Re: RX580 Gigabyte Gaming 8GB - Micron - Boots only if I do the BIOS short
Here is my theory about this error, I may be very well wrong but I'll give it a go.
Pin 1 of the BIOS is the enable signal. It is active low
I am assuming that your shorting of the pins (1-5) is somehow grounding the pin 1 therefore the chip gets enabled. Therefore the card is able to boot. I am not sure of the model of your bios chip but maybe grounding pin 1 (short to pin 4) would have the same effect? Not sure
If the above assumption is correct than it means that the Enable signal is somehow wrong.
Enable signal is generated by the GPU, goes through 3.3v and also has a pull-up resistor.
You already measured this resistor (R35) to be 1k. According to a schematic on a similar card this value should be 2.2k. I am not sure if this could be an issue. Do you see 3.3V on this resistor?
If both resistance values are ok then it would be the GPU, or the GPU connection that is the problem.
Here is my theory about this error, I may be very well wrong but I'll give it a go.
Pin 1 of the BIOS is the enable signal. It is active low
I am assuming that your shorting of the pins (1-5) is somehow grounding the pin 1 therefore the chip gets enabled. Therefore the card is able to boot. I am not sure of the model of your bios chip but maybe grounding pin 1 (short to pin 4) would have the same effect? Not sure
If the above assumption is correct than it means that the Enable signal is somehow wrong.
Enable signal is generated by the GPU, goes through 3.3v and also has a pull-up resistor.
You already measured this resistor (R35) to be 1k. According to a schematic on a similar card this value should be 2.2k. I am not sure if this could be an issue. Do you see 3.3V on this resistor?
If both resistance values are ok then it would be the GPU, or the GPU connection that is the problem.
I see exactly what you are thinking, but I am pretty sure with SPI the chip enable signal has to be correctly timed, it is part of the protocol and to prevent spurious reads/writes, you cannot just randomly pull it low and read/write the entire chip.
Of course I could be wrong, feel free to enlighten me
Here is my theory about this error, I may be very well wrong but I'll give it a go.
Pin 1 of the BIOS is the enable signal. It is active low
I am assuming that your shorting of the pins (1-5) is somehow grounding the pin 1 therefore the chip gets enabled. Therefore the card is able to boot. I am not sure of the model of your bios chip but maybe grounding pin 1 (short to pin 4) would have the same effect? Not sure
If the above assumption is correct than it means that the Enable signal is somehow wrong.
Enable signal is generated by the GPU, goes through 3.3v and also has a pull-up resistor.
You already measured this resistor (R35) to be 1k. According to a schematic on a similar card this value should be 2.2k. I am not sure if this could be an issue. Do you see 3.3V on this resistor?
If both resistance values are ok then it would be the GPU, or the GPU connection that is the problem.
I took a brake from it as it gotten me nowhere , but I will check it again later today
Here is my theory about this error, I may be very well wrong but I'll give it a go.
Pin 1 of the BIOS is the enable signal. It is active low
I am assuming that your shorting of the pins (1-5) is somehow grounding the pin 1 therefore the chip gets enabled. Therefore the card is able to boot. I am not sure of the model of your bios chip but maybe grounding pin 1 (short to pin 4) would have the same effect? Not sure
If the above assumption is correct than it means that the Enable signal is somehow wrong.
Enable signal is generated by the GPU, goes through 3.3v and also has a pull-up resistor.
You already measured this resistor (R35) to be 1k. According to a schematic on a similar card this value should be 2.2k. I am not sure if this could be an issue. Do you see 3.3V on this resistor?
If both resistance values are ok then it would be the GPU, or the GPU connection that is the problem.
Also when booting after I short the pins (pin 1+ pin 5 or pin 8- has same effect) . it looks like this in HiveOS , but when starting to mine the GPU chip it gets pretty hot but I presume there is an error somehow on a sensor reading because the chip cannot go 511 degreees celsius
I see exactly what you are thinking, but I am pretty sure with SPI the chip enable signal has to be correctly timed, it is part of the protocol and to prevent spurious reads/writes, you cannot just randomly pull it low and read/write the entire chip.
Of course I could be wrong, feel free to enlighten me
Well I have absolutely no expertise in this area, this was just my train of thought. However it is a fact that when those pins are shorted this card boots somehow. I am just curious to understand what is happening.
Might be worthwhile to check with an oscilloscope what is happening on data pins with and without the short.
Well I have absolutely no expertise in this area, this was just my train of thought. However it is a fact that when those pins are shorted this card boots somehow. I am just curious to understand what is happening.
Might be worthwhile to check with an oscilloscope what is happening on data pins with and without the short.
Also tried the osciloscope on pin 1 of the bios chip and nothing happens (without bios being in short)
Comment