Big thanks to Guy for the help with my small issue. As always, SPT is the best in post sales support.
You know where to vote 
Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 173. (Read 1260290 times)
January 29, 2015, 10:34:22 AM
January 29, 2015, 10:32:41 AM
Big thanks to Guy and the team for the help with my small issue. As always, SPT is the best in post sales support.
January 29, 2015, 08:33:26 AM
I completely agree with your opinion. In fact I would say when comparing the sound issue to the sp-20 and the s-5 the s-5 with mods can be made quieter then the sp-20. I have a thread that shows mods under 20 bucks to make the s-5 pretty quiet. If you are willing to do mods the s-5 is easier to lower the sound.
That's nothing inherent in the SP20 though, the S5 just has about half the power density as the SP20 so it takes a lot more air to cool. If you were to put the SP20 in a double wide case, you could considerably reduce the noise for a couple reasons. First off, you could use two slower fans to move the same amount of air as the singe 120mm fan with a much better acoustical profile, which would help quite a bit. Second, the lower velocity with dual fans would decrease turbulence and grill noise, further helping it.Lastly, running four boards with two ASICs each would not only stop having the second ASIC running on hot air, it would allow you to use cheap heatpiped CPU coolers. Decreasing the thermal resistance between the die and the air can make a huge difference, as you then need a lot less airflow to maintain the same die temps.
I'm genuinely curious why Syke's original point that the SP20 could not be made much quieter without watercooling by decreasing density. I haven't noticed any coil whine or anything specific to the boards that would lead me to that conclusion.
That's $800 of cooling gear for a $400 miner, as well as spagetification 
Maybe it would be viable with 'northbridge' off the shelf blocks and chain them all with a syscooling level [$50] kit, although even then we're talking 8x $25 + $50 
Yeah, it was more just a joke. There's no need to use watercooling when talking about dissipating 100W/chip. Maybe it would be viable with 'northbridge' off the shelf blocks and chain them all with a syscooling level [$50] kit, although even then we're talking 8x $25 + $50 
January 29, 2015, 07:42:51 AM
You don't need water cooling to have quiet. You just need to stop designing for maximum density. Spread the chips out and use nice quiet slow big fans. That's really all water cooling is.
If the SP20 was twice as large (same number of chips, without putting 2 chips in line of the airflow) it would run much better and quieter with 2 fans.
If the SP20 was twice as large (same number of chips, without putting 2 chips in line of the airflow) it would run much better and quieter with 2 fans.
I agree with the sentiment on spreading chips out, in every product other than the SP20 though. The SP20 has other issues / coexisting hurdles required to get it silent at stock speeds which wouldn't necessarily be solved with a decrease in density. On the SP10, SP30, SP31, SP35 though, almost certainly. Of those only the SP10's fans were actually loud, everything else was PSU noise from using 1U PSUs.
That's classified information
Doesn't matter his response, what matters is: is there any chance of watercooled Spondoolies miner?!?!?!
There are significant challenges to a watercooled SP20.
Primarily, I'm too busy to set it all up.
Cable ties are your friend
That's $800 of cooling gear for a $400 miner, as well as spagetification
Maybe it would be viable with 'northbridge' off the shelf blocks and chain them all with a syscooling level [$50] kit, although even then we're talking 8x $25 + $50 I completely agree with your opinion. In fact I would say when comparing the sound issue to the sp-20 and the s-5 the s-5 with mods can be made quieter then the sp-20. I have a thread that shows mods under 20 bucks to make the s-5 pretty quiet. If you are willing to do mods the s-5 is easier to lower the sound.
January 29, 2015, 07:05:04 AM
You don't need water cooling to have quiet. You just need to stop designing for maximum density. Spread the chips out and use nice quiet slow big fans. That's really all water cooling is.
If the SP20 was twice as large (same number of chips, without putting 2 chips in line of the airflow) it would run much better and quieter with 2 fans.
If the SP20 was twice as large (same number of chips, without putting 2 chips in line of the airflow) it would run much better and quieter with 2 fans.
I agree with the sentiment on spreading chips out, in every product other than the SP20 though. The SP20 has other issues / coexisting hurdles required to get it silent at stock speeds which wouldn't necessarily be solved with a decrease in density. On the SP10, SP30, SP31, SP35 though, almost certainly. Of those only the SP10's fans were actually loud, everything else was PSU noise from using 1U PSUs.
That's classified information
Doesn't matter his response, what matters is: is there any chance of watercooled Spondoolies miner?!?!?!
There are significant challenges to a watercooled SP20.Primarily, I'm too busy to set it all up.
Cable ties are your friend
That's $800 of cooling gear for a $400 miner, as well as spagetification
Maybe it would be viable with 'northbridge' off the shelf blocks and chain them all with a syscooling level [$50] kit, although even then we're talking 8x $25 + $50 
January 29, 2015, 06:46:46 AM
Apologies if this has been asked before..I have 2 SP20s that I'm having an issue enabling API access on. When I add my monitoring server ip into the cgminer.conf file, save it and restart minergate, the ip that I've set disappears from the cgminer.conf file. This is only happening on 2 of them, the other 3 work fine. Any suggestions?
To edit this file, please use SSH to log in to your miner, and then use vi /etc/cgminer.conf.template
Thanks, that worked!
January 29, 2015, 06:22:01 AM
My predictions, three months ago: Bitcoin TLV `14, #25 - Guy Corem - Mining Forecast
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
Nice video. That forecast did not include the bitcoin exchange rate, which is the most uncertain variable for mining as a business: Racks of machine can become paperweight when bitcoin price dive below $100
Home mining however are not affected as much since many of them are not profit oriented
January 29, 2015, 05:10:46 AM
My predictions, three months ago: Bitcoin TLV `14, #25 - Guy Corem - Mining Forecast
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
Nice video. That forecast did not include the bitcoin exchange rate, which is the most uncertain variable for mining as a business: Racks of machine can become paperweight when bitcoin price dive below $100
Home mining however are not affected as much since many of them are not profit oriented
January 29, 2015, 04:18:52 AM
Apologies if this has been asked before..I have 2 SP20s that I'm having an issue enabling API access on. When I add my monitoring server ip into the cgminer.conf file, save it and restart minergate, the ip that I've set disappears from the cgminer.conf file. This is only happening on 2 of them, the other 3 work fine. Any suggestions?
To edit this file, please use SSH to log in to your miner, and then use vi /etc/cgminer.conf.template
January 29, 2015, 04:14:49 AM
My predictions, three months ago: Bitcoin TLV `14, #25 - Guy Corem - Mining Forecast
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
BitFury revealed their mining address two days afterwards. I didn't expect to see AntPool grow so much.
Let's see how the rest of my predictions will be in 2015. It was based of BTC value of $400
The rest of Inside Bitcoin TLV lectures can be found here: Inside Bitcoins Tel Aviv 2014t
I think that it was one of the best Inside Bitcoin conferences ever. Kudos to Meni Rosenfeld.
Nice video. That forecast did not include the bitcoin exchange rate, which is the most uncertain variable for mining as a business: Racks of machine can become paperweight when bitcoin price dive below $100
Home mining however are not affected as much since many of them are not profit oriented
January 29, 2015, 12:06:45 AM
Apologies if this has been asked before..I have 2 SP20s that I'm having an issue enabling API access on. When I add my monitoring server ip into the cgminer.conf file, save it and restart minergate, the ip that I've set disappears from the cgminer.conf file. This is only happening on 2 of them, the other 3 work fine. Any suggestions?
January 28, 2015, 08:02:36 PM
you don't get cfm for free
Ain't that the truth!
January 28, 2015, 07:59:26 PM
thanks!
I would like to try a 38mm PWM Delta, but I doubt they will be quieter than the 1amp "JDH1238B" (also 38mm) I took from the S3.
These bigger Deltas are most commonly server fans with high cfm and they don't care about noise.
If someone can point me to a 38mm PWM 12V Delta with not more than 1 to 1.2 amp, I appreciate.
the probably 1:1 fitting connectors 1.5mm spacing (I posted an ebay link) are still on the way from china to me, because I could not find another source.
In the meantime I cut the original fan wires and inserted a standard fan connector, only way not to touch the controller board.
I would like to try a 38mm PWM Delta, but I doubt they will be quieter than the 1amp "JDH1238B" (also 38mm) I took from the S3.
These bigger Deltas are most commonly server fans with high cfm and they don't care about noise.
If someone can point me to a 38mm PWM 12V Delta with not more than 1 to 1.2 amp, I appreciate.
the probably 1:1 fitting connectors 1.5mm spacing (I posted an ebay link) are still on the way from china to me, because I could not find another source.
In the meantime I cut the original fan wires and inserted a standard fan connector, only way not to touch the controller board.
if you are going to chop off the connector anyway, then this is the best deal $$ wise (two for $20 with nonstandard connector)
http://www.ebay.com/itm/Delta-AFB1212SHE-Computer-Case-Power-Fan-120X120mm-Lot-of-2-/191472588964?pt=US_Computer_Case_Fans&hash=item2c94a7d8a4
if you want with standard PWM, then
http://www.ebay.com/itm/Delta-AFB1212SHE-PWM-120mm-x-38mm-4Pin-PWM-Hi-Speed-3700-RPM-151CFM-Case-Fan-/231318294284?pt=LH_DefaultDomain_0&hash=item35dba50f0c
or
http://www.amazon.com/Delta-Electronics-AFB1212SHE-120x120x38mm-connector/dp/B004Y1HLA8/ref=sr_1_8?s=pc&ie=UTF8&qid=1422465408&sr=1-8&keywords=delta+120mm+fan
these fans use 1.06A
from what I read they are 1.6A and therefor too loud for sure. anyone already tried?
you don't get cfm for free
January 28, 2015, 07:09:26 PM
ADD: Confirmed. Time in schedule needs to be UTC
I added this "problem" already into the notes field of the google sheets sp20 firmware document....
January 28, 2015, 05:47:00 PM
How exactly is schedule supposed to work?
I set it to lower the voltage in 10 minutes for example and nothing happens when the time is up.
Current firmware version: 2.6.10
I set it to lower the voltage in 10 minutes for example and nothing happens when the time is up.
Current firmware version: 2.6.10
Been working fine for me, set the times you want the changes to occur, here I have cheap rate 00:00 to 07:00 so I set it accordingly
At the allotted time it changes as you can see from the speed graph over the last few days, fast during the night, slow during the day.
@zvisha: can the fan be adjusted at the same time? I want to run at 5-10% in the day but at night i don't care if it runs at 60%
also @zvisha: is there a minimum speed for the fan? I run at 4-5% without an issue but am a bit worried that if i used 3% the fan might come to a stop when im not looking. alternatively, can the auto-fan go as low as 3% if chip temps are all 85/100C ?
January 28, 2015, 04:48:47 PM
making another post as this is not related to the previous one:
What does this mean:
What does this mean:
Code:
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 0: 4 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 at freq 890
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1: sl:1 sd:1, rate:890
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 3 at freq 990
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 3: sl:1 sd:1, rate:990
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 2 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 1: 2 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0 at freq 910
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0: sl:1 sd:1, rate:910
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 2: 1 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 4 at freq 920
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 4: sl:1 sd:1, rate:920
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 3: 1 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: Do BIST LOOP- DONE BIST_PERIOD_SECS
Jan 28 21:46:53 miner local1.warn minergate[498]: Bist took 39748 usec
I have a huge lot of these Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 at freq 890
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1: sl:1 sd:1, rate:890
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 3 at freq 990
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 3: sl:1 sd:1, rate:990
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 2 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 1: 2 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0 at freq 910
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0: sl:1 sd:1, rate:910
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 2: 1 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 4 at freq 920
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 4: sl:1 sd:1, rate:920
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 1 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: BIST 3: 1 asics down after bist
Jan 28 21:46:53 miner local1.warn minergate[498]: FAILED BIST 0 ASICS
Jan 28 21:46:53 miner local1.warn minergate[498]: Do BIST LOOP- DONE BIST_PERIOD_SECS
Jan 28 21:46:53 miner local1.warn minergate[498]: Bist took 39748 usec
January 28, 2015, 04:15:10 PM
This is the reason it works as expected. I just checked the Events section and the time there is still UTC while I set it to my time zone in the Settings. So I just have to select the times to be in UTC.
Testing now
ADD: Confirmed. Time in schedule needs to be UTC
Testing now
ADD: Confirmed. Time in schedule needs to be UTC
2.6.9 on mine.
My timezone is set to Europe/London which is the same as UTC
My timezone is set to Europe/London which is the same as UTC
January 28, 2015, 04:13:27 PM
2.6.9 on mine.
My timezone is set to Europe/London which is the same as UTC
My timezone is set to Europe/London which is the same as UTC
January 28, 2015, 04:07:16 PM
exactly what I did. then I did a couple of test runs to make sure... nope
What is your FW version?
Did you change the time zone in the settings? I did...
What is your FW version?
Did you change the time zone in the settings? I did...
January 28, 2015, 04:00:38 PM
How exactly is schedule supposed to work?
I set it to lower the voltage in 10 minutes for example and nothing happens when the time is up.
Current firmware version: 2.6.10
I set it to lower the voltage in 10 minutes for example and nothing happens when the time is up.
Current firmware version: 2.6.10
Been working fine for me, set the times you want the changes to occur, here I have cheap rate 00:00 to 07:00 so I set it accordingly
At the allotted time it changes as you can see from the speed graph over the last few days, fast during the night, slow during the day.
Jump to: