Pages:
Author

Topic: Unofficial Spondoolies SP20 thread - page 18. (Read 126204 times)

hero member
Activity: 544
Merit: 507
April 03, 2015, 07:20:14 AM
Hi,

Sort of new at this.  I have 2 units and I seem to can't get a good hash/temp ratio.  Both units are hitting 120 on one loop.  I want the most hashing power while keeping the temps below 120.  Any recommendations?





Thanks,

Vince

those rates seems pretty good to me. The key is the inlet air temp. You have 31° so it's quite nice for them.

1,6 TH! I wish i could make them that high.

Do you have them on A/A?
newbie
Activity: 3
Merit: 0
April 03, 2015, 12:43:48 AM
Hi,

Sort of new at this.  I have 2 units and I seem to can't get a good hash/temp ratio.  Both units are hitting 120 on one loop.  I want the most hashing power while keeping the temps below 120.  Any recommendations?

http://s14.postimg.org/lu12hxx35/Screen_Shot_2015_04_03_at_1_40_51_AM.png

http://s14.postimg.org/3v1t6z6pt/Screen_Shot_2015_04_03_at_1_40_41_AM.png

Thanks,

Vince
legendary
Activity: 1638
Merit: 1005
March 30, 2015, 11:32:40 AM
Just letting you know,

My unit will sent as RMA tomorow.

Looks like i have a faulty blade.

Will let you know the outcome !
full member
Activity: 224
Merit: 100
March 29, 2015, 12:13:14 PM
Is it possible to clock Loop 3 lower than 0-2 ? Loop 3 always runs hotter and I'd like to run it a bit lower as the fan Auto setting adjusts itself based on how hot or cold Loop 3 is running. Would like to try to get the fans running quieter without it varying its speed as much as it is now (ranges from 6-39 depending)

 Is it possible to set a Voltage Limit per loop instead of having be applied globally ?

 Could that be done by limiting the Max PSU Power on Unit 4, it's safe, and the SP20 will try to run as fast as it can with what power is available to that loop ?

 Eg: Have Max Power for each unit set to 200,200,200,100 and that's perfectly acceptable ?




That works fine.  I clock the rear loops on all my SP20s about 20% less than the front loops to balance out temps.
legendary
Activity: 1652
Merit: 1067
Christian Antkow
March 29, 2015, 11:59:02 AM
 Is it possible to clock Loop 3 lower than 0-2 ? Loop 3 always runs hotter and I'd like to run it a bit lower as the fan Auto setting adjusts itself based on how hot or cold Loop 3 is running. Would like to try to get the fans running quieter without it varying its speed as much as it is now (ranges from 6-39 depending)

 Is it possible to set a Voltage Limit per loop instead of having be applied globally ?

 Could that be done by limiting the Max PSU Power on Unit 4, it's safe, and the SP20 will try to run as fast as it can with what power is available to that loop ?

 Eg: Have Max Power for each unit set to 200,200,200,100 and that's perfectly acceptable ?


legendary
Activity: 1456
Merit: 1000
March 27, 2015, 11:42:37 PM
After getting my first Antminers up and running over the last few days, I can't say how much I miss SP20s being available.  Sad

Ain't that the truth. Spondoolies are really in a league of their own. Everything else is Fisher Price.

Really? try to sell here a used SP20-SP30 for anything more than a direct Th comparison with bitmain and this forum will bore you with complains about price.
So, this argument is not working here- a presumable place of mining connoisseurs-and it's just a fact of life.

You might try Ebay.   The boards are good for selling quick on hardware, but you will selling price will be beaten down pretty low.

I have yet to sell a miner here as I got more elsewhere.  Yes you do have to deal with ebay, paypal, etc.  And even in some cases teach someone how to mine from scratch.  But so far I always got more money this way.  If you do a "make a offer" option with high price it allows you to look at buyers offer and feedback and pick one you like.
legendary
Activity: 3892
Merit: 4331
March 27, 2015, 10:46:50 PM
After getting my first Antminers up and running over the last few days, I can't say how much I miss SP20s being available.  Sad

Ain't that the truth. Spondoolies are really in a league of their own. Everything else is Fisher Price.

Really? try to sell here a used SP20-SP30 for anything more than a direct Th comparison with bitmain and this forum will bore you with complains about price.
So, this argument is not working here- a presumable place of mining connoisseurs-and it's just a fact of life.
legendary
Activity: 1456
Merit: 1000
March 27, 2015, 09:39:52 PM
After getting my first Antminers up and running over the last few days, I can't say how much I miss SP20s being available.  Sad

Ain't that the truth. Spondoolies are really in a league of their own. Everything else is Fisher Price.

They do make quality gear no one can take that away from them.   I love solid metal miners.
legendary
Activity: 1596
Merit: 1000
March 27, 2015, 09:09:23 PM
After getting my first Antminers up and running over the last few days, I can't say how much I miss SP20s being available.  Sad

Ain't that the truth. Spondoolies are really in a league of their own. Everything else is Fisher Price.
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
March 27, 2015, 07:34:07 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

Ahhh ok that is better then I thought then.   What PSU do you currently have on trying to power the SP20?

I normally use a Corsair RM 1000. I tried to slipt the load with an Corsair CX750. (RM on connector 1 and 2 + CX750 on the connector 3 and 4).

The PSU and the unit won't start if I plug a PCI-e in position 4.

So I'm pretty sure it's a physical problem on the blade ...

EDIT: I can see the PSU trying to start but it doesn't ...
EDIT2: As soon asI only plug 1-2-3, both setup works (Split PSU or just the RM 1000)
EDIT3: Anyway check the ASIC4 ... DC2DC chip is overheating for nothing and won't clock up either ... sucksssssssss

what speeds did you use with this one?  1300 or less ?    1400 or more.

I am doing a survey on dead or dyeing sp20's  .   I want to know   the death rate for :

under     1300
between 1300-1400
over       1400

please let me know.  thanks


Around 1550. so over 1400
  thanks
legendary
Activity: 1638
Merit: 1005
March 27, 2015, 06:44:17 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

Ahhh ok that is better then I thought then.   What PSU do you currently have on trying to power the SP20?

I normally use a Corsair RM 1000. I tried to slipt the load with an Corsair CX750. (RM on connector 1 and 2 + CX750 on the connector 3 and 4).

The PSU and the unit won't start if I plug a PCI-e in position 4.

So I'm pretty sure it's a physical problem on the blade ...

EDIT: I can see the PSU trying to start but it doesn't ...
EDIT2: As soon asI only plug 1-2-3, both setup works (Split PSU or just the RM 1000)
EDIT3: Anyway check the ASIC4 ... DC2DC chip is overheating for nothing and won't clock up either ... sucksssssssss

what speeds did you use with this one?  1300 or less ?    1400 or more.

I am doing a survey on dead or dyeing sp20's  .   I want to know   the death rate for :

under     1300
between 1300-1400
over       1400

please let me know.  thanks


Around 1550. so over 1400
full member
Activity: 224
Merit: 100
March 27, 2015, 06:24:40 PM
After getting my first Antminers up and running over the last few days, I can't say how much I miss SP20s being available.  Sad
legendary
Activity: 4256
Merit: 8551
'The right to privacy matters'
March 27, 2015, 05:33:12 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

Ahhh ok that is better then I thought then.   What PSU do you currently have on trying to power the SP20?

I normally use a Corsair RM 1000. I tried to slipt the load with an Corsair CX750. (RM on connector 1 and 2 + CX750 on the connector 3 and 4).

The PSU and the unit won't start if I plug a PCI-e in position 4.

So I'm pretty sure it's a physical problem on the blade ...

