Topic: ANTMINER S5: 1155GH(+OverClock Potential), In Stock $0.319/GH & 0.51W/GH - page 96. (Read 451266 times)
April 25, 2015, 05:27:50 AM
now show you pictures
April 25, 2015, 05:20:15 AM
I reversed the blades and now it works, but the problem that remains in the blade where it goes attacked the controller remains burnt, so I should have a new blade guess right? but the thing that I do not admit is that I have to pay as well the Italian customs to get back garanzia.basta write a piece in the declaration hardware parts replaced under warranty because the product has already been paid previously
I got my parts with DPD direct From China . No VAT.
artic mx-4 and not exceeding mx-2?
MX-4 is more expensive than the MX-2now and reassembled, and the small 4-pin connector that goes from the blade hashboard
This is the controller's power cord.
April 25, 2015, 05:19:23 AM
I reversed the blades and now it works, but the problem that remains in the blade where it goes attacked the controller remains burnt, so I should have a new blade guess right? but the thing that I do not admit is that I have to pay as well the Italian customs to get back garanzia.basta write a piece in the declaration hardware parts replaced under warranty because the product has already been paid previously
marvykkio, can you show us a picture of the burnt attachment point? Is it on the data cable, the PCIE power cable, or the power cable for the control board (that's the small 4 pin cable that's only attached to one hashboard)?
now and reassembled, and the small 4-pin connector that goes from the blade hashboard
April 25, 2015, 05:17:40 AM
artic mx-4 and not exceeding mx-2?
April 25, 2015, 05:16:52 AM
I reversed the blades and now it works, but the problem that remains in the blade where it goes attacked the controller remains burnt, so I should have a new blade guess right? but the thing that I do not admit is that I have to pay as well the Italian customs to get back garanzia.basta write a piece in the declaration hardware parts replaced under warranty because the product has already been paid previously
marvykkio, can you show us a picture of the burnt attachment point? Is it on the data cable, the PCIE power cable, or the power cable for the control board (that's the small 4 pin cable that's only attached to one hashboard)?
April 25, 2015, 05:13:56 AM
now that I disassembled the miner I have to find a good thermal paste to put on and dissipate heat, which pasta you advise me? the best 
Something is wrong. An Italian is asking people in the rest of the world for advice on pasta.
April 25, 2015, 05:03:16 AM
I reversed the blades and now it works, but the problem that remains in the blade where it goes attacked the controller remains burnt, so I should have a new blade guess right? but the thing that I do not admit is that I have to pay as well the Italian customs to get back garanzia.basta write a piece in the declaration hardware parts replaced under warranty because the product has already been paid previously
April 25, 2015, 05:01:22 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
Do you have a multimeter? Check the output voltage of your PSU. That voltage makes a bigger difference in power consumption than frequency does, actually.
For CMOS ASICs, power consumption is roughly proportional to clockspeed times voltage squared.
Since hashrate is proportional to clockspeed, and power is also proportional to clockspeed, that means that efficiency (or J/GH, which is actually the inverse of efficiency) is independent of clockspeed as long as voltage is kept constant. Overclocking S5s will not make the efficiency worse. Only overvolting will.
I have some S5s in our DC that use about 570 W of AC when run at 361 MHz. The trick? We're tweaking our PSUs to output 11.7 volts (open circuit). We also have a little trick with the cables that I might explain sometime when I have a free moment.
Kill-A-Watts are useful and cheap, but they are not accurate. I often see 5% to 10% discrepancies between our Kill-A-Watt and our Dell 0.2% accuracy metered PDUs. Kill-A-Watts get really confused by AC harmonics and non-sinusoidal waveforms, and computer power supplies have a lot of that.
Also, Bitmain does their testing on 230V. If you're using 120V, you should expect 3% higher power consumption.
April 25, 2015, 04:55:25 AM
this morning I packed all the miner, and reversed the blades, and I reassembled everything and went back to work, the problem is gone, now it works again, but the problem that remains is that part of the blade is not working so if you were to burn well this the miner does not work anymore. What do you think I should do? bitamain should send a new blade, but I think this is impossible, that I definitely want to send him the blade, or all of the miner to review it, and spends much time in addition to spending more money to send the miner from italy up in china. what should i do?
I sent all my broken Antminer S5 parts to Bitmaintech (Serbia).
Read more Bitmain tech support and warranty
now that I disassembled the miner I have to find a good thermal paste to put on and dissipate heat, which pasta you advise me? the best
I used the MX-2 http://www.arctic.ac/us_en/mx-2.html
65 g tube
April 25, 2015, 04:52:11 AM
now that I disassembled the miner I have to find a good thermal paste to put on and dissipate heat, which pasta you advise me? the best
April 25, 2015, 04:48:12 AM
this morning I packed all the miner, and reversed the blades, and I reassembled everything and went back to work, the problem is gone, now it works again, but the problem that remains is that part of the blade is not working so if you were to burn well this the miner does not work anymore. What do you think I should do? bitamain should send a new blade, but I think this is impossible, that I definitely want to send him the blade, or all of the miner to review it, and spends much time in addition to spending more money to send the miner from italy up in china. what should i do?
April 25, 2015, 04:43:46 AM
this morning I packed all the miner, and reversed the blades, and I reassembled everything and went back to work, the problem is gone, now it works again, but the problem that remains is that part of the blade is not working so if you were to burn well this the miner does not work anymore. What do you think I should do? bitamain should send a new blade, but I think this is impossible, that I definitely want to send him the blade, or all of the miner to review it, and spends much time in addition to spending more money to send the miner from italy up in china. what should i do?
April 25, 2015, 03:03:45 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
590W in DCS5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
kill a watt running in AC mode, so it's different.
i think using evga g2 750w is enough for your s5
i have 3 ants5 using seasonic 750w, acbel 900w, and xigmatek 1000w, all running 24hrs/day
no problem at all, but the seasonic 750 flow is more hot between the others.
April 25, 2015, 02:40:19 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
is this after the psu? i think the 590 is on the 12v line.
you have to also factor in the psu efficiency.. on a 750w psu you are at like 85% and most psu are most efficient at 50-60%
I should have probably got one of those 1300W, right?
I think a 1300 is overkill. I have had good luck with RM1000's.
I love them have a few of them on a few different miners, including a S5. If you want even better you can go with platinum PSU.
Which platinum PSU would you recommend? The EVGA 1000 P2?
April 25, 2015, 02:27:50 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
is this after the psu? i think the 590 is on the 12v line.
you have to also factor in the psu efficiency.. on a 750w psu you are at like 85% and most psu are most efficient at 50-60%
I should have probably got one of those 1300W, right?
I think a 1300 is overkill. I have had good luck with RM1000's.
I love them have a few of them on a few different miners, including a S5. If you want even better you can go with platinum PSU.
April 25, 2015, 02:23:26 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
is this after the psu? i think the 590 is on the 12v line.
you have to also factor in the psu efficiency.. on a 750w psu you are at like 85% and most psu are most efficient at 50-60%
I should have probably got one of those 1300W, right?
April 25, 2015, 02:08:07 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
is this after the psu? i think the 590 is on the 12v line.
you have to also factor in the psu efficiency.. on a 750w psu you are at like 85% and most psu are most efficient at 50-60%
April 25, 2015, 02:03:05 AM
I have several of those S5's and I'm experiencing a huge difference in power consumption.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
S5 running at 350M = 630W
Why am I not getting the 590W with the default frequency of 350M?
I'm using EVGA G2 750W power supply. Measuring power using a Kill A Watt.
April 24, 2015, 10:22:54 PM
Odd problem with my S5 just started.
I haven't really used this S5 for about a month as I'm more into Spondoolies, but I fired up two of them.
One works fine, the other I can connect to, so it's on the LAN, check the pool setup, even flashed the newest firmware per today, but no matter what, it will not hash. It just keeps increasing the 'Rejected' #'s slowly.
I reset it to factory settings, no difference. Anyone have something like this happen before?
I haven't really used this S5 for about a month as I'm more into Spondoolies, but I fired up two of them.
One works fine, the other I can connect to, so it's on the LAN, check the pool setup, even flashed the newest firmware per today, but no matter what, it will not hash. It just keeps increasing the 'Rejected' #'s slowly.
I reset it to factory settings, no difference. Anyone have something like this happen before?
What pool are you mining to? With all rejects it reminds me of what some devices do with pools such as nicehash.
Well, I'll be. I removed westhash and it works fine now. WTH? Works fine on the other s5. Same firmware, exact pool settings, etc..
Thanks for the idea.
That is the problem with some miners with westhash. I have the Thu Dec 18 20:21:28 CST 2014 firmware on my S5. It seems to work with westhash.
Smit makes a special firmware for nicehash/westhash. It is not an official firmware though. I have ran it on my S4's an C1's, but with S5 I'm not brave enough as it's firmware is not running off SD card.
You could always use a proxy on a pc aswell, and use that to connect to westhash. Then you mine to your proxy.
April 24, 2015, 10:13:32 PM
A random observation:
1. It seems that upon loss of internet connection, S5 sometimes latches on a funky state: fan is not spinning or spinning minimally, but temperature is rising.
2. We had intermittant comcast problems in the last few days, so i observed it on a few occasions.
3. I felt that air coming out was very hot, and upon restarting the miner by PSU, miner reported 0 hashing, but temperatures of above 80 degrees at the sensor, so i had to switch it off for 10-20 min, then it worked as it should.
I believe that someone described something similar before.
The solution should be in software, which should maintain fan speed at a certain (closer to high) speed until internet connection is reestablished.
The fact is that the S5 miner does not fully switch itself off on internet drop or, at least, does not maintain fan speed long enough to coool the machine after internet dropoff.
1. It seems that upon loss of internet connection, S5 sometimes latches on a funky state: fan is not spinning or spinning minimally, but temperature is rising.
2. We had intermittant comcast problems in the last few days, so i observed it on a few occasions.
3. I felt that air coming out was very hot, and upon restarting the miner by PSU, miner reported 0 hashing, but temperatures of above 80 degrees at the sensor, so i had to switch it off for 10-20 min, then it worked as it should.
I believe that someone described something similar before.
The solution should be in software, which should maintain fan speed at a certain (closer to high) speed until internet connection is reestablished.
The fact is that the S5 miner does not fully switch itself off on internet drop or, at least, does not maintain fan speed long enough to coool the machine after internet dropoff.
I mentioned this exact same thing happened to me before. Maybe 3 weeks ago or so in this thread. Even a couple others mentioned the same thing and had screen shots showing over 80 degrees, which shouldn't even happen if the shutdown is 80. All my rigs stopped because of a sporadic internet connection. Miners fans stopped and they all turned into little heaters, getting hotter and hotter. All I needed to do was put my hand or face near one of them and felt the heat spewing off. Thank god I was home to power them off or all my rigs would have over heated. I mentioned and tried to bring this up but nothing came about because I had no "proof" it seems. This happened to me twice on all 6 of my S5's and both times I was lucky enough to be home and power off.
I've also seen this. Last weekend while doing network changes I was off the internet for about an hour. 10 minutes in all of my s5s started either maxing fans or the fans shut down, but the heat kept going up.
Jump to: