I spoke with Michael I think really late Friday night. He said he was going to look into an RMA, or maybe a board replacement,  as he tapped into my SP20, and couldn't get the stuck ASIC #4 to work at all.

My SP20 arrived early November and the order # was 2754. I don't have a ticket support number.

Oh, thanks for the T-shirt, by the way.

Well, got up early today to talk to support. It's 14:37 Jerusalem time, nobody there.

Now I gotta go to work.

I get off work too late to talk to them at 1900 hours...some I'm screwed for another week.

I'm still getting the following with the new firmware patch 2.5.26 on an SP20.

21/11 04:32:10:---- RUNNING SLOWSTART -----
Cgminer died and restarted
21/11 04:32:10:Started!
21/11 04:32:12:ASIC ERROR PLL on ASIC:4 (255)
21/11 04:32:12:restart_asics_full (stuck PLL B)
21/11 04:32:12:Recursive restart_asics (stuck PLL B)
21/11 04:32:12:FATAL: miner_gate.c:1504
21/11 04:32:12:Abort
Cgminer died and restarted
21/11 04:32:14:restart_asics_full (stuck ASIC)
21/11 04:32:14:Recursive restart_asics (stuck ASIC)
21/11 04:32:14:FATAL

I'll try getting up even earlier this morning so maybe I can catch somebody in tech support.

It's still mining on 7 out of 8 asics. No matter which firmware it is, asic #4 is still giving the same error (stuck PLL B).

I've tried tuning down the starting voltages so maybe it will be unstuck, but no luck so far.

Ok, I tried to add you on Skype....so far no response...is it zvi_shteingart with an underscore or zvi shteingart with a space?

Skype doesn't report anybody by the zvi_shteingart handle...

I'd like to get this problem resolved for an apparently damaged in transit SP20. Asic #4 is not hashing at all, no matter what I do, what voltage I set it at, whether I reboot the unit 4 times or more, etc. Running 2.5.12 latest firmware. I've already reseated the boards, and same symptoms.

Running on a known good power supply right now at 1300 GH/s with 7 out of 8 asics at 250w per module limit.

I have no idea what this is, but I'm posting it anyway.

14/11 16:35:44:---- RUNNING SLOWSTART -----
14/11 16:35:45:Started!
14/11 16:35:46:Asic disable 4: User disabled
15/11 00:27:32:Run time failed crutial 6
15/11 00:27:32:restart_asics_full (disable asic while running)
15/11 00:27:32:Testing LOOPs
15/11 00:27:32:Testing LOOPs done
15/11 00:27:32:Testing DC2DC
15/11 00:27:32:Asic disable 4: User disabled
15/11 00:27:32:Asic disable 4: User disabled
15/11 00:27:33:JOB 6000001 stuck, NOT killing ASIC

15/11 00:27:34:JOB 6000001 stuck, NOT killing ASIC

Watchdog caught problem: soft handle
spond-manager stop
Stopping MG
15/11 00:30:55:Signal
15/11 00:30:55:Pre read garbage 45ffff01 (0)/int 41114
15/11 00:30:55:Pre read garbage 8000006d (1)/int 41114
15/11 00:30:55:Data corruption 1 on read 6fe:32 (32837f00:20) got fffe:45 (45ffff01:45ffff01) - ql=0
15/11 00:30:55:Data corruption 1 on read fffe:41 (41ffff00:20) got 7fe:32 (3283ff00:14442d) - ql=2
15/11 00:30:55:Pre read garbage 41ffff00 (0)/int 41114
15/11 00:30:55:Data corruption 2 on read 7fe:32 (3283ff00:20) got 0:0 (2d:45ffff01) - ql=0
15/11 00:30:55:Data corruption 1 on read fe:32 (32807f00:20) got fffe:45 (45ffff01:8000006d) - ql=4
15/11 00:30:55:Data corruption 2 on read fffe:41 (41ffff00:20) got fffe:45 (45ffff01:8000006d) - ql=4
15/11 00:30:55:Data corruption 3 on read fffe:45 (45ffff00:20) got fe:32 (32807f00:14442d) - ql=0
15/11 00:30:55:Pre read garbage 41ffff00 (0)/int 41114
15/11 00:30:55:Pre read garbage 2d (1)/int 41114

Well, I still have an ASIC #4 down, and I guess Zvi is try to catch some sleep this Friday night.
 
But, at least it's mining. I reseated the boards, but it didn't really do anything.

I'll try to catch you on Skype next week, and I'll send Barbara another e-mail or something.

Tried rebooting, changing voltage values, turning ASICs on and off, but ASIC #4 just doesn't want to activate.

I will try to contact Zvi or jtoomin friday morning sometime.

