Author

Topic: [ANN] Spondoolies-Tech - carrier grade, data center ready mining rigs - page 380. (Read 1260354 times)

legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform

Ahummmm is this bad news ??


Uptime:2366 | FPGA ver:13
-----TOP-BOARD-----
PSU-TOP[NONE]: 0->0w[0 0 0] (lim=1360) 0c cooling:0/0x27000000
LOOP[0] OFF (no PSU)
0: disabled (i2c BAD, btw!)
1: disabled (i2c BAD, btw!)...

A breaker might have tripped. It's also possible that one of your PSUs overheated and got stuck in the overheated state -- that happens on occasion too.

If the PSU has a flashing orange LED going, then a full power cycle (unplug both PSUs, wait 30 seconds (this actually matters! Don't wait only 10 seconds, as that isn't enough), plug them back in) should fix it. If the LED is dark, then your PSU isn't getting powered, and you should check the cables and the circuit breakers.

Please everybody read the red part - important!

Edit:

I will change the configuration that we ship the miners to be 1300W by default.
This way the clients will get the units running at safe rate, without failures, but it will take about 1%-2% of their maximum hash rate. After the client made sure the miner is working, and he has access to the miner for a few hours, he can increase it up to 1350W in settings page and let the system learn.
seems alot more stable with the newest firmware
hero member
Activity: 818
Merit: 1006

I will change the configuration that we ship the miners to be 1300W by default.
This way the clients will get the units running at safe rate, without failures, but it will take about 1%-2% of their maximum hash rate. After the client made sure the miner is working, and he has access to the miner for a few hours, he can increase it up to 1350W in settings page and let the system learn.

I suggest something higher than 1300. A sample of some of the machines we received earlier today, after 10 hours and having stabilized their power consumption:

Hashrate, Top max watts, Bot max watts
4559 1345 1350
4431 1340 1350
4507 1335 1350
4535 1340 1355
4458 1350 1340
4470 1340 1345
4568 1360 1350
4670 1350 1340 (record holder, fastest I've seen)
4565 1350 1350
4461 1340 1345
4294 1355 1350 (LOOP[7] OFF (no data on loop) -- fixed by a power cycle, now 4521 1355 1350)
4530 1340 1345
4564 1345 1355
4589 1345 1360
4553 1345 1335
4494 1345 1345
4553 1360 1350
4549 1340 1345
4455 1340 1335

We probably have better cooling and power in our facility than most people, but even so, I would guess most SP30s in most buildings should be able to handle 1320W or 1330W as a default starting point. I actually vote for keeping the starting point at 1360W, because I think that the overclocking algorithm is awesome. But that's just me.
full member
Activity: 238
Merit: 100
Wanna buy a Tesla? Visit TeslaBargain.com first!
I never said out next product is the SP50. Who knows what kind of goodies we have lined up for you?
With 14nm out of reach, I predict 20nm, at least 6TH, lower power consumption, lower price, less noise, named SP40 (with SP50 reserved for the <20nm generation). Smiley
hero member
Activity: 560
Merit: 500

Ahummmm is this bad news ??


Uptime:2366 | FPGA ver:13
-----TOP-BOARD-----
PSU-TOP[NONE]: 0->0w[0 0 0] (lim=1360) 0c cooling:0/0x27000000
LOOP[0] OFF (no PSU)
0: disabled (i2c BAD, btw!)
1: disabled (i2c BAD, btw!)...

A breaker might have tripped. It's also possible that one of your PSUs overheated and got stuck in the overheated state -- that happens on occasion too.

If the PSU has a flashing orange LED going, then a full power cycle (unplug both PSUs, wait 30 seconds (this actually matters! Don't wait only 10 seconds, as that isn't enough), plug them back in) should fix it. If the LED is dark, then your PSU isn't getting powered, and you should check the cables and the circuit breakers.

Please everybody read the red part - important!

Edit:

I will change the configuration that we ship the miners to be 1300W by default.
This way the clients will get the units running at safe rate, without failures, but it will take about 1%-2% of their maximum hash rate. After the client made sure the miner is working, and he has access to the miner for a few hours, he can increase it up to 1350W in settings page and let the system learn.
legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform
My box keep rebooting as well its been up and down since the latest firmware


I see some people having this issue. It is related to systems that boot slowly, for example blocked NTP, and watchdog restarts them. I released version 2.3.40 that solved the issue to a client. Please try and see if it helps. I will get more into this issue tomorrow, write if it helped.

Edit**
spoke too soon, same issue still with the different PDU

It wasn't the NTP issue on my end I have a CISCO PDU that seemed to be the issue my WU rate was also low in the 20s on my pool.

Put the both PSU's on a different strip and it seems to be more stable.

I did check the SSH with t hat command to test for NTP issues and that seems not to be it since I did not get the alarm.

One thing I noticed on the previous PDU was wild voltage swings from 230-245 on the PSU's on the CISCO whereas the ADP holds both at 233-232
hero member
Activity: 560
Merit: 500
My box keep rebooting as well its been up and down since the latest firmware


I see some people having this issue. It is related to systems that boot slowly, for example blocked NTP, and watchdog restarts them. I released version 2.3.40 that solved the issue to a client. Please try and see if it helps. I will get more into this issue tomorrow, write if it helped.
hero member
Activity: 818
Merit: 1006
Hey, we've got a p2pool node up for testing. Anyone want to try it out?

74.82.237.66:9332

0% fees, 0% author contribution, hosted in central WA on one of our 100 Mbps symmetrical fiber links. IP address will change eventually, and we might be rebooting it, so don't expect it to be reliable just yet.

We now have around 173 TH/s running on our p2pool node, and our server is barely keeping up with the load. (GetBlockTemplate latencies are getting a bit high, and orphan rates aren't as low as they should be). We should have a new server in a few days with 2.5x faster CPUs, 4x as much RAM, and SSDs. Until then, we probably shouldn't add any more load to the machine.

In related news, the Toomim Brothers' 37 SP30s from the August group buy arrived today. FedEx Freight was a bit slow. They're all running now.
hero member
Activity: 818
Merit: 1006
I have lost power in one of my PSUs in my SP30. There is no green light and hashing went down to 2TH/s. Have you got any ide what is wrong?

I'm going to open up teh case and check up the cables inside

Swap the power cables on the PSUs first. It might be your power strip or extension cord that got blown out, or it might be a breaker that tripped.
hero member
Activity: 818
Merit: 1006

Ahummmm is this bad news ??


Uptime:2366 | FPGA ver:13
-----TOP-BOARD-----
PSU-TOP[NONE]: 0->0w[0 0 0] (lim=1360) 0c cooling:0/0x27000000
LOOP[0] OFF (no PSU)
0: disabled (i2c BAD, btw!)
1: disabled (i2c BAD, btw!)...

A breaker might have tripped. It's also possible that one of your PSUs overheated and got stuck in the overheated state -- that happens on occasion too.

If the PSU has a flashing orange LED going, then a full power cycle (unplug both PSUs, wait 30 seconds (this actually matters! Don't wait only 10 seconds, as that isn't enough), plug them back in) should fix it. If the LED is dark, then your PSU isn't getting powered, and you should check the cables and the circuit breakers.
hero member
Activity: 818
Merit: 1006
Edit: Noob question. Can I power the SP30 with regular PSU power cords that have 10A on them? I see that the cables SP-Tech used for PSUs have 15A. By my math plugging the PSUs on 220V should only require ~7A per PSU, but I want to be sure. Thx.

10A cables are fine if you're on 200+V. If you're on 120V, though, they're somewhat sketchy. I used some 10A cables on 120V back in the day without too much trouble, though.
sr. member
Activity: 512
Merit: 250

I dont know about the noise being lower it sounds like a leafblower , I'm going to need to do some sort of sound proofing or silent cabinet.  Running two would be an insane amount of noise.
It is by no way silent, but one should be able to run it in a cellar with little to no sound-proofing required.

I also saw a community effort for a silent cabinet some months ago, I think by the same guy who built the open graphic card rigs.

It was fairly cheap (compared to the ludicrous amounts some companys charge) and had a very decent sound reduction.


+1   Roll Eyes
legendary
Activity: 1098
Merit: 1000
Thanks, you can apply them to the SP50 now Smiley

I never said out next product is the SP50. Who knows what kind of goodies we have lined up for you?

Gadi

Gadi, you tease Smiley

Whatever it is I am now number 1 in the order queue !!
legendary
Activity: 1904
Merit: 1007
My box keep rebooting as well its been up and down since the latest firmware

Do you have NTP protocol blocked by chance? Check this: https://bitcointalksearch.org/topic/m.8674302
legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform
My box keep rebooting as well its been up and down since the latest firmware
sr. member
Activity: 462
Merit: 250
I have lost power in one of my PSUs in my SP30. There is no green light and hashing went down to 2TH/s. Have you got any ide what is wrong?

I'm going to open up teh case and check up the cables inside
donator
Activity: 1414
Merit: 1051
Spondoolies, Beam & DAGlabs
Due to a technical restriction the discount coupons are valid only for SP31. Should you choose to hang on to them for future products please contact us and we will replace the coupon.

Thanks, you can apply them to the SP50 now Smiley

I never said out next product is the SP50. Who knows what kind of goodies we have lined up for you?

Gadi
hero member
Activity: 572
Merit: 500
What are you guys's hashrate? My sp30 runs at around 4600 Ghps 24h average, is that below, above or precisely average?

I'm at 4502.69Ghs rock solid, after the initial PSU capability detection Smiley  I finally retrieved the miner yesterday evening from DHL so I can't tell you the 24h average. Wiring it was fun .. now it runs on 214V (under load). The temps are 32 °C / 80,78 °C.  After I powered it up, it took under 10 minutes to have it hashing (including firmware updates).
sr. member
Activity: 473
Merit: 250
All of the August customers who requested a coupon were sent one.

We are now going over the September list.

Gadi

Can I ask in this thread as I'm sure other will be interested too.

I have my vouchers thanks, it says they are valid for 6 months, are they restricted to the SP30/31 ?

I'd probably take a gamble and wait for what you have coming next if they aren't.



Due to a technical restriction the discount coupons are valid only for SP31. Should you choose to hang on to them for future products please contact us and we will replace the coupon.


I am thinking of purchasing an extra coupon (giftcard) from someone.  Will I be able to apply multiple gift cards on a single order?
hero member
Activity: 560
Merit: 500
Hi all
The latest SP30 version will not boot if your NTP protocol disabled by the firewall.
Please make sure your NTP not blocked.

If you see miner rebooting all the time, do the following:
- disconect ethernet
- let the miner boot till the yellow led blinks twice every second
- connect ethernet
- ssh to miner
- type "ntpd -d -n -q -p 1.openwrt.pool.ntp.org"
- If you see the text "Alarm clock" - your ntp is blocked. Open it.

I will prepare a fix but it will not be this weekend, sorry.

 
legendary
Activity: 2408
Merit: 1102
Leading Crypto Sports Betting & Casino Platform
What are you guys's hashrate? My sp30 runs at around 4600 Ghps 24h average, is that below, above or precisely average?


That is wonderful, seems you have +5%   Wink

mine runs at the spec 4500 + or - 50GHZ

i cant get it to use over 1355 watts though otherwise I could get it to run much higher, not sure why
Jump to: