[Setup & Troubleshoot] Bitmain AntMiner S1 180GH/S miner

[klondike_bar]

Ok, well, I guess I found the source of the problem - the left side (aka blade #2) on that unit is bad. I had the unit on by itself on the same PSU as the fully working unit. Yes I did disconnect the other unit entirely. The faulty unit was all by its lonesome. Left the miner cool for a few minutes, powered it on - the same xxxxx on the second module. I tried powering just the faulty blade alone, but I guess the first blade needs to be one as it seems the controller board is fed through it so the unit will not power on fully without the "master" blade.

ASIC# says 16 for the faulty blade, so half of the chips are not functional, although if you go by counting x's and o's more than half are throwing errors or non functional.

xxxxxoxo xoxoxoxo xxxxxxxx xxxxxxxx

Swapping the PSU at this stage would be a pointless exercise.

so you are running a single antminer on what PSU? (brand, model, wattage, etc)

alternatively, use a multimeter and check the volatge on the bad board. to do this:
red (+) probe goes to the bottom connection of the large grey cube-shaped inductor (there are 4 of these per blade, test on all 4 on the bad side please)
black (-) probe goes to any of the small copper dots on the board marked with 'GND', or to the GND terminals for your PSU cables

you should see 1.10V on all 4 inductors. if not, it narrows down the possible sources/causes of the problem.

I mentioned the brand model wattage before..It's an OCZ EliteXstream 1000W, 1000 watts single 12v rail.  Hmm, well I'm going to the hosting location today and I'll bring my multimeter with me. I'm going to note down what you said about the voltage test points.

my apologies then, info gets buried in the thread easily. sounds like power isnt an issue in that case, so I'm stumped. Checking for the 1.10V signal is your best bet as it will at least indicate whether the chips are being fed the correct voltage to operate (if they are given voltage but refuse to run, it at least narrows the issue.

I did an RMA before on my first auction unit because the first section of my secondary board had no 1.1V signal from the TPS53355 chip, and as a result the entire chain (all 4 sections) would not function or be detected. Bitmain did an excellent job on the RMA and reimbursed me fully, even for the outrageous costs of shipping to china (seriously - how does it cost 1.5x as much to send a single PCB without heatsink to china as it did for them to ship the entire antminer to canada??)

alternatively, do temps seem to be any kind of issue? it may be worth checking hte tension of the hex screws holding down the pcb to the heatsink. if thermal contact is poor it may be possible the chips overheat and switch off very quickly, but i dont think this is the case

[1715320528]

1715320528

[1715320528]

1715320528

[1715320528]

1715320528

[1715320528]

1715320528

[1715320528]

1715320528

[1715320528]

1715320528

[rammy2k2]

use min 100cfm and ~3000 rpm (pwm). But I suggest to use 2 fan. One is connected to the miner and regulated with pwm mode.
Other is placed on opposite side, soaking air out and powered direct from PDU molex at highest speed (its silent so it wont be so laud).
Or you can power them both at higest speed and from PDU, no need of connecting to the miner fan connector and pwm regulation.

i need silent ones so i can sleep, so replacing the standard ones is a must , thanks for telling me

[alfabitcoin]

Allinvain
maybe to try firmware update? Who knows, its worth a try.
Go on bitmain github for the recent firmware.

[jermwerty]

Just FYI for those first timers that are setting up ANTs with muli-rail PSUs...  Try to find out how the rails are wired!

• For example I have a 2 rail PSU that one rail is dedicated to all molex/PCI-E/SATA, the other is to the mobo.  
• I have a 4 rail PSU that has one rail to mobo, one to PCI-E, one to Molex/SATA, and another PCI-E
• I have another 2 rail PSU where molex and PCI-E are on different rails.

With the newest ANTs I do like the included connector as I was able to re-purpose the 6pin PCI-E to connector be compatible with the 8-pin mobo connector as long as you change the +/- polarity when plugged into the blade (NOTE YOU WILL FRY THE ANT IF YOU GET +/- backwards, PAY ATTENTION!)  That allowed me to use a relatively weak 600W dual rail supply easily - one rail has mobo 8-pin and the other rail has PCI-E.

What I think most people don't realize is CHECK THE FRIGGIN OUTPUTS WITH A VOLTMETER UNDER LOAD!

If the 12V rail is dropping to something like 11.2V under load, you're gonna have a bad time!  I have a Corsair CX500 that drops to 11.6V under 435W load of an OC'd ANT, and thats as low of a 12V rail as I can stand... but what do you guys think?

How do you know how the rails are wired do you ask?  Some of the manufacturers will show you right in the specs.  Many don't though.  Power up the PSU and check different connectors and note each lead with a digital multimeter.  Then plug in one ANT blade (while powered off, then power on) and re-run the tests.  The rail with load will most likely drop .1V or so, and it should become obvious.  If the PSU has two rails, but both PCI-E connectors show a drop then most likely you are only using half the PSU's power by default.  Do my mobo connector trick (or molex, or whatever connector is on a different rail!) and life is good!

[Xer0]

With the newest ANTs I do like the included connector as I was able to re-purpose the 6pin PCI-E to connector be compatible with the 8-pin mobo connector as long as you change the +/- polarity when plugged into the blade

haha, i did the same but reverse - cut a 8pin momo cable adapter and plugged in 6pin pcie with swapped polarity
(could not get pcie extenders in any local store here as the miner came before my ebay adapters order)

[allinvain]

Ok, well, I guess I found the source of the problem - the left side (aka blade #2) on that unit is bad. I had the unit on by itself on the same PSU as the fully working unit. Yes I did disconnect the other unit entirely. The faulty unit was all by its lonesome. Left the miner cool for a few minutes, powered it on - the same xxxxx on the second module. I tried powering just the faulty blade alone, but I guess the first blade needs to be one as it seems the controller board is fed through it so the unit will not power on fully without the "master" blade.

ASIC# says 16 for the faulty blade, so half of the chips are not functional, although if you go by counting x's and o's more than half are throwing errors or non functional.

xxxxxoxo xoxoxoxo xxxxxxxx xxxxxxxx

Swapping the PSU at this stage would be a pointless exercise.

so you are running a single antminer on what PSU? (brand, model, wattage, etc)

alternatively, use a multimeter and check the volatge on the bad board. to do this:
red (+) probe goes to the bottom connection of the large grey cube-shaped inductor (there are 4 of these per blade, test on all 4 on the bad side please)
black (-) probe goes to any of the small copper dots on the board marked with 'GND', or to the GND terminals for your PSU cables

you should see 1.10V on all 4 inductors. if not, it narrows down the possible sources/causes of the problem.

I mentioned the brand model wattage before..It's an OCZ EliteXstream 1000W, 1000 watts single 12v rail.  Hmm, well I'm going to the hosting location today and I'll bring my multimeter with me. I'm going to note down what you said about the voltage test points.

my apologies then, info gets buried in the thread easily. sounds like power isnt an issue in that case, so I'm stumped. Checking for the 1.10V signal is your best bet as it will at least indicate whether the chips are being fed the correct voltage to operate (if they are given voltage but refuse to run, it at least narrows the issue.

I did an RMA before on my first auction unit because the first section of my secondary board had no 1.1V signal from the TPS53355 chip, and as a result the entire chain (all 4 sections) would not function or be detected. Bitmain did an excellent job on the RMA and reimbursed me fully, even for the outrageous costs of shipping to china (seriously - how does it cost 1.5x as much to send a single PCB without heatsink to china as it did for them to ship the entire antminer to canada??)

alternatively, do temps seem to be any kind of issue? it may be worth checking hte tension of the hex screws holding down the pcb to the heatsink. if thermal contact is poor it may be possible the chips overheat and switch off very quickly, but i dont think this is the case

I did some further troubleshooting today and disconnected the fully working blade and made the faulty one primary as sushi suggested and the same thing happens. I tried a different power supply - that did not help either. I did not test the voltage as I was a bit chicken on that front. To test that the unit has to be on I presume so I was worried that I'd short circuit something and end up frying the other working module as well or the control board. I noticed some oily residue on the back of the board (where the chips are) so I cleaned that up but that clearly had no bearing on the problem. As far as the heatsink it seems like it's pretty tightly held on. I saw no obvious assembly issues or broken capacitors or anything such as that.

So I think I will just RMA it and be done with it. Sushi gave me an RMA address that is located in USA so no shipping to China! Yay! Once I get approval from sushi it's off to UPS or FedEX I go.

I am just thankful this is the only dead blade out of all the units that I ordered.

[allinvain]

Allinvain
maybe to try firmware update? Who knows, its worth a try.
Go on bitmain github for the recent firmware.

It already came with the latest firmware so there was nothing newer to update to.

[allinvain]

any idea whats the CFM of the fan ?

i want to replace them with some silent ones, but need to know the CFM, anyone knows ?

You want to replace the main fan with a silent one? I'd be careful cause most silent fans simply cannot push the CFM rate required to properly cool the miner. I intend to leave the main PWM controlled fan on and put a more standard PC case fan at the back sucking air. I'd say 80 CFM should be enough to give the main fan a little bit of help.

In case anyone wants to know, this is the stock fan. Made by NMB:



Uploaded with ImageShack.us

[helipotte]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

[ineedit]

Has anyone update the firmware on an Ant yet especially from 29/11/13 to 12/12/13 ?

Any benefit?

[allinvain]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

I may do the same in the summer if they need additional cooling. I'm having a hard time finding retailers who sell Sanyo Denki fans in Canada; I wonder why.

[helipotte]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

I may do the same in the summer if they need additional cooling. I'm having a hard time finding retailers who sell Sanyo Denki fans in Canada; I wonder why.

I think I got my fans from a retailer in Canada.  This was about 9 years ago though.  These SD fans are unkillable.  Mine have ran nonstop cooling something since I have

Had them.  First thing they moved air across was a Pentium 4 overclocked.

[allinvain]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

I may do the same in the summer if they need additional cooling. I'm having a hard time finding retailers who sell Sanyo Denki fans in Canada; I wonder why.

I think I got my fans from a retailer in Canada.  This was about 9 years ago though.  These SD fans are unkillable.  Mine have ran nonstop cooling something since I have

Had them.  First thing they moved air across was a Pentium 4 overclocked.

Wow, are you serious?! Shit, are these made of adamantium or something. These are just simple single bearings, not the dual bearings kind, no?

I'm going to look for some too, but even on ebay I'm having a hard time finding the 4 pin connector kind. Most seem to be 5 pin.

[Uncle Scrooge]

Offering a Bounty to get my 3 AntMiners up and running.

I need someone to help me setup my 3 AntMiners. They hang at Miner Configuration screen when you hit save&apply.
A few days ago Sushi via teamviewer disabled something inside cgminer to get one Antminer running. Have 3 other units I need to get running.

Send me a PM if you're interested.

[sushi]

teamviewer please!!

Offering a Bounty to get my 3 AntMiners up and running.

I need someone to help me setup my 3 AntMiners. They hang at Miner Configuration screen when you hit save&apply.
A few days ago Sushi via teamviewer disabled something inside cgminer to get one Antminer running. Have 3 other units I need to get running.

Send me a PM if you're interested.

[bmoconno]

My Ants are mining very happily, but because of heat concerns, I decided to move them into another room and run them off of WiFi.  They seem to be mining and are visible when I check the "Connected Devices" on my router, but I can't remote into them (via the web interface or ssh), unless I connect them via ethernet.

I read through all these pages, and it seems that a few people are having the same problem that I am, but I never saw it addressed or solved.  There was some suggestion that maybe we can't remote into them over WiFi because of some setting in the Ants Firewall, but I'm not seeing anything that would likely be causing this.  I was just wondering if anyone has this problem or if they came to a solution.  Thanks!

[Uncle Scrooge]

Thank You Sushi for your time and effort. All 4 of the miners are up and running

Your tip is incoming!

[helipotte]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

I may do the same in the summer if they need additional cooling. I'm having a hard time finding retailers who sell Sanyo Denki fans in Canada; I wonder why.

I think I got my fans from a retailer in Canada.  This was about 9 years ago though.  These SD fans are unkillable.  Mine have ran nonstop cooling something since I have

Had them.  First thing they moved air across was a Pentium 4 overclocked.

Wow, are you serious?! Shit, are these made of adamantium or something. These are just simple single bearings, not the dual bearings kind, no?

I'm going to look for some too, but even on ebay I'm having a hard time finding the 4 pin connector kind. Most seem to be 5 pin.

I think mine are two bearing.  I bought 10 at the same time for about $5 each off the bay.  Not one has given any trouble at all.

These are the fans I have.

http://www.ebay.com/itm/NEW-lot-4-Sanyo-Denki-San-Ace-12-Volt-DC-fan-109R-H-102-CFM-JAPAN-120mm-PC-READY-/130849558955

[allinvain]

My Ants are mining very happily, but because of heat concerns, I decided to move them into another room and run them off of WiFi.  They seem to be mining and are visible when I check the "Connected Devices" on my router, but I can't remote into them (via the web interface or ssh), unless I connect them via ethernet.

I read through all these pages, and it seems that a few people are having the same problem that I am, but I never saw it addressed or solved.  There was some suggestion that maybe we can't remote into them over WiFi because of some setting in the Ants Firewall, but I'm not seeing anything that would likely be causing this.  I was just wondering if anyone has this problem or if they came to a solution.  Thanks!

That's likely cause ssh and remote access has been disabled on that interface. Oddly enough the WAN interface in the firmware is actually the ethernet interface physically and the LAN in the firmware is the wi-fi. Not sure why they did that. Initially when I was trying to setup my first unit this threw me off and I ended up manually setting an IP on the wrong interface so surprise surprise when I was not able to reach the unit at the newly assigned IP.

Check on "System" and then "Administration" to make sure ssh is enabled for all interfaces.

[allinvain]

I replaced the fan on mine with two sanyo denki units (DC san ace).  The stock fan on mine has a Dell P/n on it.

My temps are around 35c.  This is at full speed connected directly to 12V.  Kinda loud but don't care.

I may do the same in the summer if they need additional cooling. I'm having a hard time finding retailers who sell Sanyo Denki fans in Canada; I wonder why.

I think I got my fans from a retailer in Canada.  This was about 9 years ago though.  These SD fans are unkillable.  Mine have ran nonstop cooling something since I have

Had them.  First thing they moved air across was a Pentium 4 overclocked.

Wow, are you serious?! Shit, are these made of adamantium or something. These are just simple single bearings, not the dual bearings kind, no?

I'm going to look for some too, but even on ebay I'm having a hard time finding the 4 pin connector kind. Most seem to be 5 pin.

I think mine are two bearing.  I bought 10 at the same time for about $5 each off the bay.  Not one has given any trouble at all.

These are the fans I have.

http://www.ebay.com/itm/NEW-lot-4-Sanyo-Denki-San-Ace-12-Volt-DC-fan-109R-H-102-CFM-JAPAN-120mm-PC-READY-/130849558955

Nice. Apparently these have a built in temp sensor that auto adjusts the fan's speed.