Topic: ANTMINER S7 is available at bitmaintech.com with 4.86TH/s, 0.25J/GH - page 313. (Read 528055 times)

I think there are two separate things being discussed.

1. There is an issue with the S7s where they will not hash higher than 4600 ish. It has been suggested that slightly low input power may cause this. I recommend checking your input load at the psu connection at the miner. If you are handy with a meter you may have meter leads with needle points and you may get in the back side of the connector, or, if you have a single going to a Y cable connecting to the miner, measure the unused end on the Y connector. Phillip shows pictures of this in a couple of threads and he has presented this theory regarding the voltage.

2. Phillip also brought up the fact that a soft restart may help this issue, and I believe it has for one or two people with the S7. I was commenting in this thread to the other gentleman that I have seen this issue with S3s where it is pronounced. IT is usually the worst if I have a quick power outage, just long enough for the miner to know it was off for a second. on some of the S3s it literally shows 3 or 4 ghs for the main average, not the 5s average, it usually shows a normal number for the 5s. I do a soft restart and it begins working correctly. I saw some of this with my S4s, but not the literal 4 or 4 GHS on hard boot, just poor performance until I performed a soft boot. The same with some S5s. I cannot say on the S7 as I am waiting.

3. Then of course the power on the hash boards before the controller. This is an instruction bitmain began passing with the S5+ and has now continued with the S7, but unfortunately we have been unable to get any clear indication from bitmain as to why so we understand their intent. In the latest statement I read, I cannot remember if it was their website or the S7 manual, but whichever states to the effect of if using multiple power supplies power on the hash board power supply first and then the power supply for the controller. The way it is worded it It didn't seem to be an issue if you happen to be powering hash boards and controller if it happens to be on one power supply, but if it is separate power on the controler last. I am sure their engineering team have a good reason for this, but because we have experienced issues with translation along with the fact bitmain has a poor record of communication we are left to theorize and try to interpret what they mean by these things.    

I know my input power is strong to the power supplies, and to the miners, and I have no issues with #2 because I began seeing that issue with my first S3 purchases and have been soft start everything after a hard start anyway. Hell I like to be up close with my miners after I boot them. I have shortcuts setup to get me there quickly so I can look at everything. I'm a bit anal like that so it works.

I think your main point here would be to check what Phillipma mentioned regarding power at the pci connector to the miner.

Feel free to PM if I can help in any way.

These s7's are in Russia

Haha. I Don't think burnt socket is a high rate of RMA seen in diagnostic, compared to all the other points of failures, but yes its not worth it even though there is a very little chance for proper connector to burn at 170watts per connectors.

Exactly or the board if they look at traces on it they likely can tell I'm guessing.   Sure some would sneak through.  But I have a feeling that a good amount that try this if they have a disaster within warranty will get caught.

They are very very pushy on it just use 3 per.  Better safe then risk a 1.5k machine over saving a few bucks (or even 50 bucks).

How is the hashrate?


Ufo

https://69.img.avito.st/640x480/1962599169.jpg


Found some miners

The hashboard thing they have started since S5+'s.   They make sure to warn you about it.   I was really scared at first thinking of if miner loses power, etc.

But my S5+ has handled a power outage and turned on all at once without issue.  It's just happened one time but was no damage.    So it's probley being to safe.  But I will still follow it as I feel there has to be a reason for it.

Ok here we go.
I have a batch one running 575 as shipped.
Has been 4600 for days (5 or so)

Power on corsair with controller and one hashboard.
Wait 5 seconds
Power on corsair with two hashboards.
After 15 minutes 4600

Soft reboot
After 15 minutes 4600

I don't understand the hashboard first thing.
Maybe 15 or 30 second delay between power on?

No no!
Power hash boards first, and then controller! <- This is what bitmain states
Personally, I then do a soft reboot on every bitmain miner after the hard boot completes.

Have ONE s7.
I'm running it on one bitmain 1600 watt psu.
I'm going to try using two corsair rm1000 psu's with one on two hashboards and the other on one hashboard and the controller.
I'm going to fire up the controller psu first and then the 2 hashboard psu and see what it does.
Then I will do a soft reboot and see what THAT does.

Just throw a reminder here to power the control board after the hash boards are up if possible.
The bitmain wording was recently changed to something more along the lines of apply power to the power supply feeding the hash boards first, and then the control board.

Many have suggested this has to do with the issues people are seeing with low hashrates.

Personally, I follow their instructions to the letter, and even to the point of using a different power supply to power my control board on my S5+ so there is no doubt about my hash boards being powered first. I am not wasting the leftover power on the power supply I use on the control board, as I am also powering a couple of S5s from it was well.

To comment on the low hash people are experiencing where the soft boot seems to pickup, I have this exact same experience with some S3s where it is truly exaggerated behaviour. Literally a hard boot shows a hash rate of a few ghs, where a soft restart allows them to operate correctly. I have seen some of the same but not as bad behaviour with my S4s, and once I knew what to look for I see it with the S5 as well. I have no doubt there is an issue across the board which causes this scenario, but as for it being tied to when the control board is powered on, I cannot say. Whether this has to do with the S7 as well I cannot say as mine is late.

Good luck to you!

Two burnt sockets, one not. 

Rich

At 400watt per boards, 2 plugs per boards would be logically safe, but with 3 per boards you further increase the safety margin. It would be good to keep in mind that, if you don't plug all 3 plugs per board, you void your warranty. I have no idea how they could possibly know that, but better safe than sorry.

TLDR: Yes 10 PCI-e 6 pin plugs.

Thanks so much, the S7 needs 10 pcie 6 pin plugs correct?

I've been asking for a full Refund since Sept 29th for this Pile.. Still no response to my tickets..

+1

I wouldn't mind joining you on your crusade, I got 2 weeks vacation left. Send me a PM if interested.

You are doing what?

This is the company so many people ran to defend when people were complaining about 3rd, 4th, and 5th S4 power supplies dying.
This is the company some people are still being so optimistic about.

This is the company you get on your knees and bow to them so you can play a political game and hope you get your miner in working condition, or many people do not even have a miner on order.

Again, I say if you cannot appreciate what it is to be treated correctly, you need to get out of your own abusive relationship. You deserve better!
... and that is exactly what we have here folks, an abusive relationship.
We are staying with someone who beats, yells, and morally abuses us, and we go back saying "please sir, may I have another?"

Well, I can tell you people and companies continue to treat you this way as long as you let them. I am speaking of myself as well.
This is a steaming pile of dung served to us on a cold plate and we are told to eat, you better smile and act like you enjoy it.

This is an obvious sign this company has customer service pulling down everything and they are not going up, but down. I *thought* they have tons of money, but now, no, they don't have any and are scraping to meet current orders. This is the only explanation. I've read enough of the hashfast and others, but this is the first time I have seen a successful company go the other direction.

To offer us the few dollars per day for a few days or even weeks is ridiculous. What if we are religiously opposed to mining on Bitmain's PPS pool? They are basing that amount on something they will be pulling in plus. The "test" machines in their mines will be getting sent to us. They are mining with them during the most profitable times.

Look at what suchmoon states. He is eloquently pointing out how we are true sheeple now. There is no denying it. I will join the others who have requested a full refund. I have no delusions I will receive it, but I agree with the gentleman who stated the original sales agreement was not met. Therefore this has nothing to do with warranty, and everything to do with honoring the original sale. 

I urge others to investigate other ways to support the bitcoin ecosystem. I began doing so recently as bitmain will see us all ripped of our dignity and continue stealing from the community.

Do you think for one second they didn't know about this problem when they took our money?
if this were a local business in your area would you be treated this way? Would you tolerate it? I hope not, because I would be there the second the door opened with an opportunity for them to take advantage of, and it would only involve them looking me in the eyes after they bend me over. They certainly would not be able to hide behind the internet.

I blame no one but myself for my stupidity. However after so many in this community came to the defense of bitmain after the S4 debacle, and the way they treated people I began to come around, and my curious side gets the better of me so I say sure and ordered an S5+. It has been running ok except for a few minor issues which would not be acceptable in any other business except this one because we have allowed our standards to become so low.

It is OK if they bend you over the table as long as they knock you out with drugs and give you a reach around when you wake up? If not, tell them, and tell anyone who will listen.

The ONLY precedence in something like this has been set by SCAMS, con artists and the lowest trash type of people in the world.

I understand bitmain doesn't care what we say, and that makes it even worse. I will see them at a conference, and like Josh Garza they better have bodyguards or be ready to sit and listen to what I have to say. Big words from someone behind a keyboard eh? Well, I have the money and time to devote to a personal crusade, so I may as well make it this one.

What stinks the most is knowing there are people out there who spent money they shouldn't have, you've seen it if you've been around any amount of time and lived through any type of scam. People are sucked in by the magic money boxes and know they will make X dollars, or *think* they will, and who suffers? Their fault sure, the person who spent the money I mean, but it doesn't change the fact that this company lied, they are cheating every person who ordered batch 1 and hasn't received what they were told they would.

I firmly believe bitmain is in the toilet financially, they do not and never will have the appropriate leadership to run a proper business, they have no clue in how to treat customers, and the people that make the decisions are too inexperienced to operate anything other than a local check cashing service. At least people know they are getting ripped off when they walk in the door.

Rush out another underperformer bitmain. Keep telling us how wonderful your latest chip technology is. Let me tell you how impressive this empty space on my rack is doing:
Not a fucking thing is how great it is.

Yoshi has made an effort with some people over the years and that is obvious, but he makes it clear he has no control over bitmain, he does what he can. He sent me a fan because bitmain refused to even acknowledge an email. This is after I spent 2,500 USD. The fan failed after a week. They offered to sell me one. At Least Yoshi was honest, which is what I asked for him to be, I said just tell me if I'm screwed, just tell me you aren't going to honor your own warranty. This is while we were on the phone. I was polite, professional, but like a good investigator I do ask the tough questions. I asked Yoshi I said personally Yoshi, not as a bitmain employee but your own personal opinion is this right? He was honest and sincere when he replied and said no, it isn't right, but it is what they do in these situations.

I should have known better, and I let myself make bad choices. I will no longer do so. I will not stop discussing this either. I do not care how many "people" begin to cower down again after they begin receiving their units and are happily discussing the latest mod or hashrate and I will do everything possible to warn every nub that even though there is not another manufacturer, do not go here because you will be what? Say it with me, you will be FUCKED.

EVGA 1600 GOld has 14 PCIe connectors, so you are covered, but it costs >$300 while mildly used EVGA 1300 Gold (has 8 connectors) would cost you ~$125 and it will run S7 with an addition of two of these:
http://www.amazon.com/gp/product/B004NNTVT6

I am running my S7 on EVGA 1300 as described above and it is stable so far (~two weeks)
Edit: I have an extra EVGA 1300 with two additional startech PCIE splitters-can sell to someone if needed (reason-did not order second S7 that i planned to use)

this has nine pcie cords some are 2x headed plug all into the 3 boards  leave the 2 headed pcie cords  ½ empty just use 1 head.

the empty heads are great to show you full load volt measurements

this is easy to fix  use 1 sata to pcie adapter   and plug it into the controller


http://www.amazon.com/Branded-8inch-15pin-Express-Power/dp/B005NJXY7O/ref=sr_1_2?ie=UTF8&qid=1444581473&sr=8-2&keywords=sata+to+pcie

Would this work for the S7?
How many VGA cords would I be short on, and would there be a way to fix that?
http://www.evga.com/Products/Product.aspx?pn=120-G2-1600-X1