Hello ! Thanks for your awnser..
I changed my LDO to try and the thing is identical, when I plug the multimeter I have 1.80 then a micro fall to 1.599 then 1.780 and remains stabilized there and I think that is what puts me in fault RI, I imagine that it comes from a resistor yet they are all functional. At least they show a value lower than the others but proportional to 1.72 and not 1.8 I do not understand too much
In the video I test a track next to it which is good and the LDO of the defective track
https://youtu.be/ZvnSXT-e7l4
Topic: S9 tmp451 error i2c (Read 160 times)
On this picture you have in grreen area CLK0.9 & RI 1.8 and in RED CLK 0.9 and RI at 0v
https://i.ibb.co/xM9Zskr/unnamed.jpg
On the S9 hash board, when the RI signal crosses the voltage domain, it needs to pass through the resistor before being transmitted to the test point and then to the chip. Have you checked the LDO of the 11th voltage domain and the surrounding resistors, if the resistor is damaged, the RI signal will not be transmitted to the next voltage domain.
I understand, not easy without the PCB board
Thanks anyway, I tested all the components (fig 19) they are all ok. I have to buy the pick3 to flash in HEX, I don't have this item yet !
And if ever I would do a replacement of the sensor !
Thanks anyway, I tested all the components (fig 19) they are all ok. I have to buy the pick3 to flash in HEX, I don't have this item yet !
And if ever I would do a replacement of the sensor !
I sent the card in question to a friend so I don't have any logs, but I update as soon as I have a new error (I have the same one on another card)
Also on another card, I have all the chips in CLK at 0.9 but the RI signal breaks level 26 25 24. And is at 0 until chip 00.
I tested the LDO, they are well at 1.8v and functional. The capacitor 330u 2.5v also functional.
Where can this come from according to you ?
On this picture you have in grreen area CLK0.9 & RI 1.8 and in RED CLK 0.9 and RI at 0v
Also on another card, I have all the chips in CLK at 0.9 but the RI signal breaks level 26 25 24. And is at 0 until chip 00.
I tested the LDO, they are well at 1.8v and functional. The capacitor 330u 2.5v also functional.
Where can this come from according to you ?
On this picture you have in grreen area CLK0.9 & RI 1.8 and in RED CLK 0.9 and RI at 0v
Sorry but it's hard to troubleshoot without the hashboard on hand.
And those signals you talking about are for ASIC chip not for sensor temp.
And also, I just want to you to try, and maybe it's only a software issue like I said above try to flash the unit with stock firmware from Bitmain and then try to flash the hashboard with a hex file that came from working hashboard.
If you don't know how to flash the hashboard check this link below.
- https://bitcointalksearch.org/topic/repair-a-t9-after-a-bad-firmware-lost-chain-5032987
And what I'd like you to check is the temp sensor checking them visually is not enough you need a working temp sensor if you have extra hashboard you can desolder the temp sensor to replace them. This is the way how I repair replacing the sensor with a known working one is my way to test them as according to your logs it's pointing to your temp sensor.
Please read figure 19 for the schematic and how to check them I already post the zeusbtc link above.
I sent the card in question to a friend so I don't have any logs, but I update as soon as I have a new error (I have the same one on another card)
Also on another card, I have all the chips in CLK at 0.9 but the RI signal breaks level 26 25 24. And is at 0 until chip 00.
I tested the LDO, they are well at 1.8v and functional. The capacitor 330u 2.5v also functional.
Where can this come from according to you ?
On this picture you have in grreen area CLK0.9 & RI 1.8 and in RED CLK 0.9 and RI at 0v
https://i.ibb.co/xM9Zskr/unnamed.jpg
Also on another card, I have all the chips in CLK at 0.9 but the RI signal breaks level 26 25 24. And is at 0 until chip 00.
I tested the LDO, they are well at 1.8v and functional. The capacitor 330u 2.5v also functional.
Where can this come from according to you ?
On this picture you have in grreen area CLK0.9 & RI 1.8 and in RED CLK 0.9 and RI at 0v
https://i.ibb.co/xM9Zskr/unnamed.jpg
Hello again,
Sorry for the delay in answering, I was busy.
I tested all these capacitors and they output the right voltage. I tried all the chips and they are working (RI 1.8 and CLK 0.9) I can't find any burnt trace. I really don't understand where this is coming from.
Sorry for the delay in answering, I was busy.
I tested all these capacitors and they output the right voltage. I tried all the chips and they are working (RI 1.8 and CLK 0.9) I can't find any burnt trace. I really don't understand where this is coming from.
I'm talking about the resistance of those capacitors near the temp sensor voltage output is different.
Would you mind to reflash your unit to stock firmware and letting it run for a while and copy the kernel logs and post it here?
If you don't have knowledge about repairing this I suggest try to contact zeusbtc and maybe they can help you with your issue
Hello again,
Sorry for the delay in answering, I was busy.
I tested all these capacitors and they output the right voltage. I tried all the chips and they are working (RI 1.8 and CLK 0.9) I can't find any burnt trace. I really don't understand where this is coming from.
Sorry for the delay in answering, I was busy.
I tested all these capacitors and they output the right voltage. I tried all the chips and they are working (RI 1.8 and CLK 0.9) I can't find any burnt trace. I really don't understand where this is coming from.
That is not the capacitors that im talking about the ceramic one brown color it looks like this below.
So the previous owner did broke the hashboard by leaving it without a fan that can cause burn parts.
Also check the guides here https://www.zeusbtc.com/manuals/Antminer-S9-Hash-Board-Repair-Guide.asp
So the previous owner did broke the hashboard by leaving it without a fan that can cause burn parts.
Also check the guides here https://www.zeusbtc.com/manuals/Antminer-S9-Hash-Board-Repair-Guide.asp
Hello BitMaxz,
You are right ! I test all capacitors and this one (2R5TPE) close to chip 61 show me 0v in output :
https://ibb.co/cQydy1n
EDIT: I tested the resistance of the capacitor in question, it is good. I think of a short circuit on the whole track 21 with the chips 60 61 62.
Also, I put a video attached of the red diode that flashes once the sensor is triggered (a new hashboard starts to make the same worry of sensor tmp451).
The previous owner confessed to me that he overheated without a fan (which he forgot to plug in again) before disconnecting it (like 5 minutes)
https://streamable.com/qgmfjs
You are right ! I test all capacitors and this one (2R5TPE) close to chip 61 show me 0v in output :
https://ibb.co/cQydy1n
EDIT: I tested the resistance of the capacitor in question, it is good. I think of a short circuit on the whole track 21 with the chips 60 61 62.
Also, I put a video attached of the red diode that flashes once the sensor is triggered (a new hashboard starts to make the same worry of sensor tmp451).
The previous owner confessed to me that he overheated without a fan (which he forgot to plug in again) before disconnecting it (like 5 minutes)
https://streamable.com/qgmfjs
It might be capacitors near the temp sensor would you mind trying it to check if shorted or not? You can also find the shorted part by touching those capacitors while the miner is set to sleep mode.
Or if not you don't have another choice but to replace the broken temp sensor or bypass the temp sensor scan but the problem if you bypass the temp sensor it may damage other components.
Or if not you don't have another choice but to replace the broken temp sensor or bypass the temp sensor scan but the problem if you bypass the temp sensor it may damage other components.
Hi everyone, I'm on a s9 13th and a card shows me signs of a hashrate drop after 5 minutes of mining (OS braiins ). I look at the logs everything is top, starts correctly the only concern being this line which appears before the fall of hashrate:
I then looked more closely with a magnifying glass by desoldering the heat sinks, it seems okay to me on tmp541 chip, no dust, the welds are not cracked… a friend told me that as for the s17 the welds crack over time and make false contacts.
I noticed that on the hashboard when the sensor goes wrong, the little red light flashes more slowly like a bad contact (if it helps I say so)
also in the line of code I saw write Chip 61, so I decided to test the chip as well as its surroundings and I came across the correct values of CLK: 0.9 and RI 1.8 in 12v plug.
from what I understood it is, the more its heating (so after 5min to heat up) the closer we get to a crash point because of the sensor (or other)
for the moment I am there without too much additional info… an idea maybe?
Big Thanks 🫡
Code:
2022-11-30T13:20:12.562165Z ERROR bosminer_hal::sensor: Sensor hb8[ii_hwmon::tmp451::TMP451]: read failed: I2C error: general error Hashchip: no response for read_register(reg=0x20) from chip One(61)
I then looked more closely with a magnifying glass by desoldering the heat sinks, it seems okay to me on tmp541 chip, no dust, the welds are not cracked… a friend told me that as for the s17 the welds crack over time and make false contacts.
I noticed that on the hashboard when the sensor goes wrong, the little red light flashes more slowly like a bad contact (if it helps I say so)
also in the line of code I saw write Chip 61, so I decided to test the chip as well as its surroundings and I came across the correct values of CLK: 0.9 and RI 1.8 in 12v plug.
from what I understood it is, the more its heating (so after 5min to heat up) the closer we get to a crash point because of the sensor (or other)
for the moment I am there without too much additional info… an idea maybe?
Big Thanks 🫡
Jump to: