Temprature readings for S7 - How to check correct chip Temp.

[najeebs]

Hello all ,

I have an S7 Unit running for some time now and i would like to ask if the following situation is normal and if not, what sould be my next step.

Hashboard 1 always seems to be running 20 degrees hotter than Hashboard 2 & 3 .

Current temperatures on hasboard 1 always seem to hover around 65 / 70 Degrees and hasboard 2/3 are always stable at 45 to 50 and their temp readings are always identical.

is a temp sensor on board 1 not good or is it the case for the other two boards ..?

any feedback much appreciated.

Thanks,

[1714850504]

1714850504

[1714850504]

1714850504

[1714850504]

1714850504

[fanatic26]

While it is normal for one board to be hotter than the others in the S7 the temperature difference shouldnt be that much. Have you disassembled and cleaned the miner?

[najeebs]

Hi ,

Yes, i thought of it as well initially and had it cleaned thoroughly inside.

although overall temperatures decreased a bit, the difference still remains as i mentioned.

thanks .

[MinerMEDIC]

I was wondering about this too my S7 dosn't seem to get much above 35°, something I find odd to impossible. Seems to me that at that temperature they wouldn't even need active cooling.

[najeebs]

Yep ,

I am not sure whats happening, I have attached a snapshot of my miner status, is it all normal, i mean to say as being for an S7. i am pretty new to this stuff I am not sure about the different sections and what they mean, i.e discarded / WU / Utility ..etc etc ..

Thanks

https://ibb.co/hgaKhR

[lightfoot]

Hm. Have you tried switching the boards around a bit? I've noticed that chain 1 can run 10c hotter than the other two, I think it's due to air cavitating against the side of the box (board 3 has the heat sinks in a better air flow).

You can always slow it down 50mhz or so, they're more efficient at around 600-650mhz as opposed to 700

[najeebs]

hey lightfoot ,

that's a good idea, will try swapping them around to see if it makes a difference.

Currently I am running my miner at 575M   , the boards have a small stickers which says 525M , I would love to overclock them , how ever have noticed that the hash rates start to decrease at 600 and above, any pointers in that direction ?? i have been to a max of 650 and the hash rates still are not as good as they are at 575 . should i try going higher like 700 and 750 ... the temperatures at 600 range seems to be stable .

Thanks,

[MinerMEDIC]

Speaking of getting detailed information from an S7 where are the logs when you SSH into the machine? I didn't see any in /var/logs, Anyhow I'm a hardware guy and know about just enough linux to get into trouble.

[fanatic26]

hey lightfoot ,

that's a good idea, will try swapping them around to see if it makes a difference.

Currently I am running my miner at 575M   , the boards have a small stickers which says 525M , I would love to overclock them , how ever have noticed that the hash rates start to decrease at 600 and above, any pointers in that direction ?? i have been to a max of 650 and the hash rates still are not as good as they are at 575 . should i try going higher like 700 and 750 ... the temperatures at 600 range seems to be stable .

Thanks,

The 54 chip boards have stock frequencies of up to 650 depending on the batch. (650mh makes the miner run at 5.67TH, they only ever had internal batches like this and none released to the public. ) I would probably not push the machine that hard if you want it to last for an appreciable amount of ttime.

Another thing of note, while you do have one board running hotter than the others, even that board is so far inside the safe range I wouldnt even question it. These boards will run to 80c before shutdown.

[lightfoot]

As your clock speed goes up errors might go up as well. Or engines could fail to start, reducing overall hash rate. Find the best speed and rock with that.

[SGMPhil]

Just FYI - I run a group of S9s and they all do the same thing - board 1 always runs hotter that the rest. I attribute this to airflow through the devices and think  is just normal due to the physical design of the miner, board positions and airflow from the fans. My guess is that moving the boards around will most likely not change the symptom as the position of board 1 simply does not get the same airflow as the other two boards. Just my $0.02 worth.

[lightfoot]

That's what I am thinking: Board one has the heat sinks on top of the chips right on the wall, 2 and 3 have the heat sinks more in the main airflow. Not the end of the world, just is what it is.