[CLOSED] ASICMiner Prisma 1.4th/s - 1.47 BTC

[bigredbutton]

^^ Happen to know where one can find a tutorial on how to set-up a BE Controller to run with a Prisma?

[1715301325]

1715301325

[1715301325]

1715301325

[1715301325]

1715301325

[1715301325]

1715301325

[Xtra7973]

^^ Happen to know where one can find a tutorial on how to set-up a BE Controller to run with a Prisma?

From page one, CrazyGuys post:
BE controller information

1) The prisma and Tube BE controller are the same, but tube controllers are setup by default at a frequency of 270. If you receive a tube controller, drop the frequency down to 240 before connecting a prisma.
2) BE controller's default address is 192.168.0.254. The easiest way to access it the first time is by connecting to your PC or laptop via crossover cable or put it on a network with the same subnet.
3) BE controllers require a power source of their own. They can reliably handle anywhere from 5v-12v. I have tested with both a 5v 2a and 12v 1a power adapter. The power adapter will need to have a 2.5mm barrel connector.
4)  The BE controller supports ghash.io and slush by default. BTC GUILD and ckpool are also supported, but they use different addresses for tubes and prismas. See dogie's tube guide for more info on supported pools.

do you need help with bfgminer?

[bigredbutton]

I tried that. I connected the ethernet directly from the controller to my computer or router and nothing from the default IP address. I tried the :8000 port as well.

I'm not having an luck with my BE controller or Rasp Pi.

when I tried to run the command


sudo /var/www/minera/./build_miner.sh cgminer-AM



it said doesn't recognize cgminer-AM.

[CrazyGuy]

Also I asked in this thread earlier if anyone got the prismas running minera to work with mobileminerapp on a cell phone as I had no luck running it. Well once again user error on my part. In the settings in minera where you enter your API key from mobileminer and email address you signed up with thats where I made the mistake. The API key I copy and pasted, so my error was my email address. https://i.imgur.com/GqdpDin.png instead of gmail.com I entered gnail.com and I just kept overlooking the obvious. Now mobile miner works on my cell and I can now check stats, change pools. start, stop and restart the miners from anywhere. Here is a screenshot of it from my cell phone https://i.imgur.com/7GkB19D.jpg If anyone has any problems setting it up just look at the obvious first, as I overlooked it. It's very simple to setup and get going. If anyone needs help with it just ask, I will try to do what I can. Also everyone with the BET0 problems, I just refreshed my setup last night and I had the BET0 and then a BET1 problem. All I did was just hit restart miner under miner in minera, now it did take about 10 trys, but it works, so you dont have to keep changing back and forth from Cgminer (rockminer fork) and Cgminer (asicminer fork) just keep spamming the restart button. It works and seems to be a faster way then doing all those other steps.

Strange but might be true about the restart button spam , sometimes i get up to 300gh/s , sometimes 500gh/s , but now it got up to incredible 1200 gh/s .. I guess i won't touch it for now and leave it so i can enjoy the hairdryer sound for a while.
Thanks for the advice man.

Im glad its working for you. I was just so frustrated trying it the other way with all those steps, and to get the same result each time. So I tried it that way, and about the tenth time it worked, hashing away now for a little over 22 hours without problems. I removed any arguments in the Startup extra commands (rc.local) box so im assuming its hashing at 230 and both units are around 1450 GH each with an error rate of about 5% total.

Was nice to see big numbers for a bit , but it did not last very long , i dont know if it was because the debug was on or any other reason , but it made the cgminer crash after like 30 minutes. Now the Minera is stuck , can't restart cgminer and it wont start even after restarting Rpi. It will need new image on the SD i guess.
Back in the old story. Woohoo.

Tried spamming restarting the miner and not luck.  Keep getting these errors:

[2014-11-16 21:49:12] Pushing work from pool 0 to hash queue
 [2014-11-16 21:49:12] Got work from get queue to get work for thread 0
 [2014-11-16 21:49:12] BET0: Sent work 23 to board 0
 [2014-11-16 21:49:12] USB: BET0 read1 buffering 54 extra bytes
 [2014-11-16 21:49:12] BET0: Nonce 000000ff from board 16, asic 12 for work 22
 [2014-11-16 21:49:12] BET0: Nonce Fix Failed
 [2014-11-16 21:49:13] USB scan devices: checking for BET devices
 [2014-11-16 21:49:13] BET looking for and found BET 10c4:ea60
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 0424:ec00 instead
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 0424:9512 instead
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 1d6b:0002 instead
 [2014-11-16 21:49:13] USB scan devices: checking for ICA devices


When I tried CrazyGuy's build from yesterday, I  can't even see the miner now in Minera.

Maybe 1 or 2 days more of trying to get it to hash then to the market to load this piece of crap.

There is obviously something wrong with the miners, does'nt seem like many people are getting these. I guess people have to be lucky , to get properly working one 

If you left debug and log file on, you may have filled up your SD card.

[yxt]

I tried that. I connected the ethernet directly from the controller to my computer and nothing from the default IP address. I tried the :8000 port as well.

I'm not having an luck with my BE controller or Rasp Pi.

Have you set your PC IP fit the BE Controller subnet?

[ManeBjorn]

I normally have it on BTC Guild and I have tried BTC PoolParty, and ckpool and no luck.  It had worked on all three prior to it's stoppage today.
Very strange.  I have tried 3 different controllers and 2 different versions of the RPi image as well as 3 different RPis.
I have also torn it down and put it back together and tried one board at a time.  It just does not want to work.  As an extra test I have tested the controllers and images on a Tube and a T1 as well with no issues. Tried several PSUs as well.
Just so dam weird.  No changes then all the sudden it stops and nothing is bringing it back.

Noted.
I will try it.
I am at a loss as to why it just stopped to begin with so weird after running so well for so long.

I just tried this now.  The weird thing is the unit ran until this afternoon then just detects one board or BET0 and does not hash.
Same error.  Just nothing.  I have tried a new PSU and cables.  Always the same thing.
Any ideas anyone?

Run this command to pull the latest and rebuild cgminer-AM

sudo /var/www/minera/./build_miner.sh cgminer-AM


ASICMiner checked in some updates to their fork about 5 hours ago surrounding Block erupter initialization. Worth a shot.

I actually had the same thing once I upgraded. I had to restart cgminer a few times to get it up and running. Might be a better idea to just go back to the last build.

many moons ago I had gridseed  build worked for 2-3 months then stopped on wemineltc. still worked on cex.io   a suggestion try 2 or 3 or 4 other pools

[CrazyGuy]

I tried that. I connected the ethernet directly from the controller to my computer and nothing from the default IP address. I tried the :8000 port as well.

I'm not having an luck with my BE controller or Rasp Pi.

when I tried to run the command


sudo /var/www/minera/./build_miner.sh cgminer-AM



it said doesn't not recognize cgminer-AM.

Subnet needs to match on the be controller. Also, you would need to use a crossover cable in that scenario.

For minera, you are using the default minera image without my changes. You need to run the manual installation steps and pull from my repository... or, just rebuild the main branch of cgminer with the default image and give it a shot.

[bigredbutton]

The problem is when I run those commands it doesn't seem to update. After I download the latest update what do I need to type in minera to pull from your repository? I think this is where I am going wrong.

[bigredbutton]

sudo apt-get install -y lighttpd php5-cgi
sudo lighty-enable-mod fastcgi
sudo lighty-enable-mod fastcgi-php
sudo service lighttpd force-reload

When your web server is ready you can install Minera:

sudo apt-get install -y redis-server git screen php5-cli php5-curl
cd /var/www
sudo git clone https://github.com/crazybitcoinguy/minera
cd minera
sudo ./install_minera.sh

Didn't seem to work. I have no idea what I am doing when it comes to command lines and compiling and repositories. I need it explained as simple and as possible for you see I am part retard...

[bigredbutton]

I tried that. I connected the ethernet directly from the controller to my computer and nothing from the default IP address. I tried the :8000 port as well.

I'm not having an luck with my BE controller or Rasp Pi.

Have you set your PC IP fit the BE Controller subnet?

Sorry for flooding this thread but it is more active and responsive then the Minera thread. Getting desperate due to the lost mining time.

My knowledge of networks and their set-up is limited and that could be a big part of my frustration. IP Fit, like in ready to work for? subnet? I should probably phone this one in...

[ManeBjorn]

Ok here is some info that may help us all get to the bottom of the miners stopping running and not working right either on BE200 or RPis.

I just noticed on the BE200 that all the board disappear from the test status screen and the stats screen and then come back up after 30 to 60 secs.
I had not noticed this before.  I was refreshing and they all disappeared.  Then they come back.

Now I am seeing it happen again this is all within about 1/2 hr.
All the cables have been checked.  The PSU has been tested and replaced as well to eliminate all factors.

So I wonder what causes this.
Very strange.

[aluam]

Tried spamming restarting the miner and not luck.  Keep getting these errors:

[2014-11-16 21:49:12] Pushing work from pool 0 to hash queue
 [2014-11-16 21:49:12] Got work from get queue to get work for thread 0
 [2014-11-16 21:49:12] BET0: Sent work 23 to board 0
 [2014-11-16 21:49:12] USB: BET0 read1 buffering 54 extra bytes
 [2014-11-16 21:49:12] BET0: Nonce 000000ff from board 16, asic 12 for work 22
 [2014-11-16 21:49:12] BET0: Nonce Fix Failed
 [2014-11-16 21:49:13] USB scan devices: checking for BET devices
 [2014-11-16 21:49:13] BET looking for and found BET 10c4:ea60
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 0424:ec00 instead
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 0424:9512 instead
 [2014-11-16 21:49:13] BET looking for BET 10c4:ea60 but found 1d6b:0002 instead
 [2014-11-16 21:49:13] USB scan devices: checking for ICA devices


When I tried CrazyGuy's build from yesterday, I  can't even see the miner now in Minera.

Maybe 1 or 2 days more of trying to get it to hash then to the market to load this piece of crap.

There is obviously something wrong with the miners, does'nt seem like many people are getting these. I guess people have to be lucky , to get properly working one 

If you left debug and log file on, you may have filled up your SD card.

I checked the log file through SSH, it didn't have more than 1MB , deleted that and cgminer started right again , now i have to get it running by spamming restart again and hopefully it will last longer.
By the way , just came home and the other 2 miners still nowhere to be found :/ i thought they would come today already , but no luck.

Would you happen to have a BE controller to use instead of the pi? If so I would try using the BE controller and bfgminer 4.10 as a proxy. I had that setup before I got the USB to UART adapters and to be honest it was rock solid. Running the prismas off of minera and the USB to UART adapters seem to be a little bit more temperamental then just using the BE controller with bfgminer. I tried multiple versions of bfgminer and 4.10 was really solid. If you need any help or guidance setting it up I will do whatever I can with my limited knowledge to help you.

Nope I dont have BE controller. Unfortunately i didn't thought of that when i was buying prismas. I was naive enough that the miners would be easy enough to setup with the USB controllers...
And they are nowhere to be found in the UK , was thinking about getting one and try it. But no idea where to look for it.
I'm quite familiar with setting up the proxy from previous miners , but thank you for the help anyway .

[aluam]

For anyone having the BET0 error problems , for me the best working cgminer is 4.7.0 , compiled on Rpi from git : https://github.com/blockerupter/cgminer . Don't know why , but seems to be working better than the AM one for some reason. Also for the spamming button , better is to open up the minera terminal under the "System" , cd /var/www/minera/minera-bin/src/cgminer or AM or whatever your minera uses. And just start it up with the arguments or json config which is at /var/www/minera/conf/miner_conf.json , then its easier to spam it and see what is actually going on without having to turn on the log all the time , works better for me that way.
This way i managed to start it up after few tries again, hashing 1400gh/s this time , 1170W power usage. 6% errors so far. Not sure if its good idea with my 1200W platinum corsair, its close enough.
If i keep --bet-clk 23 i was hashing 1200ghs with no errors and 1070W , not sure whats better.

Sorry for keeping the spam in the topic , i thought this could be helpfull.

[Chris_Sabian]

For anyone having the BET0 error problems , for me the best working cgminer is 4.7.0 , compiled on Rpi from git : https://github.com/blockerupter/cgminer . Don't know why , but seems to be working better than the AM one for some reason. Also for the spamming button , better is to open up the minera terminal under the "System" , cd /var/www/minera/minera-bin/src/cgminer or AM or whatever your minera uses. And just start it up with the arguments or json config which is at /var/www/minera/conf/miner_conf.json , then its easier to spam it and see what is actually going on without having to turn on the log all the time , works better for me that way.
This way i managed to start it up after few tries again, hashing 1400gh/s this time , 1170W power usage. 6% errors so far. Not sure if its good idea with my 1200W platinum corsair, its close enough.
If i keep --bet-clk 23 i was hashing 1200ghs with no errors and 1070W , not sure whats better.

Sorry for keeping the spam in the topic , i thought this could be helpfull.

If you are running the prisma on 1 PSU then this a very bad idea.  PSU really don't like running more than about 80% rated power continuously.  You have a very high risk of damaging/destroying your PSU.   

[aluam]

For anyone having the BET0 error problems , for me the best working cgminer is 4.7.0 , compiled on Rpi from git : https://github.com/blockerupter/cgminer . Don't know why , but seems to be working better than the AM one for some reason. Also for the spamming button , better is to open up the minera terminal under the "System" , cd /var/www/minera/minera-bin/src/cgminer or AM or whatever your minera uses. And just start it up with the arguments or json config which is at /var/www/minera/conf/miner_conf.json , then its easier to spam it and see what is actually going on without having to turn on the log all the time , works better for me that way.
This way i managed to start it up after few tries again, hashing 1400gh/s this time , 1170W power usage. 6% errors so far. Not sure if its good idea with my 1200W platinum corsair, its close enough.
If i keep --bet-clk 23 i was hashing 1200ghs with no errors and 1070W , not sure whats better.

Sorry for keeping the spam in the topic , i thought this could be helpfull.

If you are running the prisma on 1 PSU then this a very bad idea.  PSU really don't like running more than about 80% rated power continuously.  You have a very high risk of damaging/destroying your PSU.   

I thought so , i should probably lower the clocks down a bit.

[philipma1957]

I have the best single psu you can get to runs this (money no object)  the evga 1600 watt platinum psu.

At setting 22=  freq 230 (slight underclock)  this gear does great.  I get 88.727gh x 16 = 1420 gh  with 2.63% errors poolside at mmpool.org  it is about 1400gh watts

watts bounce from 1100 to 1130 on a kill-a-watt   

so best number is  1100/1420 = .7746 watts at the k-watt-meter 
worst number is 1130/1400 =    .8071 watts at the k-watt-meter   


 the psu can do 1600 watts and it cost me 300 usd .
with a 10year warranty I plan to use it in some type of mining for years to come.

This prisma really needs quality power supply and good cooling. 

I can only  manage  the heat correctly  with an open door setup. 

Thoughts are this you need to be a little more advanced to use this then an s-3.
It is loud.
For now I will not run this if I am not at home  as my mining area can't quite be made fire resistant.

If you can set this in an outside building  with proper cooling it is pretty nice.
You avoid the sound problems.
You avoid the burn up risk .

If you are an apartment person this is a no go.



@ crazyguy 

I sent a pm today with a few questions . Please let me know if you got it.
 

Thanks phil

[mrpark]

I sold my Prisma 1.4Th/s for 1.5 BTC. It's been a fun experiment while it lasted!

[philipma1957]

I sold my Prisma 1.4Th/s for 1.5 BTC. It's been a fun experiment while it lasted!

It is good gear if you can host it in the right place. Selling it was not a bad idea on your part.

[PlanetCrypto]

I have the best single psu you can get to runs this (money no object)  the evga 1600 watt platinum psu.


Thoughts are this you need to be a little more advanced to use this then an s-3.
It is loud.
For now I will not run this if I am not at home  as my mining area can't quite be made fire resistant.

If you can set this in an outside building  with proper cooling it is pretty nice.
You avoid the sound problems.
You avoid the burn up risk .

If you are an apartment person this is a no go.

In the PC power supply arena I absolutely concur. The EVGA's are Top Notch.
However, what we feel is a more more cost effective solution follows:



Connected to this adapter Gekko Science Sells ( http://gekkoscience.com/products/IBM2K_board.html )


Bottom side view:


What the look like wired up and running:


But this is what we call a "Desktop Computer".

8 6 core AMD's (48 cores), 64G memory, Dual nVidia Quadro's w/1.5G each, . . .

These are 1800W supplies (147.5A * 12.2V) They have adjustable OCP (Over Current Protection) and adjustable output voltage, in addition to all the "bells 'N' whistles" the Gekko Science adapter brings to the table.
We have paid anywhere between $18.75 -$25 each, the Gekko Science adapter sells for $60.
We make our own PCIe cables, although Gekko Science sells 3' cables for, if memory serves, $3.75 each.
The fans we scarfed in the salvage market which flow 24 CFM each and are required to run these supplies outside the IBM Blade Center enclosures they were designed to be in.
I believe Gekko Science offers fans for these supplies but don't know their cost.
So assuming one purchased the cables from Gekko (8 * 3.75) for $30, the total cost of this supply is ~$115

Just a thought.

[philipma1957]

I have the best single psu you can get to runs this (money no object)  the evga 1600 watt platinum psu.


Thoughts are this you need to be a little more advanced to use this then an s-3.
It is loud.
For now I will not run this if I am not at home  as my mining area can't quite be made fire resistant.

If you can set this in an outside building  with proper cooling it is pretty nice.
You avoid the sound problems.
You avoid the burn up risk .

If you are an apartment person this is a no go.

In the PC power supply arena I absolutely concur. The EVGA's are Top Notch.
However, what we feel is a more more cost effective solution follows:



Connected to this adapter Gekko Science Sells ( http://gekkoscience.com/products/IBM2K_board.html )


Bottom side view:


What the look like wired up and running:


But this is what we call a "Desktop Computer".

8 6 core AMD's (48 cores), 64G memory, Dual nVidia Quadro's w/1.5G each, . . .

These are 1800W supplies (147.5A * 12.2V) They have adjustable OCP (Over Current Protection) and adjustable output voltage, in addition to all the "bells 'N' whistles" the Gekko Science adapter brings to the table.
We have paid anywhere between $18.75 -$25 each, the Gekko Science adapter sells for $60.
We make our own PCIe cables, although Gekko Science sells 3' cables for, if memory serves, $3.75 each.
The fans we scarfed in the salvage market which flow 24 CFM each and are required to run these supplies outside the IBM Blade Center enclosures they were designed to be in.
I believe Gekko Science offers fans for these supplies but don't know their cost.
So assuming one purchased the cables from Gekko (8 * 3.75) for $30, the total cost of this supply is ~$115

Just a thought.

 

 I have a discount on psus most don't get them at the price I do.  
  My evga 1300 g2's were all 154 vs 180-190
 my evga 1600 p2 was well under 300 vs 330 for most buyers.  
I still build pc's for a few people.  So high quality psu's have other uses for me.

But anyone like you with a lot of gear running  should consider  the server gear.

  My gear location is an attached garage  with about 4k power in the winter and 2 k power in the summer.  We are in different leagues.  although a lot of gear has passed muster on the wooden bench  in my photos here.
 More then 200k since summer of 2012.  I liked these tubes long and short.  Just could not quite get them to work for my setup. So they were more of a move them in  move them out item.  

So far the best gear I have owned has been  the s-3's. I have 8 on hand right now with 4 others in other locations.