EDIT: I can see the PSU trying to start but it doesn't ...
EDIT2: As soon asI only plug 1-2-3, both setup works (Split PSU or just the RM 1000)
EDIT3: Anyway check the ASIC4 ... DC2DC chip is overheating for nothing and won't clock up either ... sucksssssssss

what speeds did you use with this one?  1300 or less ?    1400 or more.

I am doing a survey on dead or dyeing sp20's  .   I want to know   the death rate for :

under     1300
between 1300-1400
over       1400

please let me know.  thanks
legendary
Activity: 1638
Merit: 1005
March 27, 2015, 05:17:48 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

Ahhh ok that is better then I thought then.   What PSU do you currently have on trying to power the SP20?

I normally use a Corsair RM 1000. I tried to slipt the load with an Corsair CX750. (RM on connector 1 and 2 + CX750 on the connector 3 and 4).

The PSU and the unit won't start if I plug a PCI-e in position 4.

So I'm pretty sure it's a physical problem on the blade ...

EDIT: I can see the PSU trying to start but it doesn't ...
EDIT2: As soon asI only plug 1-2-3, both setup works (Split PSU or just the RM 1000)
EDIT3: Anyway check the ASIC4 ... DC2DC chip is overheating for nothing and won't clock up either ... sucksssssssss
legendary
Activity: 1456
Merit: 1000
March 27, 2015, 05:04:12 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

Ahhh ok that is better then I thought then.   What PSU do you currently have on trying to power the SP20?
legendary
Activity: 1638
Merit: 1005
March 27, 2015, 03:39:06 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)

There is no smell at all, I was just saying I have a bad feeling about this ...

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.
legendary
Activity: 1456
Merit: 1000
March 27, 2015, 03:37:54 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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

Smell scares me personally.   I can say I have had good luck with RM1000's on other miners but I have never used on SP20.

I would ask SP support if you can open it up to see if something is melting or what's causing the smell.  The reason I say ask them is I don't  want you to lose warranty from opening without their consent. (This could have changed but I'm like 75 percent sure opening on your own equals lost warranty, but I could be wrong)
legendary
Activity: 1638
Merit: 1005
March 27, 2015, 03:31:35 PM
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)

Looks like my RM1000 is not able to take the load anymore... which explain the loop 0 failed.

But still one ASIC is acting strangely ? Any idea ?

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
legendary
Activity: 1848
Merit: 1001
March 22, 2015, 09:48:22 PM
[We don't cut corners (e.g. embedded controller, etc) and we use dc2dc which means that the BOM is higher.]

If you don't cut corners then why didn't you install an internal microSD card Huh

I had to go buy one in an attempt to revive several DOA SP20s. I had to go to 3 stores before I found a card and reader ($19) for this thing I'll never use for any other purpose.

AntMiners have an internal SD card so one can reset factory settings by pressing a button.

VERY good point!!

Also, why do you (SP-T) fail to reply to multiple @support mails for weeks on end & 'secretly/privately' offer to ship experimental SD cards for $35.00USD while publicly offering to post the firmware for free but never actually doing it???

its all a bit peculiar...

alh
legendary
Activity: 1846
Merit: 1052
March 22, 2015, 05:13:25 PM
Thank you for that kind advice. I don't have to pay for electricity though.
Whats more interesting to me, is to get the best performance with the lowest noise.

Without some significant cooling changes to the SP20, you will be forced to reduce the hash rate in order to keep the fan speed down. You can try replacing the fan with a quieter one, but the SP20 design requires a fair amount well pressured airflow to keep cool. Generally that means that a quieter fan will not cool as well as the stock fan. With the power cables at one end, there doesn't seem to be a good way to work out a push pull configuration. Other options are to "drill out" part of the of the honeycomb, though some folks claim it's really the fan itself that makes the noise. Another option is to reverse the airflow, though that then pushes warm air over the PCIe cables.

Best of luck in finding a combination of quiet operation with a high hashrate.
Pages:
Jump to: