Topic: Unofficial Spondoolies SP20 thread - page 23. (Read 126198 times)

maker and model of the psu? must be single rail, must use 16awg wires on the 6pin!!!
psu shutting down it's usualy because of too much load on it; might not the best psu and it needs an consumer on the 5v rail...
etc

Looking for any ideas on this ... I have an SP20 that is being troublesome ... Set to max settings under the basic settings tab.

It works fine for 10 minutes and then shuts off the PSU or PSUs, have tried various PSU configs:

4 x 550
2 X 850
1 x 1350 + 1 x 850
2 x 850 + 2 x 550
1 x 1350 + 2 x 850

I'm guessing its not looking good, but any ideas would be welcome.

I'm in Canada, but yes you're right.  I'll have to adjust the temps in May or June

At least now I know the ratio for each loop.

summer is comming! Front Temp wil go over 30c, maybe 40c and all temps will be yellow

After some gruelling trial-and-error tuning, I've got my SP20 running at ~1480Ghz without any yellow asic temps or frequencies.  The goal is to keep temps at or below 115 C while attaining as high a hashrate as possible.  On the 'Last Day' chart you can see how many incremental changes I made to find this sweet spot.  We'll see how things look after a week  

Notes:
-Ambient temp is approx 15 C
-ASICs 1 & 3 always run cool  (In this config, ASIC 1 is almost at the max recommended settings yet still only 85 C - potentially overclockable?)
-ASICs 2 & 4 always run hot  (lowered the voltages accordingly)
-At the wall watts is at 1028 (~80% of my PSU rating).  This is close to the PSU recommended limit


TIP!!: If you want to access your spondoolies admin page from outside your LAN, you can create a virtual server on your home router to redirect traffic.  Just be sure to use a strong password!!



http://i.imgur.com/ms7BY56.png
http://i.imgur.com/taZwxei.png
http://i.imgur.com/uxXfuET.png

That should work fine.

Ok reduced to 165 each loop, after 40 min steady at 1212 Ghs with ASIC reporting 647 watts total.  Temps are good but all 4 loops are reporting slight under power but not affecting restart so far.

Personally I think you are still too close to the 750W limit of your power supply with a total maximum of 720W across the 4 loops. That puts you at 96% of the power supply rating. I would step down another 15W per loop and give yourself some more margin on the power supply. Yes, you'll probably lose some hashrate in doing so, and get a bit more peace of mind. Just my $.02.

Bronze CX750M has no more than 85% efficiency, so at the wall you are using ~800W. ASIC stats calculation assume higher efficiency-so whatever number you see there might be inaccurate. I prefer to run my SP20 slightly lower-at 160w limit with 1229-1231 typical hashing.

Others might attest on whether it is safe to run 750W PSU at 800W at the wall. I prefer to add more cushion as power usage fluctuates.

So I'm a little new at this, have my first SP20 running almost a week now (took most of that week to read all of this thread  ).  I had one of my New CX750M's die (sending it back for exchange), so in the mean time trying to set things in the safe zone with only 1 PSU with 750.

I need some input from the pros to know if I'm in the zone.

Current Settings:

Fan Speed: Auto

Start Volts Unit 1 .665
Start Volts Unit 2 .65
Start Volts Unit 3 .645
Start Volts Unit 4 .628

Voltage Limit  .67

Max PSU Power Unit 1 180
Max PSU Power Unit 2 180
Max PSU Power Unit 3 180
Max PSU Power Unit 4 180

I'm getting consistent 1250 - 1275 Ghs and my temps are good, one yellow @780hx on loop 0.  Just need to know if I'm missing something or Good to go?

2.6.14
Uptime:58563 | FPGA ver:100 | BIST in 79
-----BOARD-0-----
PSU[UNKNOWN]: 0->(169w/169w)[169 169 169] (->169w[169 169 169]) (lim=180) 0c 308GH cooling:0/0x0
-----BOARD-1-----
PSU[UNKNOWN]: 0->(176w/176w)[176 175 175] (->176w[176 175 175]) (lim=180) 0c 314GH cooling:0/0x0
-----BOARD-2-----
PSU[UNKNOWN]: 0->(177w/177w)[177 177 177] (->177w[177 177 177]) (lim=180) 0c 322GH cooling:0/0x0
-----BOARD-3-----
PSU[UNKNOWN]: 0->(175w/175w)[175 175 175] (->175w[175 175 175]) (lim=180) 0c 322GH cooling:0/0x0
LOOP[0] ON TO:0 (w:8335)
 0: DC2DC/1/:[vlt1:667 vlt2:669(DCl:794 Tl:794 Ul:669) 66W  98A  55c] ASIC:[ 85c (125c) 780hz(BL: 780) 4023 (E:193) F:0 L:0]
 1: DC2DC/1/:[vlt1:664 vlt2:669(DCl:794 Tl:794 Ul:669) 69W 104A  71c] ASIC:[ 85c (125c) 820hz(BL: 820) 4312 (E:193) F:0 L:0]
LOOP[1] ON TO:0 (w:8409)
 2: DC2DC/1/:[vlt1:662 vlt2:664(DCl:794 Tl:794 Ul:669) 70W 105A  79c] ASIC:[105c (125c) 810hz(BL: 810) 4239 (E:193) F:0 L:0]
 3: DC2DC/1/:[vlt1:664 vlt2:666(DCl:794 Tl:794 Ul:669) 70W 106A  85c] ASIC:[115c (125c) 820hz(BL: 820) 4170 (E:193) F:0 L:0]
LOOP[2] ON TO:0 (w:8567)
 4: DC2DC/1/:[vlt1:662 vlt2:664(DCl:794 Tl:794 Ul:669) 71W 107A  54c] ASIC:[ 85c (125c) 840hz(BL: 840) 4359 (E:193) F:0 L:0]
 5: DC2DC/1/:[vlt1:658 vlt2:664(DCl:794 Tl:794 Ul:669) 71W 107A  70c] ASIC:[ 85c (125c) 830hz(BL: 830) 4208 (E:193) F:0 L:0]
LOOP[3] ON TO:0 (w:8581)
 6: DC2DC/1/:[vlt1:654 vlt2:656(DCl:794 Tl:794 Ul:669) 72W 109A  75c] ASIC:[110c (125c) 850hz(BL: 850) 4347 (E:193) F:0 L:0]
 7: DC2DC/1/:[vlt1:650 vlt2:656(DCl:794 Tl:794 Ul:669) 68W 105A  80c] ASIC:[105c (125c) 820hz(BL: 820) 4234 (E:193) F:0 L:0]

set it to work as best as it is possible and whip it hard

good luck!

I bought it used   

if you bought it less than 3 months ago, i would contact spondoolies

got one myself that 1 loop does not want to start from time to time
another one has problem with only one asic, keeping that loop under 220w seems to be ok.  high temps makes the miner restart itself very badly! because of that asic

Yes, this only happens occasionally when I restart the miner.  Restarting again clears the issue.  However as I mentioned above, this loop (especially asic 7) runs much hotter than the others.

is that loop starting at all? try changing 6pin cables and see if it is the same
if it is the same try disabling individually asic 6 and after that asic 7

I should also mention that; occasionally when I restart the SP20 I get the following error:
http://i.imgur.com/LGH4rsg.png

Is this normal?

I'm running 2.6.14 as well.

I'm wondering if lowering the the Max PSU Power for Unit 4 will reduce the temp on that last asic.  For example my settings would looks something like this:

Max PSU Power Unit 1 (70W - 288W)  :  240
Max PSU Power Unit 2 (70W - 288W)  :  240
Max PSU Power Unit 3 (70W - 288W)  :  240
Max PSU Power Unit 4 (70W - 288W)  :  100


Will this help?  Is it safe to do?

looks like hash dropped 120gh and  watts went down 162 watts.

 fan dropped from 90 to 60  

what is your firmware?  


I have 2.6.14   .

If you are on 2.6.14 firmware


  I would still dial it back  to my settings.  I still think you have not found your units sweet spot.
But you are much closer to the right setting.  and I know your psu's are saying thank you.  Compare the numbers  bump up a bit down a bit.

Extreme pushing of the sp20e  can always be done if coins jump up to 400 or 500 usd.  if they say under 300 usd be relaxed

So I took your advice philipma1957 and lowered my voltages.  Currently running at:

0.68V (start)
0.69V (max)
Fan 60
240W max

1364.51Ghs


One of the asics is STILL sitting at 120C.  I'm thinking there may be some dust in the unit limiting air flow.  Hopefully it's not a defective chip. 

Looks like this weekend I'll be opening for a cleaning.


http://i.imgur.com/arjtncA.png

Got one with a Scythe Ultra Kaze, the other has a Gentle Typhoon 4250rpm

Its located in the same orientation and location as the original fan, both running at 100% full speed off the psu

From here: