Brand new innosilicon T2T-30T stops hashing within minutes, autotunes constantly

[emeric1]

Hello everyone,

I am having trouble mining bitcoin with a innosilicon T2T-30T. Here is the problem:
- Firmware always starts auto tune, gets to about 25TH in minute 1 (factory performance setting), then slowly drops from there down towards 0.005 TH. The maximum I did it is for 8 hours, and auto tune is never completed, stays within hashing speeds of 50MH to 5GH.

- I tried the following fixes:
   - Two different firmwares (the one recommended by innosilicon, and one "PSU" firmware also recommended by an innosilicon chinese word document)
   - 3 different mining pools, including the default one included with their setting - The miner connects to my pool, but of course, at these speeds, nothing is mined.

More info on miner:
Version
Type  T2TZ
Controller Version   soc
Build Date      8th of June 2019 08:15 AM
Platform Version          t2tz_soc_20190608_081559

Any help is highly appreciated... I have searched the web for hours without finding a fix..
Thank you so much in advance,
Emeric

[1715374387]

1715374387

[1715374387]

1715374387

[1715374387]

1715374387

[philipma1957]

set to balance and pool to  a good pool

btc.viabtc.com:3333 would be a pool that 'likes' inno's

you should stay stable  on balance or balance+ setting

very few innos like to be set to factory performance setting.

[emeric1]

Thanks for your help...

I have tried your recommended viabtc pool and the "balance setting." The miner has now been autotuning for 40 minutes, started circa 20+TH in minute 1-3, then all the way down to under 50GH/s. It has been stuck at an essentially meaningless hash rate for 30 minutes. The pattern is the same.

I am beginning to think I may have a defective unit...
Any other ideas?
Much appreciated if so.

[NotFuzzyWarm]

One basic question that has to be asked: Does the miner have good power feeding it? Voltage should be between 208-240VAC, preferably around 220VAC. Running on low voltage can cause PSU's to drop out which of course totally screws with the tuning process and general running.

[emeric1]

Hi Thank you for your insight.

It is connected on a 30amp breaker, 2 x 120 rods + Ground. I am using a Kentek 1 ft AC Power Cable IEC-60320 C13 to C14 Extension 14 AWG 15A 250V SJT for my connector. I will try to do some power measurements and test that. Thanks for the advice.


So after measurement, at the plug that goes into the miner, I got exactly 244 volts. So the lack of power does not seem to be an issue.

In the miner status section, there is a "Miner info" box with 3 hash boards highlighted. All of them say "Alive" when the miner turns on, and them move on to "Dead" status after a few minutes.

The issue appears to be that each board reaches a 93-94degree C and then turns to "Dead." Yet, the fan is always running at 100% and the miner is nowhere near hot, which leads me to think there is a heat sensor issue. Could I be onto something?

Thanks.

[mikeywith]

So after measurement, at the plug that goes into the miner, I got exactly 244 volts. So the lack of power does not seem to be an issue.

But you are slightly above specs which is as bad as being under it, the specs on their website say " 210~240V", it's only a 4volt difference but might be worth looking into.

... which leads me to think there is a heat sensor issue. Could I be onto something?

Each hash board has its own sensor/s, the likelihood hood of all temp sensors dying together on the same miner is pretty low.

Doesn't Inno provide some sort of Kernel log? I suppose they do, please post the full logs here.

My initial guess is power-related issues, second, comes the network, but I am hoping to find some clues in the kernel log.

[NotFuzzyWarm]

But you are slightly above specs which is as bad as being under it, the specs on their website say " 210~240V", it's only a 4volt difference but might be worth looking into.

Yesss - and no.

The issue with a high line is that if the input side of the PSU has poorly rated input diodes and/or bulk filter capacitors a high line *could* exceed their voltage ratings causing them to fail (short out) with very obvious results. A low line results in pulling more current from the wall and again, that could result in blowing the input diodes if their current rating is exceeded but more often just results in the PSU not being able to deliver enough voltage when under full load.

In short, if input voltage is too high you could blow the PSU but assuming none of the Majik Smoke has been released it should still operate just fine.

As Mikey said, you need to provide us a log of what the miner is reporting as it starts up and then fails. Be sure to use the code function # from the toolbar here to paste in the system logs.

[emeric1]

There does not appear to be a Kernel log in the interface, but I was able to get an error code with the InnoMonitor. Error code is 32 - which means hash boards overheated according to Inno's support team.

They told me to check on the fans. I did, and they both work, and both blow in the same direction. However, is there a specific direction that fans are supposed to blow? (i.e., could they have been mounted backwards at assembly). What I find interesting is that the built-in PSU blows air one way, and the fans blow air the other way--but maybe that's by design.

One of the things they told me is to check whether "Fans blow in the same direction as other miners"--but I do not have other miners from that brand and cannot find information online.

Thanks,
Emeric

[wndsnb]

What I find interesting is that the built-in PSU blows air one way, and the fans blow air the other way--but maybe that's by design.

If the PSU fans are blowing in the opposite direction then they are taking in the heated air from the miner exhaust, and heated air from the PSU exhaust is getting sucked into the miner intake.... so not a good scenario. Pretty sure that isn't how it's supposed to work.

[emeric1]

Thank you all for the help...

My problem is now officially fixed. One of the fans was not installed properly by Innosilicon (air intake fan was in the wrong direction!). I was hesitant to change the direction of the fans because of their "Warranty invalid if torn" sticker, but I am glad I ignored the warning.

I find it interesting that a brand new factory-sealed item could have been mounted improperly from the start--not sure how good their quality control department was on my miner batch.

Hopefully this conversation can help others who face the same problem.
Thanks all for your time!
Emeric