seriouscoin
|
|
October 21, 2014, 11:35:37 PM |
|
Update:
I think i fixed the issue with dead chips, however i still need SP Tech to confirm the the power consumption is actually accurate.
Common sense told me that the reason for chips to become dead is due to low voltage. Some chips dont have the same tolerance as others.
I decided to raise the voltage up to 0.62, hoping this will give the bad chips more headroom. At the same time i lower the Maximum power consumption (to somewhat the same with what the stats told me when my voltage was set at 0.6)
After 30mins, i check the ASIC stats and no dead chips!. Power consumption state at what i set.
Can SP confirm the setting of power consumption indeed work? and the reading is accurate? I dont have wattage reader at the moment (since i use 240v plug )
|
|
|
|
klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
|
October 21, 2014, 11:43:46 PM |
|
Update:
I think i fixed the issue with dead chips, however i still need SP Tech to confirm the the power consumption is actually accurate.
Common sense told me that the reason for chips to become dead is due to low voltage. Some chips dont have the same tolerance as others.
I decided to raise the voltage up to 0.62, hoping this will give the bad chips more headroom. At the same time i lower the Maximum power consumption (to somewhat the same with what the stats told me when my voltage was set at 0.6)
After 30mins, i check the ASIC stats and no dead chips!. Power consumption state at what i set.
Can SP confirm the setting of power consumption indeed work? and the reading is accurate? I dont have wattage reader at the moment (since i use 240v plug )
accurate power consumption at the wall should employ a kill-a-watt. the SP readings are only as accurate as the PSU provides, whcih likely has a +/-5% range. If low temps are an issue, why not reduce fan speed to make the unit a bit hotter internally?
|
|
|
|
wh00per
|
|
October 22, 2014, 12:13:54 AM Last edit: October 22, 2014, 12:32:26 AM by wh00per |
|
accurate power consumption at the wall should employ a kill-a-watt. the SP readings are only as accurate as the PSU provides, whcih likely has a +/-5% range. If low temps are an issue, why not reduce fan speed to make the unit a bit hotter internally?
My Kill-A-Watt is only working for 125V, has a 5%-15% variance in accuracy -- 15% at the ends of the measured range and is good up to 15A. The unit has a quality check but is not calibrated in any way. IMHO, the PSU readings are as good as Kill-A-Watt readings. More expensive Wattmeters might give you what you need, but do NOT rely on a $20.00 piece of electronics for high accuracy measurements. Look for units which come with NIST calibration certificates, as they usually will check the whole instrument span and tell you the expected variation in readings (3 points, 5 points, 9 points certificates are usually available) under the form of a curve. Funny, my SP30 started to suffer from the 2.5TH drop recently .. where the bottom board completely shuts off >> http://powerprice.info/SPT/graphic.phpLOL Now it died..
|
CSA/cUL Certified Power Distribution Panels - Basic, Switched, Metered. 1-3 phases. Up to 600V. NMC:N4F9qvHz11BHcc4nh1LCJFsrZhA1EWgVwj
|
|
|
RoadStress
Legendary
Offline
Activity: 1904
Merit: 1007
|
|
October 22, 2014, 12:14:25 AM |
|
I decided to raise the voltage up to 0.62, hoping this will give the bad chips more headroom. At the same time i lower the Maximum power consumption (to somewhat the same with what the stats told me when my voltage was set at 0.6) Ok now I get what are you talking about. I thought you are talking about the Basic Settings. My mistake. You shouldn't lower your maximum power consumption because it will be lower since you are lowering your voltage. And yes the dead chips are because of the low voltage, but personally I just ignore them as long as I am happy with the hashrate/power ratio.
|
|
|
|
seriouscoin
|
|
October 22, 2014, 12:34:21 AM |
|
Update:
I think i fixed the issue with dead chips, however i still need SP Tech to confirm the the power consumption is actually accurate.
Common sense told me that the reason for chips to become dead is due to low voltage. Some chips dont have the same tolerance as others.
I decided to raise the voltage up to 0.62, hoping this will give the bad chips more headroom. At the same time i lower the Maximum power consumption (to somewhat the same with what the stats told me when my voltage was set at 0.6)
After 30mins, i check the ASIC stats and no dead chips!. Power consumption state at what i set.
Can SP confirm the setting of power consumption indeed work? and the reading is accurate? I dont have wattage reader at the moment (since i use 240v plug )
accurate power consumption at the wall should employ a kill-a-watt. the SP readings are only as accurate as the PSU provides, whcih likely has a +/-5% range. If low temps are an issue, why not reduce fan speed to make the unit a bit hotter internally? Which part of 240v plug dont you understand? Please dont ever reply to my post. Frankly i find all your posts to be nonsense. Especially when you even argue the Bitmain C1s might have different waterblock process.... all that BS.
|
|
|
|
seriouscoin
|
|
October 22, 2014, 12:39:32 AM Last edit: October 22, 2014, 12:54:52 AM by seriouscoin |
|
I decided to raise the voltage up to 0.62, hoping this will give the bad chips more headroom. At the same time i lower the Maximum power consumption (to somewhat the same with what the stats told me when my voltage was set at 0.6) Ok now I get what are you talking about. I thought you are talking about the Basic Settings. My mistake. You shouldn't lower your maximum power consumption because it will be lower since you are lowering your voltage. And yes the dead chips are because of the low voltage, but personally I just ignore them as long as I am happy with the hashrate/power ratio. To be honest, after wasting lots of time today i find voltage setting is the utterly most useless. This is why i asked SP-T and got the same answer (loweing voltage.....) Why didnt i just set power consumption right from the start if what you propose is true. I just applied the same setting to several SP10s and ....... got different results. Some still have dead chips. I think i will never find out the minimum voltage on these chips. Waste time and efficiency is still terrible. PS. even the intake temp is 15c, the chips temp under ASIC stats still read 77c. Same when intake temp is 24c.
|
|
|
|
RoadStress
Legendary
Offline
Activity: 1904
Merit: 1007
|
|
October 22, 2014, 01:03:17 AM |
|
To be honest, after wasting lots of time today i find voltage setting is the utterly most useless. This is why i asked SP-T and got the same answer (loweing voltage.....) You are lowering your max voltage too right? Because I don't see how it's useless. As I stated just a few posts ago I found out that optimal settings are min voltage 0.612 and max voltage 0.616 (or 0.618) and I get 1Th/s at 700W. You want better than that? Edit: Also the 77c reading means <=77 degrees, it's not actual 77.
|
|
|
|
seriouscoin
|
|
October 22, 2014, 02:13:17 AM |
|
To be honest, after wasting lots of time today i find voltage setting is the utterly most useless. This is why i asked SP-T and got the same answer (loweing voltage.....) You are lowering your max voltage too right? Because I don't see how it's useless. As I stated just a few posts ago I found out that optimal settings are min voltage 0.612 and max voltage 0.616 (or 0.618) and I get 1Th/s at 700W. You want better than that? Edit: Also the 77c reading means <=77 degrees, it's not actual 77. Uhhh how did you measure your power consumption? at 0.60-0.61 setting, i got ~760w from the stats page. I was told thats DC, so AC would be like 900w. Thats terrible isnt it?
|
|
|
|
RoadStress
Legendary
Offline
Activity: 1904
Merit: 1007
|
|
October 22, 2014, 02:19:11 AM |
|
Uhhh how did you measure your power consumption?
at 0.60-0.61 setting, i got ~760w from the stats page. I was told thats DC, so AC would be like 900w. Thats terrible isnt it?
I got a steady 700W in the stats page. What batch is your miner? 760W seems a bit too much for those settings. Are you sure you are having your max voltage set to 0.610?
|
|
|
|
seriouscoin
|
|
October 22, 2014, 03:45:56 AM |
|
Uhhh how did you measure your power consumption?
at 0.60-0.61 setting, i got ~760w from the stats page. I was told thats DC, so AC would be like 900w. Thats terrible isnt it?
I got a steady 700W in the stats page. What batch is your miner? 760W seems a bit too much for those settings. Are you sure you are having your max voltage set to 0.610? Yes i'm positive on the setting. Mine is first batch . I thought only SP30 has issues with first batch.
|
|
|
|
Zelek Uther
|
|
October 22, 2014, 06:50:37 AM |
|
Uhhh how did you measure your power consumption?
at 0.60-0.61 setting, i got ~760w from the stats page. I was told thats DC, so AC would be like 900w. Thats terrible isnt it?
I got a steady 700W in the stats page. What batch is your miner? 760W seems a bit too much for those settings. Are you sure you are having your max voltage set to 0.610? Yes i'm positive on the setting. Mine is first batch . I thought only SP30 has issues with first batch. May batch SP10s had slow-corner* ASICs. * I think that was the right term
|
Run a Bitcoin node, support the network.
|
|
|
RealMalatesta
Legendary
Offline
Activity: 2338
Merit: 1124
|
|
October 22, 2014, 07:56:56 PM |
|
|
|
|
|
stan258
|
|
October 22, 2014, 08:17:24 PM |
|
I don't think so unless you have a breakout board for it. Evga 1300 watt for off the shelf. Newegg had them for 168 free shipping with 20 rebate few Weeks ago. You'll want to look at the breakout boards threds. Most server psu only work well on 220v or better.
|
|
|
|
Koontas
|
|
October 22, 2014, 08:31:45 PM |
|
At this point I'd suggest waiting to see some updates from Spondoolies... Been pretty quiet lately and no SP20's have shipped yet with 9 days left in October.
|
Don't trust any exchange!
|
|
|
|
klondike_bar
Legendary
Offline
Activity: 2128
Merit: 1005
ASIC Wannabe
|
|
October 22, 2014, 09:22:22 PM |
|
To be honest, after wasting lots of time today i find voltage setting is the utterly most useless. This is why i asked SP-T and got the same answer (loweing voltage.....) You are lowering your max voltage too right? Because I don't see how it's useless. As I stated just a few posts ago I found out that optimal settings are min voltage 0.612 and max voltage 0.616 (or 0.618) and I get 1Th/s at 700W. You want better than that? Edit: Also the 77c reading means <=77 degrees, it's not actual 77. Uhhh how did you measure your power consumption? at 0.60-0.61 setting, i got ~760w from the stats page. I was told thats DC, so AC would be like 900w. Thats terrible isnt it? Im not sure the exact efficiency of the PSU, but 90% efficiency is probably a good guess if operating at 240V. for 760W DC thats about 845W at the wall. 85% efficiency would be 895W at the wall
|
|
|
|
zvisha
|
|
October 23, 2014, 04:55:26 AM |
|
Good morning to all. After 2 weeks of holidays and low availability, I will be available today on my Skype, zvi_shteingart, to handle clients support calls.
Don't send mails, just ping me on Skype - but have access to your machine ready (team-viewer or port forwarding ONLY, please) and ability to physically restart the miner.
Regards
|
|
|
|
promojo
|
|
October 23, 2014, 06:07:52 AM |
|
Hey guys... When can we expect October Batch 1 shipping? Now they are shipping at the end of October? Why is that? How long will shipping be to Canada once it leaves the door in Israel?
Thanks
|
|
|
|
Bicknellski
|
|
October 23, 2014, 08:41:06 AM |
|
... GO FREICOIN! This is going to be an interesting time for Bitcoin and even more interesting for FRC! Imagine a technology that makes it possible to trust anyone. Now imagine your bank, your mortgage and all your assets are all built with that technology. This is certainly not our present, but it could be our future. Blockstream is a group of people who share a vision of how to transform global systems of value exchange that, by design, put you first. 5.1.2 Economic experimentationBitcoin’s reward structure assigns new coins to miners. This effectively inflates the currency but it winds down over time according to a step-wise schedule. Using this inflation to subsidise mining has been a successful complement to transaction fees to secure the network. An alternate mechanism for achieving block rewards on the sidechain is demurrage, an idea pioneered for digital currency by Freicoin ( http://freico.in).
|
|
|
|
dplusf
|
|
October 23, 2014, 09:09:56 AM |
|
... GO FREICOIN! This is going to be an interesting time for Bitcoin and even more interesting for FRC! Imagine a technology that makes it possible to trust anyone. Now imagine your bank, your mortgage and all your assets are all built with that technology. This is certainly not our present, but it could be our future. Blockstream is a group of people who share a vision of how to transform global systems of value exchange that, by design, put you first. 5.1.2 Economic experimentationBitcoin’s reward structure assigns new coins to miners. This effectively inflates the currency but it winds down over time according to a step-wise schedule. Using this inflation to subsidise mining has been a successful complement to transaction fees to secure the network. An alternate mechanism for achieving block rewards on the sidechain is demurrage, an idea pioneered for digital currency by Freicoin ( http://freico.in). without reading the whitepaper because i am at work at the moment (but still bum around in bitcointalk :-)) What does Freicoin have to do with that and what do you mean by its even more interesting for FRC?
|
|
|
|
|