools currently you are using to detect those faulty chips?

Have you tried to flash it first to known working hex file? If not try it first.

There is a guide on zeusBTC that might be able to help you check this link below.

- https://www.zeusbtc.com/manuals/Antminer-S17-plus-Hash-Board-Repair-Guide.asp

[ardeverb]

Thanks for your answer. I reflashed eeprom already with my tools. I worked around 5 hours on the board and have more information now.

The RO signal from chips 65 and lower is 1,8 volt and is fine. The clk signal is fine till chip nummer 61, this is the first chip in the last voltage domain. The voltage of clk is 0 v on chips 61 and 1,8 v on chip 62 0v on chips 63 1,8 v on chip 64 and i am unsure on chips 65. When the board is heating up the clk voltages are changing to 0,9 volt sometimes.I replaced al the 5 chips of the last voltage domain already but nothing helps.  I think the strange issue with the clk voltage is connected but i do not know how.

[BitMaxz]

I replaced al the 5 chips of the last voltage domain already but nothing helps.

How many times did you replace those chips?

Would you mind to tell us the exact model of the chip or the chip type? Maybe the ASIC chip you are trying to put on the s17+ hashboard has a compatibility issue?

[wndsnb]

I'd check the 1.8V supply voltages on the 2nd to last voltage domain, and the CO from chip 60. Sometimes a poor solder connection can cause a voltage drop that creates flakey things to happen between voltage domains. Anything more than a few mV drop in the 1.8 across a chip is a problem. Re-flowing the chip normally fixes those issues.