Author

Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 108. (Read 1260226 times)

legendary
Activity: 1638
Merit: 1005
LOOP[2] ON TO:0 (w:79)
 4: DC2DC/1/:[vlt1:660 vlt2:664(DCl:794 Tl:664 Ul:727) 12W  18A  87c] ASIC:[ 85c (125c) 770hz(BL: 770)   35 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:724 vlt2:727(DCl:794 Tl:794 Ul:727) 106W 146A  60c] ASIC:[ 85c (125c) 1050hz(BL:1050)   44 (E:193) F:0 L:0]

What is wrong ? That ASIC was all good before tonight ... :-(

then after like 15 minutes ..

LOOP[0] OFF TO:0 (test serial failed)
0: disabled (serial problem)
1: disabled (serial problem)

same behaviour after 2-3 reboot (power off/on)

Allright ... so now I made a few tests with another power supply.

As soon as I put power in connector #4 ... 3 and 4 doesn't startup.

If I put power only in the connector 3 ... 3 is working.

Smells ... really bad (smells actually nothing ... just that I have a bad feeling)

EDIT: well in fact the PSU do not start if there a connector plugged in the connector 4. Tested with one PSU and two PSU... I think i have DC2DC chip issue ... Like ASIC 4 ... i think the DC2DC chip is overheating.

Just letting you know ... The unit will be sent as RMA. Will let you know the outcome !

I can say they have always treated me great.  I met some of them in person, all were very nice guys.   Hope you get back to hashing soon Smiley.

So far, I have nothing to say except good thing. A1 support.
legendary
Activity: 1456
Merit: 1000
LOOP[2] ON TO:0 (w:79)
 4: DC2DC/1/:[vlt1:660 vlt2:664(DCl:794 Tl:664 Ul:727) 12W  18A  87c] ASIC:[ 85c (125c) 770hz(BL: 770)   35 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:724 vlt2:727(DCl:794 Tl:794 Ul:727) 106W 146A  60c] ASIC:[ 85c (125c) 1050hz(BL:1050)   44 (E:193) F:0 L:0]

What is wrong ? That ASIC was all good before tonight ... :-(

then after like 15 minutes ..

LOOP[0] OFF TO:0 (test serial failed)
0: disabled (serial problem)
1: disabled (serial problem)

same behaviour after 2-3 reboot (power off/on)

Allright ... so now I made a few tests with another power supply.

As soon as I put power in connector #4 ... 3 and 4 doesn't startup.

If I put power only in the connector 3 ... 3 is working.

Smells ... really bad (smells actually nothing ... just that I have a bad feeling)

EDIT: well in fact the PSU do not start if there a connector plugged in the connector 4. Tested with one PSU and two PSU... I think i have DC2DC chip issue ... Like ASIC 4 ... i think the DC2DC chip is overheating.

Just letting you know ... The unit will be sent as RMA. Will let you know the outcome !

I can say they have always treated me great.  I met some of them in person, all were very nice guys.   Hope you get back to hashing soon Smiley.
legendary
Activity: 1456
Merit: 1000
Hello everyone. I'm a CoinTerra refugee and new to the Spondoolies camp. Just got an SP31 up and running over the weekend.

I can't seem to find any definitive explanation for setting the "Fan Speed" parameter. The documentation says the fan speed affects the unit's hash rate.  But is the effect of fan speed on hash rate direct or consequential?  That is, is the fan speed setting directly used in setting the ASIC clock rates, or does a faster/slower fan simply mean cooler/warmer ASICs that subsequently mean faster/slower hashing?

The reason I ask is that my miner is in relatively cold (60F) server room and the chassis temps are holding at 21C front and 59C/59C rear.  So I was wondering if it's possible to devote less PSU power to air movement, and more power to the ASICs by reducing the fan speed. Or is such an endeavor a fool's errand?

Any sound (pardon the pun) advice on the matter would be greatly appreciated.


This might be a little off topic I appolgize.  Just wanted to warn you that you might want to switch names before you go to far. Terra Hash was a scam bitcoin maker that ran off with pre-orders.   So if anyone googles your name it has TONS of bad PR.
legendary
Activity: 1638
Merit: 1005
LOOP[2] ON TO:0 (w:79)
 4: DC2DC/1/:[vlt1:660 vlt2:664(DCl:794 Tl:664 Ul:727) 12W  18A  87c] ASIC:[ 85c (125c) 770hz(BL: 770)   35 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:724 vlt2:727(DCl:794 Tl:794 Ul:727) 106W 146A  60c] ASIC:[ 85c (125c) 1050hz(BL:1050)   44 (E:193) F:0 L:0]

What is wrong ? That ASIC was all good before tonight ... :-(

then after like 15 minutes ..

LOOP[0] OFF TO:0 (test serial failed)
0: disabled (serial problem)
1: disabled (serial problem)

same behaviour after 2-3 reboot (power off/on)

Allright ... so now I made a few tests with another power supply.

As soon as I put power in connector #4 ... 3 and 4 doesn't startup.

If I put power only in the connector 3 ... 3 is working.

Smells ... really bad (smells actually nothing ... just that I have a bad feeling)

EDIT: well in fact the PSU do not start if there a connector plugged in the connector 4. Tested with one PSU and two PSU... I think i have DC2DC chip issue ... Like ASIC 4 ... i think the DC2DC chip is overheating.

Just letting you know ... The unit will be sent as RMA. Will let you know the outcome !
newbie
Activity: 10
Merit: 0
Hello everyone. I'm a CoinTerra refugee and new to the Spondoolies camp. Just got an SP31 up and running over the weekend.

I can't seem to find any definitive explanation for setting the "Fan Speed" parameter. The documentation says the fan speed affects the unit's hash rate.  But is the effect of fan speed on hash rate direct or consequential?  That is, is the fan speed setting directly used in setting the ASIC clock rates, or does a faster/slower fan simply mean cooler/warmer ASICs that subsequently mean faster/slower hashing?

The reason I ask is that my miner is in relatively cold (60F) server room and the chassis temps are holding at 21C front and 59C/59C rear.  So I was wondering if it's possible to devote less PSU power to air movement, and more power to the ASICs by reducing the fan speed. Or is such an endeavor a fool's errand?

Any sound (pardon the pun) advice on the matter would be greatly appreciated.
newbie
Activity: 44
Merit: 0
The machines are hosted by Spondoolies at the Titan data center. Spondoolies charges $300 per machine flat rate shipping.

Anyone wants to buy 2 SP30 machines from the August group buy?

My hosting with Spondoolies has expired and I need to sell the machines or they'll be scrapped.

PM me ASAP with a price if you're interested.

It would be perfect to provide an location and approximative shipping cost for 1 or both miners Wink
legendary
Activity: 1848
Merit: 1001
both my SP20E & SP30 are STILL offline with seemingly dead MMC boards.

Multiple emails to @support have gone unanswered.

STILL waiting on the experimental SD card/firmware... STILL, BUT Now UnNeeded

STILL waiting on ribbon cables... NO LONGER, THX

STILL waiting on a reply that isnt 'karma-derp / mail support@'...

Dude, everytime I visit this thread are you having different problems or is this the same one that is just dragging out over time? Why can't it just be taken care of by SP-TECH and be done with it?



you're asking me?

ive been asking the same questions in multiple places forever, it seems.

i try not to offend (mostly) so i think some people may envision an opportunity to take the piss.

the SP30 is now running fine after eventually getting the board replaced.

i imagine the SP20E will be fine too, eventually.






legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
if a spondoolies miner is stolen - can the serial number be used to track it down?

 Undecided Did someone break into your place?

it appears that the guy in charge of colocating some of my units made a very strange series of decisions that involve the equipment of myself and several others being loaded into a uhaul without notice.

the whole things is so absurd sounding that I hope things can be sorted out as some sort of accident.

either way - curious what options are available in a situation like this

edit: got an update and seems like theres just feuding between others that my hardware got wrapped up in. should turn out okay - but it made me reconsider what options are available to protect hardware from theft. features like 'call home' when a system is moved to a different IP address could be nice, but possibly overcome by a simple factory reset or uSD recovery file
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
if a spondoolies miner is stolen - can the serial number be used to track it down?

hmm...just the kind of question you want to see from someone who has on his account
'asic wannabe"

(gets popcorn.... never saw a spondoolies 'stalker before" this could be interesting)

just kidding but I saw the 'asic wannbe with this question and it cracked me up... Smiley

heh - i was an asic wannabe back in summer 2013. still am, but have been mining since sept 2013 Smiley
legendary
Activity: 1274
Merit: 1000
Personal text my ass....
both my SP20E & SP30 are STILL offline with seemingly dead MMC boards.

Multiple emails to @support have gone unanswered.

STILL waiting on the experimental SD card/firmware...

STILL waiting on ribbon cables...

STILL waiting on a reply that isnt 'karma-derp/mail@support'...

Dude, everytime I visit this thread are you having different problems or is this the same one that is just dragging out over time? Why can't it just be taken care of by SP-TECH and be done with it?

hero member
Activity: 742
Merit: 500
Anyone wants to buy 2 SP30 machines from the August group buy?

My hosting with Spondoolies has expired and I need to sell the machines or they'll be scrapped.

PM me ASAP with a price if you're interested.

It would be perfect to provide an location and approximative shipping cost for 1 or both miners Wink
newbie
Activity: 44
Merit: 0
Anyone wants to buy 2 SP30 machines from the August group buy?

My hosting with Spondoolies has expired and I need to sell the machines or they'll be scrapped.

PM me ASAP with a price if you're interested.
copper member
Activity: 2898
Merit: 1465
Clueless!
if a spondoolies miner is stolen - can the serial number be used to track it down?

hmm...just the kind of question you want to see from someone who has on his account

'asic wannabe"

(gets popcorn.... never saw a spondoolies 'stalker before" this could be interesting)

just kidding but I saw the 'asic wannbe with this question and it cracked me up... Smiley

hero member
Activity: 882
Merit: 1003
Not track it down like radar, or last known owner, But if they contacted Spondoolies they could probably provide the date of manufacture.
legendary
Activity: 1596
Merit: 1000
if a spondoolies miner is stolen - can the serial number be used to track it down?

 Undecided Did someone break into your place?
legendary
Activity: 2128
Merit: 1005
ASIC Wannabe
if a spondoolies miner is stolen - can the serial number be used to track it down?
legendary
Activity: 2338
Merit: 1124
One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Quote
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one
I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

I _guess_ it is poor power delivery.  Right now I see it running with 4.2 th/s.

Nevertheless, the issue is that I have only ssh-access to the miner at the moment. In /etc/mnt I see the file "mg_disabled_asics" with the following content:

0:0 1:0 2:0
3:0 4:0 5:0
6:0 7:0 8:0
9:0 10:0 11:0
12:0 13:0 14:0
15:0 16:0 17:0
18:0 19:0 20:0
21:0 22:0 23:0
24:0 25:0 26:0
27:0 28:0 29:0

I assume that it should be possible to disable the ASICs here. Or am I totally wrong? And IF it is possible to disable asics, would "1" mean disabled and "2" removed?

you can disable individual asics in gui

I do know this, but I don't have access to the gui through ssh...
Why don't you have access to the gui? You could try tunnel through ssh and then enter 127.0.0.1 on your browser..

My miners are some miles away...

Thought about using corkscrew, but the connection to the miner goes through another server first, so I'm stuck at the moment. I tried to run it on the first server which would make sense if I can access start.spondoolie-tech.com on it- but I failed miserably.... will keep on trying, though...
hero member
Activity: 773
Merit: 528
One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Quote
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one
I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

I _guess_ it is poor power delivery.  Right now I see it running with 4.2 th/s.

Nevertheless, the issue is that I have only ssh-access to the miner at the moment. In /etc/mnt I see the file "mg_disabled_asics" with the following content:

0:0 1:0 2:0
3:0 4:0 5:0
6:0 7:0 8:0
9:0 10:0 11:0
12:0 13:0 14:0
15:0 16:0 17:0
18:0 19:0 20:0
21:0 22:0 23:0
24:0 25:0 26:0
27:0 28:0 29:0

I assume that it should be possible to disable the ASICs here. Or am I totally wrong? And IF it is possible to disable asics, would "1" mean disabled and "2" removed?

you can disable individual asics in gui

I do know this, but I don't have access to the gui through ssh...
Why don't you have access to the gui? You could try tunnel through ssh and then enter 127.0.0.1 on your browser..
legendary
Activity: 1484
Merit: 1004
Any news from the conference from spondoolies ??  Smiley
legendary
Activity: 2338
Merit: 1124
One of my SP31 is not behaving very nicely. Actually, it is hashing with a maximum of about 1 Th. I have access via ssh, and looking in the logs, I see the following - well, it asks me to disable "one by one". How can this be done through ssh? Or has anybody an advice what I should do?

Quote
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 15: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 16: Loop RT loss - disable us one-by-one to find the bad one
Mar 28 00:06:14 miner local1.warn minergate[1737]: Critical: Asic disable 17: Loop RT loss - disable us one-by-one to find the bad one
I have one like that where it seems that its only stable once i manually disable at least one asic from the 'trouble loop', and it generally disables 2-3 more for failing BIST, then becomes reasonably stable.

go to settings, and one of the lower categories allows you to DISABLE specific chips. disable 1-2 at a time fromt he trouble loop, as the issue is sometimes a chip and sometimes poor power delivery

I _guess_ it is poor power delivery.  Right now I see it running with 4.2 th/s.

Nevertheless, the issue is that I have only ssh-access to the miner at the moment. In /etc/mnt I see the file "mg_disabled_asics" with the following content:

0:0 1:0 2:0
3:0 4:0 5:0
6:0 7:0 8:0
9:0 10:0 11:0
12:0 13:0 14:0
15:0 16:0 17:0
18:0 19:0 20:0
21:0 22:0 23:0
24:0 25:0 26:0
27:0 28:0 29:0

I assume that it should be possible to disable the ASICs here. Or am I totally wrong? And IF it is possible to disable asics, would "1" mean disabled and "2" removed?

you can disable individual asics in gui

I do know this, but I don't have access to the gui through ssh...
Jump to: