Topic: **US** BitFury Setup Guide - page 26. (Read 53199 times)

How were you able to tell which chips were bad?

Where's the reference to how to fix the problem?

Thanks.

I fixed an EOL board by shorting the solder jumpers. no need to cut the traces. Funny, my EOL board with disabled chip is hashing at 25GH/s

Guys heads up on pencil mod for overvolting and thus aiding overclocking.
Use pencil mod on R2F, make it around 1.1 - 1.2K, voltage goes from 0.66V to 0.81V. The default value is ~1.3K. It may blow the 30AFuse if you go too high; so make sure you do it only if you have a soldering station handy

After Mod (I have only 3 H cards )

If you have your rPi set to a good ip address for your network then you should be able to plug it into your network and then access it at whatever ip address you set it to.

ie: instead of http://www.google.com in the address bar of your browser you just use the ip address you set the rPi to.

You'll get a webpage that you fill out with your credentials.

Thanks Lubah!! got the video out working and able to fix Gateway IP, but now have no idea how to reconfigure pool credential via SSH? (I have a macbook pro can i just plug an ethernet cable from my laptop to the router, and use the 192.168.1.249 with a browser??)


thanks

Could you link to cscape's suggestion? I looked back thru this thread and didn't see it.

Also, what did you use to see that there were some chips not hashing? Thanks

Okay problem solved. Inferring that the problem chips be disabled as per cscape's suggestion
I cut the SPI traces of 4th and 5th chips using a knife and shorted the jumpers using thin copper wire. Here are the results. The board now uses 14chips for hashing.

Anyone know if Punin's EU setup guide is compatible with the US versions?

http://punin.com/README_BFSB_v0.1.txt

I got my kit today.  Thanks Dave!!!

When I first started it the nonce rate as reported by admin page was 25.5GH/s

After few hours, the rate dropped to 23.5-24.5.  Not sure why, but I'll let it run overnight to see if it stabilizes.

Pool rates with btcguild/emc are 22-23 GH/s

Not bad for a little fury...

I have an idea.. and since I am a n00b in this more or less you can tell me if this could be possible.

What if I were to maximize my profits by playing my luck a bit more
Example is today's Slush's pool.

19939   2013-09-07 01:31:05   0:01:37   1131170   980   0.02131520   256525   25.02841001    93 confirmations left
19938   2013-09-07 01:29:28   0:00:21   222852   100   0.01110472   256524   25.06503077    92 confirmations left
19937   2013-09-07 01:29:07   1:41:16   73576381   53260   0.01876635   256523   25.34140029    91 confirmations left
19936   2013-09-06 23:47:51   0:55:34   40356574   29200   0.01753752   256517   25.07990007    85 confirmations left
19935   2013-09-06 22:52:17   0:31:34   22671961   17580   0.01843250   256508   25.07482132    76 confirmations left
19934   2013-09-06 22:20:43   1:01:44   45313110   31880   0.01795067   256505   25.25120010    73 confirmations left
19933   2013-09-06 21:18:59   9:46:10   443154889   308960   0.01831448   256499   25.36216592    67 confirmations left
19932   2013-09-06 11:32:49   2:13:12   97267673   71160   0.01900746   256421   25.10027617    confirmed
19931   2013-09-06 09:19:37   5:47:52   256339688   185720   0.01628119   256402   25.02380004    confirmed

For instance.. I go mine in the pool A, if block is not found in 2 hours (or whatever predetermined time)
I would roll over to start mining in pool B. etc..
This way the 9hr 46min (@ 0.0183) wouldn't let me sit there for that long.. and I would still have gotten some BTCs, then went to a different pool that "could" have brought me better luck.
Like the 21 seconds to get 0.01110 or heck even the top line.


Thoughts on something like that could be worked in? Is it even feasible?

It appears to load balance/split equally between any of the three pools you have enabled.  If you only have one pool enabled, everything goes there.  If you have all entries pointed to the same pool and enabled, everything goes there, too...

My understanding was that all three pools are used concurrently (as three separate stratum proxies).  I would fill them all out to the same pool if you want to devote 100% of your hashrate.

If you are good with linux, you can edit your sdcard on a different machine with an sdcard reader (at least I did).  I was thinking of taking a dd image of it, but I wanted to get mining quickly so I have skipped it.  Has a raw image been posted anywhere as a backup?

How does the pool page work?  There are entries for up to 3 pools.  During mining, does it use only the 1st entry, and the 2nd/3rd act as backups in the event the 1st fails?  Or are the 3 pools evenly divided across all running boards?

 have a board with only first 4 chips hashing. I think the chips start from bottom left corner and go vertically up (U40, U41, U42, U43) down (U44, U45, U46, U47) ,..., then down the second column and then up, etc. Any tips on how to make the board fully functional. It's okay if chip 5 will not hash.

You don't need hdmi.  As long as you know or can figure out the IP address you can just SSH to the raspberry pi with a tool like putty should you require command line access, else just web browse to the units IP address.   If its still at the default address, just put your pc on the same subnet and access the default address via SSH

I also have a no HDMI vid out pi as well. Composite video worked just fine so I will live with it. You can ssh into it if its on your network. If you dont have composite video on computer monitor, try your TV.

Regards

hi everybody!! got my starter kit today, but ran in to problem during set up so power down, and now the pi have no video out (hdmi) red power light are on, can anybody help please!! never use pi before ((



thanks

Thanks Dave, just filled out the form and will return it first thing in the morning. Will you be sending the address to ship it back to in the email?

edit: just realized i have box you shipped it in and will ship it back to that address. (had a noob moment).

We've seen the no power for rPi problem.  We believe its a bad M-board.  Out of over a hundred M's we've only seen this once.  Problems with M-boards were not bad, so we decided not to burn in test each one.  That policy may change if more are going out with problems.

Please use the return feature on the site and we will expedite a replacement to you.

Sorry for the disappointment.

My quote as a consultant - if there is no time to do it right the first time, how will there be even more time to do it over again, later?