After applying 2.5.12 firmware upgrade, I still have a stuck ASIC 4. I did look through the log, and I have seen a new error after applying the firmware.

Error::: PLL stuck:ffffffff asic(4/255) LOOP:127 (Initial freq), 3000
Nov 11 02:33:35 miner local1.warn minergate[1019]: Critical: ASIC ERROR PLL on address 4/255
Nov 11 02:33:35 miner local1.warn minergate[1019]: Discovered stuck PLL 7000001

If that makes that ASIC try to run at 3000 Mhz, than that would be the hangup. I highly doubt these things are rated for 3000 Mhz...

Hi. I got my miner November 7th. Got it all plugged in and the 2nd board would cause a halt in cgminer. Miner would not hash. Eventually, I disabled all the ASICS on the 2nd board, and the miner would hash with only the first module.

Noticed that the boards sat at an angle inside the case, so I opened it up and re-aligned. The boards are not screwed in or secured with anything except some grooves raised up from the enclosure. Pretty cheap design, which is prone to board shifting if handled rough in transit.

After re-aligning, I manually enabled and disabled each ASIC on board 2, and figured out that ASIC 4 will not work under any voltage, the other ASICs on the board work fine. Getting 1300 GH/s with 7 ASICS...so not too bad at max 225w per module setting. I'll turn up the juice once all the problems have been worked out, hopefully.

Got an e-mail tonight from Barbara saying they will refund 0.12 BTC because the SP20s shipped late. I'm cool with that, so I thanked her and also told her to forward my diagnosis to tech support. Also filed a form for tech support asking for directions as to how to proceed. Requested they send me a replacement board.

I'd rather not ship the whole thing back and lose hashing time. I can do major surgery on just about any electronic device without a problem.

I have an SP10 Dawson also that has been going without a problem for months now. I can't really say I'm impressed with the SP20, because I'm so far pretty underwhelmed.

Firmware is still 2.5.6, not really sure about any other versions. Hopefully they will get all the bugs worked out soon.

Right now, I have managed to get 7 out of 8 ASICS to hash. ASIC #4 is still not hashing, no matter what I do. If I enable it, the mining process will not start.

Received a pleasant e-mail from Spondoolies-tech about the refund of 0.12 BTC because of the late shipment. At least they are honest.

Sent them my address and also added a tech support note that ASIC#4 is not hashing at all, and caused the mining process to crash.

We will see what they say.

Starting voltage is 0.64, max volts is 0.75, max power setting is 225 per 2 chip module. Notice that Chip 5 is sucking more power and is running cooler because the other chip on the module is disabled. It's also running faster. Quite interesting.

LOOP[0] ON
 0: DC2DC/1/:[vlt1:699 vlt2:703(DCl:794 Tl:794 Ul:749) 86W 122A  57c] ASIC:[ 85c (125c) 955hz(BL: 955)   60 (E:193) F:0]
 1: DC2DC/1/:[vlt1:695 vlt2:703(DCl:794 Tl:794 Ul:749) 84W 120A  73c] ASIC:[ 85c (125c) 950hz(BL: 950)   55 (E:193) F:0]
LOOP[1] ON
 2: DC2DC/1/:[vlt1:687 vlt2:695(DCl:794 Tl:794 Ul:749) 88W 128A  82c] ASIC:[110c (125c) 990hz(BL: 990)   54 (E:192) F:0]
 3: DC2DC/1/:[vlt1:691 vlt2:695(DCl:794 Tl:794 Ul:749) 82W 119A  85c] ASIC:[105c (125c) 925hz(BL: 925)   55 (E:188) F:0]
LOOP[2] ON
4: disabled (User disabled)
 5: DC2DC/1/:[vlt1:716 vlt2:722(DCl:794 Tl:794 Ul:749) 97W 135A  66c] ASIC:[ 85c (125c) 1035hz(BL:1035)   45 (E:193) F:0]
LOOP[3] ON
 6: DC2DC/1/:[vlt1:693 vlt2:698(DCl:794 Tl:794 Ul:749) 85W 122A  73c] ASIC:[ 85c (125c) 955hz(BL: 955)   51 (E:193) F:0]
 7: DC2DC/1/:[vlt1:695 vlt2:698(DCl:794 Tl:794 Ul:749) 86W 124A  82c] ASIC:[110c (125c) 960hz(BL: 960)   43 (E:191) F:0]

Pulling 1300 GH/s with 7 chips. Not too shabby. The hash rate is reflected at Eligius, apparently.

Well, you're a luckier man than me. I just gone mine on the 7th also, and it seems like it will not work with the 2nd board (4 asics) enabled.

Did you set the power level for Max PSU power unit 1 through 4 higher than the default of 100 Watts each?

I set mine for 200 W each, using a 1000 W power supply with 4 PCI-E connectors for testing purposed.

So far, I'm showing 692 GH/s with only 4 ASICs enabled. Yes, the fan does cycle weirdly.

I will be checking this thread, hopefully we can figure something out. Be back after some more testing.

*******Edit***..

Right now I've got 3 pairs of ASICS going, the top 2 and the bottom pair. The 3rd pair down, every time I enable it, takes down the whole thing, cgminer won't start. Weird.  I will have to probably open this thing up and see if that board is set right (probably the same problem you had). Something in shipping maybe?

Anyway, SP20 is reporting 1061 GH/s, but I'm only seeing 400 GH/s in the pool. I replaced 2 AntMiner S1s (-400 GH/s) and supposedly added 1000 GH/s right now and my total output is not actually higher than what I was seeing before.

So, I'm seeing the same thing you're seeing. SP20 is reporting a higher hashrate than the pool is seeing. Weird. Going to try to mess with difficulty rating, but Eligius does this automatically, so I'm not really sure if I changed something on my end, if it would do anything.

So they can then own everything, cheap.

The same thing that happened before, and after the battle of Waterloo, when the British stock market was manipulated.

Except this time, they're using printed money, which costs them nothing, therefore they do not care about the losses.

It's fucking diabolically brilliant.

Just don't worry.

Keep mining.

One difficulty adjustment isn't gonna kill anybody. In case you haven't noticed, difficulty adjustments happen on a regular basis.

What's so different about 20% this time and these??

Nov 05 2013    510,929,738    30.70%    3,657,378 GH/s
Oct 26 2013    390,928,788    46.02%    2,798,377 GH/s
Oct 16 2013    267,731,249    41.45%    1,916,495 GH/s

Those were MUCH worse, yet people are still mining.

If one difficulty adjustment ruins your ROI, you're doing something wrong.

Just got my two S3's in today, and plugged them in and set them up.

They are both configured to run at 218.75 Mhz from the factory.

For whatever reason, one of them runs at 425 Gh/s and the other runs at 408 avg. The 5s averages are higher. Maybe it will be better when they are "broken in"  

I put them both on an 850W power supply I had, but they both turned off, and started hashing at 26 Gh/s. Then, I put each on it's own 800W power supply, and that seems better. Only using 2 PCIE connectors so far.

Might be the voltage on the power cord coming from the garage might have dropped due to the 100 ft length...

Running out of power in the basement....

I disagree with that assessment.

You sound just like a broker who says "Invest in these mutual funds from our company, they are better", and then gets paid commission for it.

Anybody that says to just buy BTC and hold it is just advising people to buy BTC so that more demand is created, so that they can sell the BTC they already have mined at a higher price.

No, actually I picked an example that has not fully played out. That was just the "6 months" checkup. Final tally for ROI is actually at 9 months.

I'm not really going to get into specifics, because frankly, I don't want to seem like any kind of boasting asshole. Suffice it to say that I owe a lot of stuff that I learned to people on these forums, who have tried to help other people, like myself and others. Those people are very nice and helpful, and have helped me a lot, so I try to pass on a little of what I have learned.

Cryptomining is like a box of chocolates, you never really know what you're going to get. Sure, everybody can have hindsight, and say "You should have bought BTC at price XXX, instead of buying mining hardware", but that's like saying "You should have bought Apple at $11/share back in 1995-6". That kind of advice is absolutely worthless. Kind of like the advice you get from a broker, who just wants to make his commission.

My advice, forget calculators (They have their uses to see mainly when you should unplug stuff). Physics is tough to argue with.

Buy with a 6-9 months window, and do your own thing. Try a few things out. I'm not going to get into specific amounts, because the IRS could be trolling these boards, but here is a few "projects" of mine that have turned quite well.

1) May 2013, bought a 7850 and a 7750 Radeon. Everybody was saying GPU mining was dead, blah blah blah. Kept mining with GPU's until August, Now those GPU's are mining Scrypt. ROI made beyond my wildest dreams, let's just leave it at that.

2) July and August, added 5 block erupters. ROI made, because I kept the BTC and waited for the top.

3) Bunch of K16's that I ordered finally came in early November 2013, mined with them for a month and a half,  sold them in December  on E-bay for basically what I paid for them. Net profit enough to pay my wife's student loans and mortgage one month. Wife stays off my back now on cryptomining and computers (priceless).

4) Some jally preorders came in about the same time as I got a 30Gh/s little single off E-bay (October 2013). That thing has made the money back on it in spades, and then some. Butterfly labs stuff is still going, although I will have to pull the plug soon, according to my calculations.

5) Made enough to get the Antminers, which are still going.

6) Just got an SP10 Dawson in, we'll see how that works out. The next power bill might be interesting however. That thing sucks power.

So, my point is, stop stressing about ROI for the next 6 months, and don't listen to the naysayers who say you should buy BTC and hold it. Yes, you have to hold BTC, but hold the BTC you mine yourself, until there is the right time to either re-invest in mining equipment, or sell them.

Hope this helps somebody.

Forget USD. It's a dying currency, and soon it will be completely worthless, only good enough for kindling and toilet paper.

I used to think like you, honestly. I always converted stuff to USD, and calculated profit/loss that way.

However, once I've used BTC for about a year, I've started thinking in BTC. There is no need to think of stuff in other terms.

If I paid in BTC, I calculate my profit in BTC.

I could care less about fiat, worthless currencies. Yes, my power bill is in $$, but that's chump change in my book. $300 over 6 months is going to the movie theater twice a month with another person and buying popcorn and a drink. I don't even think about it.

Yes, the price of BTC was high in December/January 2013/14, but that's irrelevant also.

The price of USD to Yen, Euros and Pesos goes up and down too, but you don't see people calculating and saying "OMG, I had more Euros in December than I did in May, if I converted my USD to Euros!"...

What's the point of conversions to other currencies, if you don't really care about the other currencies?

I use BTC as much as I can, other currencies be damned.

Yes, your analysis makes sens XScrypt, as long as you pick and choose which data points for USD/BTC conversion you use. But, I respectfully disagree with your conclusion, since it is based on hindsight.

I don't really care if  BTC goes up or down in terms of fiat currencies, because I'm minimizing their use. Hell, you can buy just about anything you want nowadays in BTC, even cars and groceries.

Plus, if all small time miners gave up and bought BTC, then BTC would be controlled by HUGE centralized mining farms, which is just like the Federal Reserve/Central Bank system. The whole point of BTC is to have a decentralized currency creation system, that cannot be brought down if one or more nodes go down.

The whole theory of "You should have bought BTC instead of mining because of ...blah blah blah" goes completely AGAINST what a decentralized cryptocurrency with no central control stands for.

Looks like stats are down, they were up yesterday.

Miners check out fine however, so it must be just the stats server again.

I think I have a spare quad core lying around somewhere, must be better than the Commodore 64 running the stats...

What kind of machine would you need to run the stats to keep up with the pool? Seems like it coughs up a furball every two weeks !

You are right, I did miscalculate my power costs.

700 W for 2 antminers (measured using KillOwatt meter),  24 hours a day, that's 16800 W or 16.8 KW.

At $ 0.10 / KW, that's $1.68/day * 180 days = $302.8 for half a year.

Subtract that from the profit, and that knocks it down to roughly 11% profit.

Still, profit is profit. You're not getting 11% by investing in treasury bills, and AntMiners are way safer than the stock market (Nothing exciting happens in my basement, heh).

Those Antminers are still running, and will be profitable until roughly October/2014.

Thank you for pointing that out.

I love these threads, because they are usually total nonsense, with just some random new website calculator to put in some assumptions into, leading to some set in stone conclusion which never changes.

You won't make ROI, ever, blah blah blah blah.

Please. Your assumptions are totally invalid.

Here, I will give you a real world example. I'm a small-time home miner, mostly do this for a little fun and a little profit.

I bought an S1 AntMiner in December 2013, and one in January of 2014. Total cost for the 2 S1 AntMiners was 5.35 BTC.

Here were are in June of 2014, and I'm looking at a wallet balance of about 6.6 BTC.

Net profit = 1.25 BTC, or (1.25/5.35 invested)*100 = 23.36-ish %.

Electricity cost per day is about 35 cents, to that's about 65 bucks worth of electricity for 180ish days.

This is not a boast, or a calculator, or a set of assumptions.

The price of BTC in $$, or rubles, or yuan doesn't matter, since I paid with BTC.

If you want something done right, do it yourself, and don't worry about the price spikes or price dips. Averaged out over 6 to 9 months, you will almost always make a small profit (assuming you don't have an excessive electricity costs).

S1 Miners are cheap now, and there is a reason for that.

If you go to thegenesisblock.com, click on mining, and enter the S1 stats, you will see that if you buy one now, chances are, by September/October time frame they will not pay for themselves in terms of power cost.

Meaning, that the chips and the miners, while still good now, won't be good for much longer, since they are 55nm technology. Now, you can undervolt them, like people have done here, to get a pretty good watt/Gh ratio, but that's just prolonging the inevitable.

If you're in this long term, you need to get 40nm and under technology.

That being said, getting an S1 now cheap, messing with it, learning about it, would be a good way to get some knowledge.

I've spent more than the wife thought possible.

I also sent her shopping with some of the proceeds, so now she's a believer.

Overall, made a small profit. Nothing earth shattering. Kind of nice to have a hobby that pays you back.