jbmod
Newbie
Offline
Activity: 13
Merit: 0
|
|
December 02, 2017, 08:15:54 PM |
|
Need some assistance/advice with my brand new Antminer D3. I'm not trying to waste anyone's time so I'm going to try and be as detailed as possible so people can diagnose the problems I have. Alright, the first day I get the miner. Runs for about 3-4 hours. Has red "fault light flash"--> check and see every chip has a red "x" restart and the entire board 3 just no longer showed up on the page. I went and upgraded firmware and now all three are back up but the third only shows "54" chips while the rest show 60. I am unsure why this is. It's been going well on Antpool now for about 24 hours but I've noticed it's got millions... MILLIONS of hardware errors. I am unsure how to fix this or what it means. Here's my situation: Miner is a room with ac + window open. (80F- 60F at night). The power supply is the APW3++ from Bitmain. I'm in USA so I believe it's 120 volts if that matters. Everything else seems to be set to default. You can see my recent miner status here: It seems to have improved hash rate for the time being? https://imgur.com/oqT12HfPlus on a side note, it randomly hits 0.00 hash or something on antpool charts: https://imgur.com/a/d54aGAnyways, should I try to use my warranty to get it fixed? What are the benefits of using this "firmware" and how is it possible that it would be quieter faster hashing, and less power? Can't Bitmain just update theirs? I'm very new to this and would love to learn so all advice is greatly appreciated. When have installed on 120V you get only 1000-1100W out of a Bitmain PSU. The HW Errors is a fault from Bitmain, they put the frequency so high only to label the miner with 19.3GH, try in the settings to reduce the frequency to 506. this was my magic number on the last 2 D3 i get from Bitmain. Blissz, change the Firmware you can setup for all 3 hashboards independent the Power and frequency, how you see in your pic the last hashboard is hotter, lot HW-Errors. With this Firmware you can tweak your D3 to get the best out of all 3 Hashboards Thank you for that understanding, but my question is what exactly does it mean to get "hardware errors" it just means it's too hot? Hasn't even hit over 70C yet on a board? Plus, If I reduce the "frequency" what exactly does that do to my hash speed? Would it result in a reduced speed or no? what're the benefits? I have the same problem : https://www.dropbox.com/s/tstpxhj3jm4kpg8/Chips%20problem.jpg?dl=0Many HW and lost 6 chips in few hours... I contacted bitmain support's I' m waitting their respons But sometime I recover one chip and the second chain disappear ... https://www.dropbox.com/s/ng8w9cy86xlxe5c/55%20chips.jpg?dl=0keep me updated. Thanks As I mentioned earlier in this forum I was able to get around this problem by sending more power to the boards. With blissz firmware you're able to set each board individual power settings. I'd suggest to try blissz firmware and set the power of that board with less <60 ascis to a power setting more than 13 (bitmain default). I was able to get a D3 with only 55 of 60 asics showing with power settings to 15 or 16 back to all 60 asics and 0 HW errors.
|
|
|
|
Boneyg001
Newbie
Offline
Activity: 81
Merit: 0
|
|
December 02, 2017, 08:56:08 PM |
|
Need some assistance/advice with my brand new Antminer D3. I'm not trying to waste anyone's time so I'm going to try and be as detailed as possible so people can diagnose the problems I have. Alright, the first day I get the miner. Runs for about 3-4 hours. Has red "fault light flash"--> check and see every chip has a red "x" restart and the entire board 3 just no longer showed up on the page. I went and upgraded firmware and now all three are back up but the third only shows "54" chips while the rest show 60. I am unsure why this is. It's been going well on Antpool now for about 24 hours but I've noticed it's got millions... MILLIONS of hardware errors. I am unsure how to fix this or what it means. Here's my situation: Miner is a room with ac + window open. (80F- 60F at night). The power supply is the APW3++ from Bitmain. I'm in USA so I believe it's 120 volts if that matters. Everything else seems to be set to default. You can see my recent miner status here: It seems to have improved hash rate for the time being? https://imgur.com/oqT12HfPlus on a side note, it randomly hits 0.00 hash or something on antpool charts: https://imgur.com/a/d54aGAnyways, should I try to use my warranty to get it fixed? What are the benefits of using this "firmware" and how is it possible that it would be quieter faster hashing, and less power? Can't Bitmain just update theirs? I'm very new to this and would love to learn so all advice is greatly appreciated. When have installed on 120V you get only 1000-1100W out of a Bitmain PSU. The HW Errors is a fault from Bitmain, they put the frequency so high only to label the miner with 19.3GH, try in the settings to reduce the frequency to 506. this was my magic number on the last 2 D3 i get from Bitmain. Blissz, change the Firmware you can setup for all 3 hashboards independent the Power and frequency, how you see in your pic the last hashboard is hotter, lot HW-Errors. With this Firmware you can tweak your D3 to get the best out of all 3 Hashboards Thank you for that understanding, but my question is what exactly does it mean to get "hardware errors" it just means it's too hot? Hasn't even hit over 70C yet on a board? Plus, If I reduce the "frequency" what exactly does that do to my hash speed? Would it result in a reduced speed or no? what're the benefits? I have the same problem : https://www.dropbox.com/s/tstpxhj3jm4kpg8/Chips%20problem.jpg?dl=0Many HW and lost 6 chips in few hours... I contacted bitmain support's I' m waitting their respons But sometime I recover one chip and the second chain disappear ... https://www.dropbox.com/s/ng8w9cy86xlxe5c/55%20chips.jpg?dl=0keep me updated. Thanks As I mentioned earlier in this forum I was able to get around this problem by sending more power to the boards. With blissz firmware you're able to set each board individual power settings. I'd suggest to try blissz firmware and set the power of that board with less <60 ascis to a power setting more than 13 (bitmain default). I was able to get a D3 with only 55 of 60 asics showing with power settings to 15 or 16 back to all 60 asics and 0 HW errors. can I do that even with 120 volt wall outlet?
|
|
|
|
blissz (OP)
|
|
December 02, 2017, 09:05:26 PM |
|
As I mentioned earlier in this forum I was able to get around this problem by sending more power to the boards. With blissz firmware you're able to set each board individual power settings. I'd suggest to try blissz firmware and set the power of that board with less <60 ascis to a power setting more than 13 (bitmain default). I was able to get a D3 with only 55 of 60 asics showing with power settings to 15 or 16 back to all 60 asics and 0 HW errors.
Good information, thanks for sharing!
|
|
|
|
dottat
Newbie
Offline
Activity: 48
Merit: 0
|
|
December 02, 2017, 09:15:27 PM |
|
Need some assistance/advice with my brand new Antminer D3. I'm not trying to waste anyone's time so I'm going to try and be as detailed as possible so people can diagnose the problems I have. Alright, the first day I get the miner. Runs for about 3-4 hours. Has red "fault light flash"--> check and see every chip has a red "x" restart and the entire board 3 just no longer showed up on the page. I went and upgraded firmware and now all three are back up but the third only shows "54" chips while the rest show 60. I am unsure why this is. It's been going well on Antpool now for about 24 hours but I've noticed it's got millions... MILLIONS of hardware errors. I am unsure how to fix this or what it means. Here's my situation: Miner is a room with ac + window open. (80F- 60F at night). The power supply is the APW3++ from Bitmain. I'm in USA so I believe it's 120 volts if that matters. Everything else seems to be set to default. You can see my recent miner status here: It seems to have improved hash rate for the time being? https://imgur.com/oqT12HfPlus on a side note, it randomly hits 0.00 hash or something on antpool charts: https://imgur.com/a/d54aGAnyways, should I try to use my warranty to get it fixed? What are the benefits of using this "firmware" and how is it possible that it would be quieter faster hashing, and less power? Can't Bitmain just update theirs? I'm very new to this and would love to learn so all advice is greatly appreciated. When have installed on 120V you get only 1000-1100W out of a Bitmain PSU. The HW Errors is a fault from Bitmain, they put the frequency so high only to label the miner with 19.3GH, try in the settings to reduce the frequency to 506. this was my magic number on the last 2 D3 i get from Bitmain. Blissz, change the Firmware you can setup for all 3 hashboards independent the Power and frequency, how you see in your pic the last hashboard is hotter, lot HW-Errors. With this Firmware you can tweak your D3 to get the best out of all 3 Hashboards Thank you for that understanding, but my question is what exactly does it mean to get "hardware errors" it just means it's too hot? Hasn't even hit over 70C yet on a board? Plus, If I reduce the "frequency" what exactly does that do to my hash speed? Would it result in a reduced speed or no? what're the benefits? I have the same problem : https://www.dropbox.com/s/tstpxhj3jm4kpg8/Chips%20problem.jpg?dl=0Many HW and lost 6 chips in few hours... I contacted bitmain support's I' m waitting their respons But sometime I recover one chip and the second chain disappear ... https://www.dropbox.com/s/ng8w9cy86xlxe5c/55%20chips.jpg?dl=0keep me updated. Thanks As I mentioned earlier in this forum I was able to get around this problem by sending more power to the boards. With blissz firmware you're able to set each board individual power settings. I'd suggest to try blissz firmware and set the power of that board with less <60 ascis to a power setting more than 13 (bitmain default). I was able to get a D3 with only 55 of 60 asics showing with power settings to 15 or 16 back to all 60 asics and 0 HW errors. can I do that even with 120 volt wall outlet? Your best bet on 120 is to lower your voltage and clock speed to ensure you are powering the rig sufficiently. I have mine both at 437 with voltage set to 5. I get about 16,500 hash out of that setting and draw about 740 watts. Give it a try and see what happens.
|
|
|
|
heythorp
Newbie
Offline
Activity: 28
Merit: 0
|
|
December 02, 2017, 09:35:17 PM |
|
For the people loosing a chain.
Try running just that chain by itself. Unplug the other chains from power and the control board.
If it works at default settings then its most likely a power issue.
If it doesn't work by itself you should warranty it. I have one board at bitmain now. So far the process has been very good and their support has be pretty quick to respond.
|
|
|
|
blissz (OP)
|
|
December 03, 2017, 12:05:35 AM |
|
v1.1 :: 02/12 :: recommended updateDownload:https://mega.nz/#!nthHEQ5R!EwtiWy94w683ohv9CjI8_9xU3JQ0AvJ_rd49G8dguEoChanges 02/12:- improved temperature sensor reading stability of all chains - improved board / chain detection - fixed an issue where the miner could hang mining the dev fee if your first pool is down - improved awesome miner compatibility (special thanks to @funminer for beta testing) happy mining!
|
|
|
|
heythorp
Newbie
Offline
Activity: 28
Merit: 0
|
|
December 03, 2017, 12:13:30 AM |
|
v1.1 :: 02/12 :: recommended updateDownload:https://mega.nz/#!nthHEQ5R!EwtiWy94w683ohv9CjI8_9xU3JQ0AvJ_rd49G8dguEoChanges 02/12:- improved temperature sensor reading stability of all chains - improved board / chain detection - fixed an issue where the miner could hang mining the dev fee if your first pool is down - improved awesome miner compatibility (special thanks to @funminer for beta testing) happy mining! thanks for the continued work! I know threads can get long and posts can get lost. So I just wanted to state again. It would be really nice if the fan speed was on the same page as the advance settings. That way it only takes one reboot when switching from silent/performance settings.
|
|
|
|
blissz (OP)
|
|
December 03, 2017, 12:15:55 AM |
|
thanks for the continued work!
I know threads can get long and posts can get lost. So I just wanted to state again. It would be really nice if the fan speed was on the same page as the advance settings. That way it only takes one reboot when switching from silent/performance settings.
Good idea! I add it to the list
|
|
|
|
dalu1
Member
Offline
Activity: 83
Merit: 10
Performing it is a better way of learning things
|
|
December 03, 2017, 12:16:58 AM |
|
Thanks Blissz for the great work.
|
|
|
|
funminer
|
|
December 03, 2017, 01:17:42 AM |
|
Need some assistance/advice with my brand new Antminer D3. I'm not trying to waste anyone's time so I'm going to try and be as detailed as possible so people can diagnose the problems I have. Alright, the first day I get the miner. Runs for about 3-4 hours. Has red "fault light flash"--> check and see every chip has a red "x" restart and the entire board 3 just no longer showed up on the page. I went and upgraded firmware and now all three are back up but the third only shows "54" chips while the rest show 60. I am unsure why this is. It's been going well on Antpool now for about 24 hours but I've noticed it's got millions... MILLIONS of hardware errors. I am unsure how to fix this or what it means. Here's my situation: Miner is a room with ac + window open. (80F- 60F at night). The power supply is the APW3++ from Bitmain. I'm in USA so I believe it's 120 volts if that matters. Everything else seems to be set to default. You can see my recent miner status here: It seems to have improved hash rate for the time being? https://imgur.com/oqT12HfPlus on a side note, it randomly hits 0.00 hash or something on antpool charts: https://imgur.com/a/d54aGAnyways, should I try to use my warranty to get it fixed? What are the benefits of using this "firmware" and how is it possible that it would be quieter faster hashing, and less power? Can't Bitmain just update theirs? I'm very new to this and would love to learn so all advice is greatly appreciated. When have installed on 120V you get only 1000-1100W out of a Bitmain PSU. The HW Errors is a fault from Bitmain, they put the frequency so high only to label the miner with 19.3GH, try in the settings to reduce the frequency to 506. this was my magic number on the last 2 D3 i get from Bitmain. Blissz, change the Firmware you can setup for all 3 hashboards independent the Power and frequency, how you see in your pic the last hashboard is hotter, lot HW-Errors. With this Firmware you can tweak your D3 to get the best out of all 3 Hashboards Thank you for that understanding, but my question is what exactly does it mean to get "hardware errors" it just means it's too hot? Hasn't even hit over 70C yet on a board? Plus, If I reduce the "frequency" what exactly does that do to my hash speed? Would it result in a reduced speed or no? what're the benefits? Hardware Error is a calculation problem in a ASIC chip, it´s no problem to have a few of them, but when there are to much like in your pics, then it´s mosttime a power problem. There are 2 ways to fix this problem when you use the Orig. Firmware from Bitmain, you must lower the frequency, the hashrate go a little down and your Hashboards not use so much power. How i say before, bitmain not care on these even with 240V in Germany i get to much HW-Error on 1 Board in all my Miner with Facory Setting of 537. use the Firmware from Blissz, here you can lower the frequency only for the Hashboard that have to many HW-Errors or when your board not get to hot you can set the Voltage higher. Depend the Powercost and for the health of the miner i will preffer to lower the frequency a little bit. And with your 120V Powerline, you must lower Power consume from your miner to get it stable.
|
|
|
|
funminer
|
|
December 03, 2017, 01:21:50 AM |
|
v1.1 :: 02/12 :: recommended updateDownload:https://mega.nz/#!nthHEQ5R!EwtiWy94w683ohv9CjI8_9xU3JQ0AvJ_rd49G8dguEoChanges 02/12:- improved temperature sensor reading stability of all chains - improved board / chain detection - fixed an issue where the miner could hang mining the dev fee if your first pool is down - improved awesome miner compatibility (special thanks to @funminer for beta testing) happy mining! Thanks for the Update!
|
|
|
|
buonleloi
Newbie
Offline
Activity: 47
Merit: 0
|
|
December 03, 2017, 01:26:15 AM |
|
Seem more stable now.
Great work !
|
|
|
|
Bajula
Member
Offline
Activity: 166
Merit: 16
|
|
December 03, 2017, 01:49:59 AM |
|
Just a little thought on algos: since x11 is set up to switch rapidly between multiple algos and the asic chips are pre rigged to understand those algos, what if someone can find a way software side to just keep telling it only use THIS algo out of the 11. Basically trick x11 into not trying to switch?? Next thing: If you were bitmain and were thinking of the future would you design the chips to do x11 only or design them for say x15 and basically shut off function for the other4 and have a way to turn on x13 or x15 (I don't remember if there are more total differences besides adding a couple of algos each time or not) -(warning next bit is fun conspiracy theory) In any case.. maybe you would design it for x11 only - especially after charlie lee (creator of litecoin for the few who don't know) quits coinbase, goes on a tour of plants that manufacture scrypt machines, they suddenly crank out tons of x11 killing the profit on them leaving scrypt to be the rockin' machines for paying your bills) - no I don't really think it is that kind of pre-planned conspiracy, but that coupled with a couple other things makes me think some folks saw and exploited an event they not only saw coming but kinda "helped along" on to hardware matters, on my d3 - I was unsure of the power draw so I have 3 rm1000 corsairs (been running all go- no quit since s1 days.. too bad they don't make them any more.) So input power limit not being as significant a factor as ambient temp (I'm in Texas) wire thickness and chip temps.. oh and even the wife doesn't complain about the fan noise so 100% is okay.. What do ya think I might be able to safely get out of them
|
|
|
|
fecell
Jr. Member
Offline
Activity: 135
Merit: 2
|
|
December 03, 2017, 02:31:27 AM |
|
bullshit. The D3 ASIC 512/010/002 Configurators use a different algorithm than the D3 ASIC 256 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 040000H. On D3 ASIC 002/040 Configurators, the sequential reading of these bytes are accom- plished by performing a Random Read at EEPROM Address 100000H
and The AT17N512/010/002 Configurators use a different algorithm than the AT17N256 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 040000H. On AT17N002/040 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 100000H. http://www.atmel.com/Images/doc3015.pdfsimilar, isn't it? 100% scam
|
|
|
|
may142012
Member
Offline
Activity: 301
Merit: 10
|
|
December 03, 2017, 03:09:13 AM |
|
bullshit. The D3 ASIC 512/010/002 Configurators use a different algorithm than the D3 ASIC 256 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 040000H. On D3 ASIC 002/040 Configurators, the sequential reading of these bytes are accom- plished by performing a Random Read at EEPROM Address 100000H
and The AT17N512/010/002 Configurators use a different algorithm than the AT17N256 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 040000H. On AT17N002/040 Configurators, the sequential reading of these bytes are accomplished by performing a Random Read at EEPROM Address 100000H. http://www.atmel.com/Images/doc3015.pdfsimilar, isn't it? 100% scam I doubt, some idiots will eventually fall for this, i am afraid.
