ANTMINER S3+ Discussion and Support Thread

[jk1966]

Just got my first ever mining machine S3+ about a month ago, so new to this whole thing.  Been reading through the threads, and haven't seen a solution to the problem, but understand the bad firmware issues. 1. Can't access any of the setup or configuration settings on any IP, could for a while when first started the machine but has gotten worse over the last few weeks to the point of no connection.

2. Can SSH in on 192.168.1.99 over LAN and that is all, the machine mines fine on a DHCP wireless connection and was checking the status every few days via the LAN and GUI.  Not 100% sure which firmware is on it, but the factory reset will not work.  Never upgraded the Firmware, but checked when new and it seemed to be the latest at the time.

What I would like to try is to SSH new and/or working Firmware to the unit.  I'm not familiar with how to do that.  mtd -r erase rootfs_data returns file not found as do others copy and edit commands I've seen in the posts.

Does anyone know the commands to SSH new firmware to the machine?  Thanks.

[dog1965]

works for my stable so far and "YES THE RESET BUTTON ON THE MINER WORKS FINE EVERYTHING GOES BACK TO FACTORY DEFAULTS". Now for the stability test which will take 4 hours or better to make sure there is no speed drops. so far so good.


Anyway if anybody wants a new s3 hash board its $100 dollars If you order it from BITMAIN so is the controller which is insane!!!!! They emailed me. I will just wait for the S5.

[jk1966]

Can someone post the contents of the etc/config/network file.  Think the problem is with that. I have no lan setting in there.  or send me link to it.  thanks

[nitrox1]

Disregard My last post sorry about that I forgot to unzip the file before I checked the MD5 and yes the MD5 Code does match sorry again.

Well no time like the present then. Once more unto the breach, dear friends.

I am holding my breath ..... don't be long!

It flashed 2 x S3 one early batch, one later with QC sticker, and they took. Did direct flash over LAN cable not WIFI.
Give me 15min to get GHs reading, I've set mine at 231.25, no change to voltage (voltage blank). This was running fine in October firmware update. I will get back ASAP

15min Update: Both sitting at 0.44 Th/s and 0.45 Th/s on 5s,  total average is a bit low, as didn't seem to spike at the beginning like they usually do.

Lets see how they do over the next couple of hours, as I was getting them dumping hash rate (50GH/s 100GH/s), and going bananas, after latest firmware.

GHash.IO 15min average 1.1 Th/s over 15min average, nice! From 2 x S3's


PS All help needed, don't supposed someone could help and potentially divert me from a bullet, and advice me on serial connection un-bricking an S3?
https://bitcointalk.org/index.php?topic=889206.msg9897190#msg9897190

I'm sorry to say, It looks like I'm getting the same instability, that came from the latest firmware. Hash rates are dropping, and fluctuating, I'm down to 800Gh/s, even on one round it was recorded at 500 GH/s, so either network issue as mentioned above, or the hashing part of the firmware isn't correct.
Can someone advise as to whether it is worth putting in something into voltage section like : "0.8"
I have a solid 1200W running two S3's, which used to run three S3's at 450GH/s each for months until this stupid firmware.

[dogie]

I'm sorry to say, It looks like I'm getting the same instability, that came from the latest firmware. Hash rates are dropping, and fluctuating, I'm down to 800Gh/s, even on one round it was recorded at 500 GH/s, so either network issue as mentioned above, or the hashing part of the firmware isn't correct.
Can someone advise as to whether it is worth putting in something into voltage section like : "0.8"
I have a solid 1200W running two S3's, which used to run three S3's at 450GH/s each for months until this stupid firmware.

Nothing to do with hashing changed on this version, so its likely something on your end. Else, the feedback looks pretty good and I'll open it up to a wider release.

[nitrox1]

I'm sorry to say, It looks like I'm getting the same instability, that came from the latest firmware. Hash rates are dropping, and fluctuating, I'm down to 800Gh/s, even on one round it was recorded at 500 GH/s, so either network issue as mentioned above, or the hashing part of the firmware isn't correct.
Can someone advise as to whether it is worth putting in something into voltage section like : "0.8"
I have a solid 1200W running two S3's, which used to run three S3's at 450GH/s each for months until this stupid firmware.

Nothing to do with hashing changed on this version, so its likely something on your end. Else, the feedback looks pretty good and I'll open it up to a wider release.

