smit1237
Full Member
Offline
Activity: 182
Merit: 100
May the source be with you.
|
|
September 14, 2014, 04:32:40 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen.
|
|
|
|
Taugeran
|
|
September 14, 2014, 04:38:55 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere?
|
Bitfury HW & Habañero : 1.625Th/s tips/Donations: 1NoS89H3Mr6U5CmP4VwWzU2318JEMxHL1 Come join Coinbase
|
|
|
J4bberwock
|
|
September 14, 2014, 04:40:44 PM |
|
this is an excerpt from bfgminer's driver-antminer.c: i have an idea, if you are confortable following code manipulation. See below Below is the modified block: *info = (struct ICARUS_INFO){ .baud = ANTMINER_IO_SPEED, .Hs = ANTMINER_HASH_TIME, .timing_mode = MODE_DEFAULT, .read_size = 5, .work_division = 1, .fpga_count = 32, };
adding these two lines ( work_division and fpga_count) should tell bfgminer detection code that there are 32 FPGAs( in this case BM1380 ASIC) and the work division is 1, each chip gets its own work item. since in vanilla main branch code, they are not specified and so bfgminer falls back to an autodetect routine which due to the speed of the asics results in fpga_count of 8. add those, save and recompile bfgminer. Post them results once it fires up. Thanks, I'll try it.
|
|
|
|
smit1237
Full Member
Offline
Activity: 182
Merit: 100
May the source be with you.
|
|
September 14, 2014, 04:42:46 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere? Won't start. But rpi gets too hot to touch. I think i close to magic smoke Now i have 1.5Th/s
|
|
|
|
J4bberwock
|
|
September 14, 2014, 04:47:35 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere? Won't start. But rpi gets too hot to touch. I think i close to magic smoke Now i have 1.5Th/s There are quite a small list of components that can have gone bad on the blade, should be quite an easy fix.
|
|
|
|
smit1237
Full Member
Offline
Activity: 182
Merit: 100
May the source be with you.
|
|
September 14, 2014, 05:08:22 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere? Won't start. But rpi gets too hot to touch. I think i close to magic smoke Now i have 1.5Th/s There are quite a small list of components that can have gone bad on the blade, should be quite an easy fix. I'm experimenting with pi onboard uart. Looks like something goes wrong on uart side. But i don't know wich component goes crazy Want free blade?
|
|
|
|
J4bberwock
|
|
September 14, 2014, 05:10:24 PM |
|
cgminer will recognize it as LIX if I enable icarus when compiling
U]SB management [P]ool management S]ettings [D]isplay options [Q]uit 0: LIX 0 : 270MHz 179C | 0.000 / 0.000h/s WU:0.0/m
now back to bfgminer with Taugeran's idea
|
|
|
|
J4bberwock
|
|
September 14, 2014, 05:25:14 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere? Won't start. But rpi gets too hot to touch. I think i close to magic smoke Now i have 1.5Th/s There are quite a small list of components that can have gone bad on the blade, should be quite an easy fix. I'm experimenting with pi onboard uart. Looks like something goes wrong on uart side. But i don't know wich component goes crazy Want free blade? I'm always interested in cheap dead miners
|
|
|
|
Taugeran
|
|
September 14, 2014, 05:32:35 PM |
|
cgminer will recognize it as LIX if I enable icarus when compiling
U]SB management [P]ool management S]ettings [D]isplay options [Q]uit 0: LIX 0 : 270MHz 179C | 0.000 / 0.000h/s WU:0.0/m
now back to bfgminer with Taugeran's idea
will be looking forward to seeing this at my lunchtime.
|
Bitfury HW & Habañero : 1.625Th/s tips/Donations: 1NoS89H3Mr6U5CmP4VwWzU2318JEMxHL1 Come join Coinbase
|
|
|
J4bberwock
|
|
September 14, 2014, 05:40:50 PM |
|
no luck with the mod in bfgminer. "no device found"
|
|
|
|
smit1237
Full Member
Offline
Activity: 182
Merit: 100
May the source be with you.
|
|
September 14, 2014, 06:34:52 PM |
|
Nice. Now i have one dead blade and one dead raspberry pi Don't know how it could happen. define dead. wont start on or you see charring/magic smoke somewhere? Won't start. But rpi gets too hot to touch. I think i close to magic smoke Now i have 1.5Th/s There are quite a small list of components that can have gone bad on the blade, should be quite an easy fix. I'm experimenting with pi onboard uart. Looks like something goes wrong on uart side. But i don't know wich component goes crazy Want free blade? I'm always interested in cheap dead miners I can send blade to you. But postage on you. If you interested in that, pm me.
|
|
|
|
J4bberwock
|
|
September 14, 2014, 06:59:36 PM Last edit: September 14, 2014, 09:28:42 PM by J4bberwock |
|
no luck with the mod in bfgminer. "no device found"
I'm removing the changes I did in the 70-bfgminer.rules.in and recompiing, just i case it coud be the reason it's not working. EDIT: Still no luck. I also need to find where is stored the default speed to set it higher and check if all the chips are getting hot or not. Changing it with bfgminer doesn't seems to have any effect. EDIT2: looks like I had a misconnected cable. Even untouched github compile wasn't working until I unplugged and plugged back the usb/uart converter. Back to test the Taugeran mod. here is the result [2014-09-14 15:22:03] Started bfgminer 4.8.0-8-gbaf6fe5Segmentation fault (core dumped) EDIT3, removing the speed setting from the commandline made it start and give crazy numbers bfgminer version 4.8.0-8-gbaf6fe5 - Started: [2014-09-14 15:25:58] - [ 0 days 00:01:44] [M]anage devices [P]ool management S]ettings [D]isplay options [H]elp [Q]uit Pool 0: ...opoolmining.com Diff:32 +Strtm LU:[15:27:20] User:Jabberwock.SHA12 Block: ...33544345 #174828 Diff:2.64M (18.91T) Started: [15:26:25] ST:34 F:0 NB:2 AS:0 BW:[ 22/ 24 B/s] E:13.41 I:20.86mBTC/hr BS:138 1/32 | 51.48/58.36/ 3.93Gh/s | A:1 R:0+0(none) HW:0/none ------------------------------------------------------------------------------------------------------------------------------------------------------------- AMU 0: | 58.78/61.76/ 3.86Gh/s | A:1 R:0+0(none) HW:0/none
|
|
|
|
J4bberwock
|
|
September 14, 2014, 09:37:05 PM |
|
Looks like the only AMU over 10 minutes getting accepted shares is 0aa.
here is what I get when I [M]anage the devices
Select processor to manage using up/down arrow keys AMU 0bf: | 533.0/577.5/ 0.00Mh/s | A: 0 R:0+0(none) HW:0/none CP2102 USB to UART Bridge Controller from Silicon Labs Serial: 0001
they are all similar except the 0aa one
Select processor to manage using up/down arrow keys AMU 0aa: | 56.02/58.12/ 5.04Gh/s | A:40 R:0+0(none) HW:0/none CP2102 USB to UART Bridge Controller from Silicon Labs Serial: 0001
|
|
|
|
Taugeran
|
|
September 14, 2014, 09:40:39 PM |
|
Looks like the only AMU over 10 minutes getting accepted shares is 0aa.
But what is the hashrate after 10 minutes?
|
Bitfury HW & Habañero : 1.625Th/s tips/Donations: 1NoS89H3Mr6U5CmP4VwWzU2318JEMxHL1 Come join Coinbase
|
|
|
J4bberwock
|
|
September 14, 2014, 09:43:32 PM |
|
Looks like the only AMU over 10 minutes getting accepted shares is 0aa.
But what is the hashrate after 10 minutes? 5GH on the pool
|
|
|
|
J4bberwock
|
|
September 14, 2014, 09:45:55 PM |
|
Before adding the 32 FPGA thing, I was getting 8 sub-units all hashing @0.650 for the same total value.
The best hashrate I've got so far poolside was 30-35Gh with cgminer and tons of HW errors. I assume it was using the 32 chips, they were all hot Bfgminer don't throw a single error but it seems to only uses 8 chips. None of the chips are actually hot, they are all more or less warm.
|
|
|
|
Taugeran
|
|
September 14, 2014, 09:50:52 PM |
|
Before adding the 32 FPGA thing, I was getting 8 sub-units all hashing @0.650 for the same total value.
Hmm. Ls sigh. I thought it would work. Guess not. Well back to my drawing boar... Altium VM to do some board analysis
|
Bitfury HW & Habañero : 1.625Th/s tips/Donations: 1NoS89H3Mr6U5CmP4VwWzU2318JEMxHL1 Come join Coinbase
|
|
|
J4bberwock
|
|
September 14, 2014, 10:00:11 PM |
|
Time for some sleep here. both cgminer and bfgminer are really close to work, but it will have to wait. If Bitmaintech would have released the code...
Using the same pic and their firmware would be easier, but I don't like to depend on them.
|
|
|
|
Taugeran
|
|
September 14, 2014, 10:05:49 PM |
|
Time for some sleep here. both cgminer and bfgminer are really close to work, but it will have to wait. If Bitmaintech would have released the code...
Using the same pic and their firmware would be easier, but I don't like to depend on them.
I know, it's a crutch. But to get walking sometimes that's what you need til you feel comfortable
|
Bitfury HW & Habañero : 1.625Th/s tips/Donations: 1NoS89H3Mr6U5CmP4VwWzU2318JEMxHL1 Come join Coinbase
|
|
|
pekatete
|
|
September 15, 2014, 09:55:22 AM |
|
Looks like it won't compile at all. But i'll try. I've connected an old laptop that I used to compile some time ago, I'll try it too. I've looked at the driver code. It doesn't designed to run on windows. As i can understand. Hope i'm wrong. Still installing debian on vm. so slow Btw you must change vid and pid in usbutils.c near #ifdef USE_ANT_S1 to match your usb bridge. This is more or less correct. If you take a look at driver-bitmain.c and .h files you will see Meaning the code requires linux...for the moment. I may as well throw in, I'm working on 2 protos similar to what jabberwock has done, but both my USB dongle design and a small USB hub design use the PIC uC that the original controller uses. going this route has allowed me the ability of setting up shop for helping Luke-JR with S1/S3 support in a less brickable environment. See this to see what I mean. All I did was try and determine a couple of OS variables and dependency structures and I bricked a controller( temporarily) I am still trying to catch up on this thread, so bear with me if this has been overcome. I managed to build a windows cgminer of the latest 4.6 release by simply uncommenting out the #ifdef LINUX bit from just one file, i.e driver-bitmain.c. I fired it up and it seems to run, but obviously have not plugged it into an S1 blade as yet. I also need to change / add the USB chip for my USB to TTL dongle. J4bberwock (or anyone), if you'd like the binary, I can provide it, though the build is simple (via MinGW). heres a screenshot:
|
|
|
|
|