|
|
|
|
RfBTC
Member
Offline
Activity: 68
Merit: 10
|
|
December 03, 2017, 03:16:56 AM |
|
I have tested my Stock D3, at 19 GHs and 1145W..... its mined around 11 ONX (Onix), in 24h, at zpool.ca
What coin (x11) are you mining at what pool / profit per day it rewards ? , i think Nicehash is the best option atm.
|
|
|
|
may142012
Member
Offline
Activity: 301
Merit: 10
|
|
December 03, 2017, 04:23:19 AM |
|
I have tested my Stock D3, at 19 GHs and 1145W..... its mined around 11 ONX (Onix), in 24h, at zpool.ca
What coin (x11) are you mining at what pool / profit per day it rewards ? , i think Nicehash is the best option atm.
You get no answers, is it? It is because no one shares any secret, because they are so desperate to make a little to recover their investment on D3 that included me, since Shitmain flooded the market.
|
|
|
|
invisiblefriend321
Member
Offline
Activity: 174
Merit: 10
|
|
December 03, 2017, 07:01:41 AM |
|
incredible project -- respect n thx
|
|
|
|
Marvell2
|
|
December 03, 2017, 07:40:30 AM |
|
so i have 4 machines on this firmware
1 seems to have a bad chain
other two chains are strong
on average i am getting around 17500 m/s and 880 watts
averrage freqency is 481 and power at 6-8 max
one D3 is the silicon lottery
it can do 18600 m/s at 900 watts
great firmware so far.
You really should do a firmware for the S9 as well to reduce power wasted and also hardware errors man you are pro !!!!
Just for reference i was doing 19600 h/s at 1200 watts on my 96% platinum pus's
now I am averaging 17600 m/s at 880-920 watts max.
whats nice is i can put each unit in a single 1000 watt PSU now if i wanted
- only big issue i see so far is chain 1
I dont think the top chain is reporting the corret temps , its ether super low at like 30 degerees or way high 95-101 when all the other chains are in the 60s or 70s so its hard to tune it
|
|
|
|
blissz (OP)
|
|
December 03, 2017, 08:48:09 AM |
|
so i have 4 machines on this firmware
1 seems to have a bad chain
other two chains are strong
on average i am getting around 17500 m/s and 880 watts
averrage freqency is 481 and power at 6-8 max
one D3 is the silicon lottery
it can do 18600 m/s at 900 watts
great firmware so far.
You really should do a firmware for the S9 as well to reduce power wasted and also hardware errors man you are pro !!!!
Just for reference i was doing 19600 h/s at 1200 watts on my 96% platinum pus's
now I am averaging 17600 m/s at 880-920 watts max.
whats nice is i can put each unit in a single 1000 watt PSU now if i wanted
- only big issue i see so far is chain 1
I dont think the top chain is reporting the corret temps , its ether super low at like 30 degerees or way high 95-101 when all the other chains are in the 60s or 70s so its hard to tune it
Good results! Try the latest firmware on the unit: It will probably help with the 1st chain temperature readings.
|
|
|
|
|