Bitcoin Forum

I'm starting this as a resource for ASICMiner Tube owners. There are currently 2-4 threads spread around the forums with various info that will be useful to have all in one place. Please post additional info here and I will try to keep the OP updated.

Note, this is entirely unofficial. I would love for someone to "officially" weigh in on these topics (Friedcat? PhaseBird?).

Anywhere you see a "???" I'd love to have some input from other users.


Getting started

Note: see the useful links section for some setup pictures and videos.

Default IP of controller = 192.168.0.254

You can change the IP from the Settings page (see below), but you'll need to configure your computer/tablet/phone to be on the 192.168.0.xxx subnet initially (or configure your router's subnet mask to include the 192.168.0.xxx range).

On first boot, you should open the controller page in a broswer at http://192.168.0.254:8000 and check the "Test Status" page to make sure the right number of boards is showing up. If it is correct, go to http://192.168.0.254:8000/FlashMega to flash the boards to the latest firmware.


Pool Compatibility

Works out of the box with:

- ghash.io
- Slush's pool (mining.bitcoin.cz)
- Bitcoin Affiliate Network (mining.bitcoinaffiliatenetwork.com)
- mmpool.org
- BTC Guild (using stratum-lb-usa48.btcguild.com:3333 to connect)

Other pools may need to use a proxy.


Using BFGMiner as a proxy (Windows, Linux, Raspberry Pi, OpenWRT)

run BFGMiner with these options: -o <real pool url> -u <real pool username> -p <real pool password> --stratum-port 3333 --set-device PXY:diff=850

While BFGMiner is running, point the tube to the IP address of the machine BFGMiners is running on. Username and password don't matter.

Full instructions here: https://bitcointalk.org/index.php?topic=735728.msg8310763#msg8310763

Note: sometimes there is a problem with bfgminer where it reports an error with bufferevent and the controller will lock up and/or reset. When this happens, it seems the only solution is to power down and power back up your whole rig.

Note2: sometimes bfgminer starts kicking back tons of duplicaates causing a high reject rate. I cannot find a pattern to this.


Using ckpool as a proxy

Download, compile, and run ckpool from https://bitbucket.org/ckolivas/ckpool and run in proxy mode.

Using slush gekko proxy

Try out the code here: http://www.gekkoscience.com/misc/slush_gekko.tar

(documentation included in archive)


Interface

Test Status

For each healthy board you will see this:


Where "Board xx" is the unique board number and "V:x.xx" designates the firmware version of the board. If you have fewer status lines on this page than connected boards (4 boards per tube unit) then you have a bad board. If any of your boards doesnt show up as the above, but instead the "O" is replaced by a "-" or "." then you have bad chips on a board.

Statistics

For each running board you will see a line like this:


Prf = hashrate of board
ExP = expected hashrate of board
Uti = Shares per minute
Efi = Efficiency = Prf/ExP
HWe = Hardware error rate (Currently there seems to be a bug where HWe is periodically, incorrectly reported at 100%)
PwrDn = number of chips on this board that are dead (equal to number of "." characters for that board on the Test Status page)

Below the lines explained above, there is a row of buttons. In the middle of the buttons is the current clock speed of the miners. It looks like this:


ReSession = restart current mining session
ForceLP = Force Long Poll - Force the board to accept a new job
Clock Up = increase clock speed of mining boards by 10 MHz
Clock Down = decrease clock speed of mining boards by 10 MHz
Wake Up = (I think this just refreshes the page, but I'm not sure)
ReClock = Adjust time difference between local and pool

You will also see lines below that give a sumary of performance for the entire unit, which is however many boards you have chained together. It looks like this:


Accepted = total number of accepted shares
Rejected = total number of rejected shares
Hardware errors = (you guessed it) total % hardware errors
PwrDn = total number of chips that are dead
Real performance = total hashrate
Expected performance = total expected hashrate
Utility = total utility (not sure how calculated)
Miner = total efficiency
Network diff = total network difficulty
Worker diff = worker difficulty
Real_LP = number of real Long Poll events ???
Running time = uptime
Session time = total time in current session (I think time since starting hashing at current pool and connection... can anyone confirm?)
RS = number of resyncs on the data cable (errors on the data cable)
RC = number of reconnects to the pool
Sha/m = (I think shares per minute???)
Subm/m = (I think submissions per minute???)
Jobs/m = (I think jobs per minute???)
Nonce/m = (nonce per minute???)
Cmd/s = ??? (related to communication between controller and boards)
Ans/s = ??? (related to communication between controller and boards)


Settings

Settings is broken into 3 sections: LAN (network settings of the miner), Stratum (pool settings), and ASIC (settings for the device itself).

LAN Section

IP = IP address of miner
Mask = subnet mask of network miner is attached to
Gateway = gateway of network miner is attached to
WEB Port = port for accessing web interface
System port = ???
Primary/Secondary DNS = DNS servers for miner to use

Stratum Section

Pool URL = URL of mining pool to connect to (format like pool.server.com -- do not include "stratum+tcp://" or the port)
Pool port = port to connect to on the mining pool
User name = mining pool user name / worker name
Password = password for the mining pool user / worker

ASIC Section

Clock[Mhz] = starting clock speed of the boards (this can be adjusted dynamically from the Statistics page
RollNtime[ s] = How many seconds the board should change on nTime ( bigger means less communications to controller, default value should be OK)
WakeUp period[ s] = ???
Normal/Smart[0/1] = ??? (all I can tell is that the firmware version on Test Status page shows difference, but no apparent impact to hash rate


Useful Links

- Setup guide (not english): http://www.iasicminer.com/index.php/action-viewnews-itemid-51
- HW assembly and SW setup guide: http://youtu.be/zEL5EKWwBx0


To-Do/Welcome Contributions

These are some things I think it would be useful to have included here but I haven't had time to put together myself. If you want to write them, I will include them. Or if they are present somewhere I can easily link to, please post a link.

- Prf and ExP of boards at different clock speeds along with power consumption at those speeds (so we can identify the most efficient clock speed to run at)
- More info on pools that work "out of the box"
- Information on anything above marked with ???, especially the ASIC settings (what do these things do?!)
- More complete guides for bfgminer and ckpool proxy
- Actual screenshots of all of the above
- Assembly instructions

I found this site useful during assembly of my tube:
http://www.iasicminer.com/index.php/action-viewnews-itemid-51

BTC Guild does work using stratum-lb-usa48.btcguild.com:3333 for the server.

I'll be running the official (yes official) ASICMiner Tube thread, it will be up tomorrow.

Miner Tested, Friedcat Approved?

Friedcat design and build.

Miner tested doesn't mean jack if it's not also miner approved.

Unless I screwed something up, running the mining_proxy.py in www.gekkoscience.com/misc/slush_gekko.tar should fix stratum packets to work on other slush-based pools. I've been running it for ozcoin connection for three hosting customers since Saturday or so.

We're still working on what it'd take to make them properly compatible with brickpools.

We also are gonna look into the board-level protocol (as these are based on an open-source design, there might be good enough reference code) and see if we can't fork cgminer to talk to them in a way that doesn't suck, using a Pi or something else decent embedded linux. Just throwing that out there. Might be nothing happens, but we're looking into it.

Friedcat commissioned, yes.

I think we need a whole lot more than a thread for these...hopefully some quality control checks and a decent controller.  Dogie will you be in charge of taking requests and getting answers?

Well I got these fans today.
http://www.amazon.com/gp/product/B00A460TK6/ref=oh_aui_detailpage_o00_s00?ie=UTF8&psc=1
I will test them tonight. These fans may cut noise down a lot.
I would really like the controllers to allow  2 pool auto switching. Much Like the s-1's or s-3's have the 3 pool auto switching feature.

Maybe we can merge the threads? Or you can just contribute here? No need to spread the info around even more. I just wanted to put it all in one place so owners don't have to go searching multiple threads.

tldr, yes.


You might as well lock this one as it will become obsolete. Mine will have ALL information in it, there will be no searching around. Pictures are ready now, took ages to process them because there are so many and they're all silver on white so have to be processed manually, will get everything else done tomorrow.

Well, I look forward to it.

I will hold off on my review then it was going to be published in a few hours.

Take a look here for youtube guide on how to assemble including some quirks with the setup:

http://youtu.be/zEL5EKWwBx0

Suggested/recommended psu? 10 psu is needed for a moq set. It's going to be a wire jungle ! Can the dps2000w support 2 miners 24x7 ? Looks like a little to close to psu max output. For stock settings i think 1 only 1 pcie needs to be connected and 2 if oc? Not too sure myself so owners please confirm.

I use EVGA 1300 Watt Supernova PSUs for them.
http://www.newegg.com/Product/Product.aspx?Item=N82E16817438011 (http://www.newegg.com/Product/Product.aspx?Item=N82E16817438011)

I am using xfx550 80+ bronze as pair per miner.

http://i62.tinypic.com/2s7xgnq.jpg

https://bitcointalk.org/index.php?topic=750363.0 more pictures
No issues :).

Bought them cheap, like 40$ each.

Utility = Shares per minute ( Utility*4.29/60 = Prf  ;D )
PwrDn = Power Down ( Chip with "." status )

ForceLP(Long Poll) = Force the board to accept a new job ( the board is doing nTime rolling by itself )
ReClock = Adjust time difference between local and pool
These two buttons are better removed... no one actually need them.


Real_LP = Guess should be the actual forced LP, but the number seems too small here...
RS = ReSync ( indicates errors on data cable)
RC = ReConnect ( reconnect to the pool )
Cmd/s and Ans/s are both looks like stats for the communications between boards and controller

RollNtime[ s] = How many seconds the board should change on nTime ( bigger means less communications to controller, default value should be OK)
WakeUp period[ s] = ???
Normal/Smart[0/1] = Firmware version would change, but I can't see any difference in hash rate

Great insightful video thank you for taking the time to record it.

I believe with these you can run 5 tubes off 4 PSU.

I've got three different installs of 5 Tubes on 2x DPS2K that have been running for a week solid. One had issues I thought might be heat-related, but it was a faulty PSU. Replacement has been going steady since Sunday.

Comprehensive guide is comprehensive. Need more red bull.

http://www.exceptionalcanine.com/expertqa/dog-drinking-more/images/large.jpg

Nice!
Did you get your Red Bull?

 Official setup guide and support thread now up in super HD! (http://goo.gl/iPFKXq)

 https://farm4.staticflickr.com/3836/14946108089_b16317d74a_b.jpg (http://goo.gl/iPFKXq)

How did you get on with the fans? I've got one tube running and another on the way and the noise is giving me a headache with the standard fan.

I need to find a quieter one especially before the other one arrives to relieve my headache and stop me getting divorced ;).

Hi all,

I bought 3 used asicminer tubes with BE-Controller for each. The run fine 24h in a pool like ghas.io but when I try to run a proxy based on BFGMiner , no difference if windows or linux, the miner or the controller suddelny stops workig. Sometimes the setup recovers on it's own, sometimes I have to unpower the controller and sometimes I have to unpower controller and the tubes.

If this happens only the first LED is flashing on the controller.

Does anyone have an Idea what I should change in the bfgminer config ?

Regards
Stefan

I'm using bfgminer 4.7.2 with no problems using proxy.  My start up is "bfgminer.exe -o stratum+tcp://uk1.ghash.io:3333 -u x -p x --stratum-port 3333 --set-device PXY:diff=512"

I also am able to mine on f2pool,btcguild, no eligius yet (keeps disconnecting with same symptoms you are having). 

The controller is fundamentally broken.  They do not work with most pools.  Several pools actually changed to work with the faulty tube controllers.

The bfgminer fix is not a good one, between bfgminer and the tube controllers you've got enough bugs to feed an army of anteaters.  I'd stick to pools that work with the tube.  If you really want to mine on an unsupported pool, you might be able to use slush's proxy, which is much more stable.  There's a version I made that fixes the extranonce2 length issue, although it still does not work with pools which give the 'h not equal zero' error.  linux only.  http://www.gekkoscience.com/misc/mining_proxy.py (http://www.gekkoscience.com/misc/mining_proxy.py)

--
novak

Ah ok - I discovered the Problem using eligius too. So it's pool dependend ?!

Anyone here mine successful in eligius using a Asicminer Tube ?

Anyone tried mining in eligius using Slushs Proxy or CKPool ?

Regards
Stefan

It is pool dependent.  The BE controller has a bad stratum implementation, which does not work with all pools.  I think Slush's proxy (the version I posted, dunno about default) and CKPool contain the extranonce2 length fix but if memory serves the error given by eligius is 'h not equal zero.' which requires bfgminer.  I couldn't get bfgminer to run as a proxy for more than an hour without crashing.

--
novak

Yep, ckpool in proxy mode also fixes the extranonce2 issue. ckolivas worked that one up for us I think the first day we had Tubes to test with, but he's far too busy to try and fix all the issues with AM's stratum implementation.

Hopefully novak can get some of those USB adapters worked up and distributed soon. These things seem to behave much better on a real cgminer.

Hi,

I have an ASICMiner Tube that I want to use with CGMiner on Windows. I have a USB adapter that is from ASICMiner (like those sent with the Prisma). It is connected and recognized by Windows. I also used the Zadig utility to change the Driver to WinUSB. When I run cgminer, I get an error stating "USB init, open device failed, err -3, you don't have privilege to access - AMU device 3:1" Any insight into what I am doing wrong? Thank you for your help

Are you running cgminer as admin? Try right clicking it and "run as administrator" Or disable UAC since that is a pain anyways.

THIS EXACTLY! It's been frustrating the heck outta me the past several months. Older controller so I assume it's software is V1, as I've seen some on ebay that specify V2.

I'm on slushes pool so did some looking there today and noticed the minimum difficulty for my worker was the default so set it up higher according to th table on the page. Will see if that helps.

I tried a different cat5 cable just to rule that out. When it happens, like above, the first LED is flashing on the controller and it is not accessible on the network though the network plug light is on.

Any thoughts on what I should try next?

Thanks,
Thomas

BUMP

Does anyone have any insight into why the block erupter ethernet controller loses connection several times a day? Is there an update to this controller? Is there another option?

Thanks,
Thomas
 

Did you ever try to run ck's modified ckproxy, or novak's modified slush proxy, which fixed some of the problems with ASICMiner's broken stratum implementation? I never tried BFG for proxy on a tube. I also never had an issue with the original controller connected to the above-mentioned proxies, or directly to slush's port set up for these. Then there's also the option to run cgminer and use a USB adapter, which should work on every pool.

For reference, I ran 31 of these in my hosting when they first came out; some of them stuck around for about 8 months and I didn't have the problems you're experiencing.

My tube is running off BFG mining proxy with no issues.

Howdy,

I picked up a handful of these tubes (4), testing the boards one at a time for issues..  Been running great, one board seemed to get the controller to lock up frequently though (no web interface, not responding to pings), as others said above, it sometimes will clear on it's own or needs a power cycle.

My question is..  What are the data cables called?  Or what are the ends called?  Where can I order some..  haha

For example the power plugs on the boards are Molex 45558-0002 or Molex 45558-0003...

I am needing basically a replacement set of the UART signal cables, but I am not successful when searching for them...

One more question..  I saw in a manual for the Prismas, they connected the controller to the fan port on one of the boards..  Isn't this 12 volt?  The controller says 5v on it..  Are there different versions of the controller?  Or is 12volts ok for them?  If so, where can I buy one of these cables..  haha

Still trying to decide if I should find PSUs for them or just sell them when I get the data cables sorted out...

I've got a bunch of cables from them laying around. Also, the controller has an internal regulator that should be good up to about 15VDC input so it's not too worried there, it'll work off whatever you're feeding it.