Probably is but why go from so stable to unstable. I'll be nosey ;-)

Any advise on the voltage is 0.8 good for for mild overclocking 231.75, or leave it blank

[dog1965]

I use 0795 which is really 0.795v or if you want to be technical 795mv millivolt

The guy that wants the CGMINER Commands log into your miner using putty then type - "cgminer --help" without using the quotas ""

[nitrox1]

I use 0795 which is really 0.795v or if you want to be technical 795mv millivolt

The guy that wants the CGMINER Commands log into your miner using putty then type - "cgminer --help" without using the quotas ""

Excuse my ignorance,  you type 0795 or  0.795 in to voltage box?

I also found this in overclocking forum

https://bitcointalk.org/index.php?topic=883197.0

[Duce]

Did anybody notice the BM1384 spec sheet. Still 28nm but reporting lower power consumption. I assume this is the chip for the S5? Or did I miss something?
https://www.bitmaintech.com/files/download/BM1384_Datasheet_v2.0.pdf

[pekatete]

Did anybody notice the BM1384 spec sheet. Still 28nm but reporting lower power consumption. I assume this is the chip for the S5? Or did I miss something?
https://www.bitmaintech.com/files/download/BM1384_Datasheet_v2.0.pdf

Eagle eye .... that is dated today! Lets have a good read then.

EDIT: Looks like a copy paste jobby ..... still has references to BM1382 ! So, which is which? Bitmain are becoming a bit lax in more than one way .... this market dominance seems to be getting to their heads; but yeah, this thing is a fair bit more efficient than the BM1382, hopefully they'll have at least one unit priced to fit our smaller wallets besides the much vaunted S5!



EDIT2: Also, the work input timings in the BM1384 datasheet have NOT been updated from those of the BM1382 .... crickey bitmain!?!

[charles2k]

Who's going to try first?

Flashed December 16 firmware (again not available on Bitmain web...) to six S3+ batch11 units and after 2 days it seems ok (no dropping to 30-50GHash as in November firmware)

[MyRig]

hummmmmm   

S3 and S4 uses BM1382 chips.  S5... I think you already caught it  BM1384 chips.  :

Some works could be copy and paste, but we will be paying more attention to details.  Each time you catch something, please let us know!

[Biodom]

hummmmmm   

S3 and S4 uses BM1382 chips.  S5... I think you already caught it  BM1384 chips.  :

Some works could be copy and paste, but we will be paying more attention to details.  Each time you catch something, please let us know!

Since you are freely talking on multiple threads about an unannounced product (S5), I assume that the announcement is inevitable (in the next 24-48hr).

[dogie]

hummmmmm   

S3 and S4 uses BM1382 chips.  S5... I think you already caught it  BM1384 chips.  :

Some works could be copy and paste, but we will be paying more attention to details.  Each time you catch something, please let us know!

Since you are freely talking on multiple threads about an unannounced product (S5), I assume that the announcement is inevitable (in the next 24-48hr).

[aarons6]

i dont know why everyone is so excited about new chips.
soon after s5s are released the difficulty will go up so much we will be making the same as we do now with our s3s.. plus the added cost of new machines which will probably never ROI.

[suchmoon]

i dont know why everyone is so excited about new chips.
soon after s5s are released the difficulty will go up so much we will be making the same as we do now with our s3s.. plus the added cost of new machines which will probably never ROI.

That depends on other things too, not just the S5 merely showing up. To declare ROI or lack thereof before the product is even released is a bit premature. S4 was declared by many (me included I think) to never ROI back in September and look where we are now with difficulty staying flat for two months.

[tonyec99]

Can I get some beta testers for a new S3 firmware please? Download link is here, it *should* be good to go but as you know....

Rolls back to the last good patch in Oct (1024), but fixes voltage adjustment and the important reset button. Hopefully.... let me know.

Link works for me...  will try as soon as I've plucked up enough courage!!

S320141219.bin installed and, so far, everything looks okay. Have we, at last, got working stable firmware?..  let's hope so.

[pekatete]

Some works could be copy and paste, but we will be paying more attention to details.  Each time you catch something, please let us know!

OK since you asked, here's another one.

The work input timing table in the new BM1384 datasheet is that of the BM1382 (copied straight from the older datasheet and clearly incompatible with the updated BM1384 power performance table!); Care to update it or remove the values completely as they are misleading?

[yslyung]

is the price gonna drop ? i think i better wait for S5 but NOT batch 1 ....

[Biodom]

is the price gonna drop ? i think i better wait for S5 but NOT batch 1 ....

batch 1-2 is ALWAYS the most profitable (if they work). My underperforming batch 1 machine made way more BTC than outperforming batch 7.