Fury/Blizzard tuning and mods

[odyssey]

I have been running a blizzard with my laptop for about a week with no issues other than some errors which are tolerable.  I got two more in this week and tried to run all 3 on a USB hub and could never get all 3 to work at the same time.  The hash rate either falls significantly or #2 and #3 die altogether.  Thought I read somewhere here about running multiple instances of the miner?  For three, how do you distinquish one from the other if they are all on one hub?  Would love to know what the batch file or command line needs to look like to run more than one.....

What miner are you using?  cgminer or bfgminer (and if bfgminer is it from Darkwinde or is it Nwoolls preliminary support)?
I haven't run mine on windows, I've been running them on a rpi with starminer.  One thing you can try is running separate mining instances -one per miner.  As for distinguishing them, they should all show up as com ports in windows if you have the drivers installed, so it would be as simple as just specifying the corresponding com port.  In linux based OS they will show up in /dev/ttyUSBx (where x is an integer).  

I am using cgminer_tag16, that Zeusminer recommended and provided.  If I plug all 3 into a hub, will they show up as three ports on my laptop control panel?  and if so, how do I run 3 instances of them?  Three separate command lines - one for each separate port in the batch file?    Eventually, I will switch to a Rpi, but just trying to get them working with the laptop to start with.

Yes - each miner should show up as an individual com port.  To run three separate mining processes, you'd have to execute three separate command lines (one per com port).

Well I did that with just two of them - two separate commands on two separate lines in my batch files with different port numbers.  It made no difference.  It appears to never see the second command line and then the hash rate is very erratic.  Both commands are identical and nothing in between either or extra parameters???

[1713961937]

1713961937

[1713961937]

1713961937

[1713961937]

1713961937

[happydaze]

with bfgminer I run a batch file for each Fury , 2 Fury's, 2 batch files
each batch file specifies a different com port

You can run multiple devices in one instance but with two miners I run one instance per miner.  I have 2 bfgminer windows running at the same time.

[nst6563]

In your batch file, try putting start in front of the command lines.  This will launch each command line in it's own process.

Example:
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM1
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM2
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM3

[nst6563]

with bfgminer I run a batch file for each Fury , 2 Fury's, 2 batch files
each batch file specifies a different com port

You can run multiple devices in one instance but with two miners I run one instance per miner.  I have 2 bfgminer windows running at the same time.

Do you get better performance from 2 miner processes?  I'm running Darkwinde's bfgminer on rpi with a single instance for 2 fury's and it seems to work just fine.  I haven't gotten any of the reported "sick" miner status' that others have been reporting. 

[happydaze]

I don't know if better.
I can run one at 381 clk and the other at 380.  Not sure if I can do that with 1 instance.
For some reason one likes 381 and the other likes 380

Just hit 1,584 kH/s 24hr avg poolside with the one running 381 clk.   

[odyssey]

In your batch file, try putting start in front of the command lines.  This will launch each command line in it's own process.

Example:
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM1
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM2
start cgminer.exe -o stratum+tcp://us.clevermining.com:3333 -u 1FC7cRPFXUXpgGoduMy1EZcA86zbQAaPKW -p x --nocheck-golden --chips-count 6 --ltc-clk 328 -S //./COM3

OK, putting the "start" in front of  both of them launches to separate windows and I think does the same thing as having tow batch files.  Each has its own port and each has its own worker at the pool.  The stats at the pool do not look right - guess they will level out after a while.  Everything seems to be ok each miner window.  Thanks so much guys!

Eventually, these need to move to a RPI so as to not tie up my laptop.  Will save that for another day......

[deus75]

Hello...

I hope someone can help me...

I Use an IGEL System With Pimp and 1 Zeus (Blizzard).
All fine.. System load. Blizzard is Found as: USB ASIC DEV 0 Bus 2 Device 2 ID: xxxx

Then i use this Config file:

"pools" : [
   {
      "notes" : "PiMP Testing/Donations - Scrypt 1",
      "url" : "stratum+tcp://de3.miningpool.co:3902",
      "user" : "User.USER",
      "pass" : "x"
   },
   {
      "notes" : "PiMP Testing/Donations - Scrypt 2",
      "url" : "stratum+tcp://de3.miningpool.co:3902",
      "user" : "USER.USER",
      "pass" : "x"
   }

],
"api-listen" : true,
"api-port" : "4028",
"expiry" : "120",
"nocheck-golden" : true,
"chips-count" : "6",
"ltc-clk" : "328",
"log" : "5",
"no-pool-disable" : true,
"queue" : "9999",
"hotplug" : "5",
"scan-serial" : [
"/dev/ttyUSB0",
"/dev/ttyUSB1",
"/dev/ttyUSB2",
"/dev/ttyUSB3",
"/dev/ttyUSB4",
"/dev/ttyUSB5",
"/dev/ttyUSB6",
"/dev/ttyUSB7",
"/dev/ttyUSB8",
"/dev/ttyUSB9"
],
"scan-time" : "30",
"scrypt" : true,
"verbose" : true,
"shares" : "0",
"kernel-path" : "/usr/local/bin",
"api-allow" : "W:127.0.0.1",
"api-description" : "cgminer v3.1.1-zm"
}

But the Miner doesn`t Started..
What i do Wrong

Greetings

[happydaze]

Does cgminer start but can't find device or does cgminer not start at all?

[happydaze]

nwoolls has been busy:

https://bitcointalk.org/index.php?topic=626361.msg7409154#msg7409154

Code:

	//max clock 383MHz, min clock 200MHz
if (chip_clk > 383)
chip_clk = 383;
else if (chip_clk < 200)
chip_clk = 200;

383 seems to work but I just started it.  I'll run it for a couple of hours to see where the hashrate stabilizes.   My first try at compiling anything  so your results may be better 

[deus75]

Does cgminer start but can't find device or does cgminer not start at all?

CGMiner doesnd Start

For Tests.. i Use an Gridseed... this miner Work..

[happydaze]

There's a missing "{" at the start of conf file you posted.  Make sure it is in yours.

Code:

{
    "pools": [
        {
            "notes": "PiMP Testing/Donations - Scrypt 1",
            "url": "stratum+tcp://de3.miningpool.co:3902",
            "user": "User.USER",
            "pass": "x"
        },
        {
            "notes": "PiMP Testing/Donations - Scrypt 2",
            "url": "stratum+tcp://de3.miningpool.co:3902",
            "user": "USER.USER",
            "pass": "x"
        }
    ],
    "api-listen": true,
    "api-port": "4028",
    "expiry": "120",
    "nocheck-golden": true,
    "chips-count": "6",
    "ltc-clk": "328",
    "log": "5",
    "no-pool-disable": true,
    "queue": "9999",
    "hotplug": "5",
    "scan-serial": [
        "/dev/ttyUSB0",
        "/dev/ttyUSB1",
        "/dev/ttyUSB2",
        "/dev/ttyUSB3",
        "/dev/ttyUSB4",
        "/dev/ttyUSB5",
        "/dev/ttyUSB6",
        "/dev/ttyUSB7",
        "/dev/ttyUSB8",
        "/dev/ttyUSB9"
    ],
    "scan-time": "30",
    "scrypt": true,
    "verbose": true,
    "shares": "0",
    "kernel-path": "/usr/local/bin",
    "api-allow": "W:127.0.0.1",
    "api-description": "cgminer v3.1.1-zm"
}

[deus75]

I have them in ...  :-(


OK 1 2 3

First:

i go to Seed Manager, and setup an New Config like "Zeus"

In this, i Put in, the Config data ..

Then i go to miner.. and set Start ?

Greetings..
 

[J4bberwock]

nwoolls has been busy:

https://bitcointalk.org/index.php?topic=626361.msg7409154#msg7409154

Code:

	//max clock 383MHz, min clock 200MHz
if (chip_clk > 383)
chip_clk = 383;
else if (chip_clk < 200)
chip_clk = 200;

383 seems to work but I just started it.  I'll run it for a couple of hours to see where the hashrate stabilizes.   My first try at compiling anything  so your results may be better 

wasn't is supposed to fail over 382?

I'm still waiting for my crystals, I should have ordered from mouser or similar with 24h shipping, but paying 20€ shipping for 2 crystals at 4€ each didn't look as a bargain.

[nst6563]

nwoolls has been busy:

https://bitcointalk.org/index.php?topic=626361.msg7409154#msg7409154

Code:

	//max clock 383MHz, min clock 200MHz
if (chip_clk > 383)
chip_clk = 383;
else if (chip_clk < 200)
chip_clk = 200;

383 seems to work but I just started it.  I'll run it for a couple of hours to see where the hashrate stabilizes.   My first try at compiling anything  so your results may be better 

I also came across that late last night (around 12am or so here) and cloned and built it.  I noticed it doesn't report the hashrate right, but the poolside is showing a combined rate of 3-3.4mhs for both fury's.  I'm running them at 380...didn't try the 383. 

[happydaze]

kramble mentioned there was a chance 383 would work.

I switched back to 380 as it seemed to be losing hashrate at 383.  I may test again later.

------------------------cut-------------------------
etc. If you have a buggy ZeusMiner unit that is not recognized, you can try using the following argument:

Code:

--set zus@/path/to/zeus:ignore_golden_nonce=1

Note that you should not use the above unless needed, and you should specify the path to the affected device. Otherwise other cp210x ASICs will be recognized as a ZeusMiner.

I can't figure out how to use:
--set zus@/path/to/zeus:ignore_golden_nonce=1

edit: going back to Darkwinde's build to try to get the hashrate back up.  24hr avg on this device has dropped to 1,518 kH/s   

[kramble]

wasn't is supposed to fail over 382?

It depends on whether they treat 255 as a valid clock speed (my FPGA code does not, see here). I'm curious about whether it works too!

I'm still waiting for my crystals, I should have ordered from mouser or similar with 24h shipping, but paying 20€ shipping for 2 crystals at 4€ each didn't look as a bargain.

In the UK I use Farnell. No delivery charge for any size order (pretty insane really, I once ordered a single device in a 144 pin TQFP package and it came packed in a 60 chip carrier array in a huge box, must have cost them more to send it than the chip was worth!).

[kramble]

I can't figure out how to use:
--set zus@/path/to/zeus:ignore_golden_nonce=1

Do you use it like this?

Code:

--set zus@/dev/ttyUSB2:nocheck_golden=1

[happydaze]

I'm running Windows - not sure if that would work.
The path seems optional.  I get an error about "doesn't have permission on the com port" (not an exact quote)

Just tried --set zus0:ignore_golden_nonce=1 and it didn't crash so maybe that is one way to use it in Windows.

edit:  so I'll try 383 again using --set zus0:ignore_golden_nonce=1 --set zus:freq=383 --set zus:cores=8 --set zus:chips=6

[kramble]

I'm running Windows - not sure if that would work.
The path seems optional.  I get an error about "doesn't have permission on the com port" (not an exact quote)

Just tried --set zus0:ignore_golden_nonce=1 and it didn't crash so maybe that is one way to use it in Windows.

I suppose Windows ought to work something like
--set zus@\\.\COM4:ignore_golden_nonce=1
or
--set zus@//./COM4:ignore_golden_nonce=1

(One of my pet niggles about DOS/windows is it's use of the wrong slash for paths, unix came first, why not follow the standards Microsoft, oh, it's because you want to do it differently just because you can. Then oops, you just broke character escaping in C, shells, etc etc </rant>)

[happydaze]

Code:

// support for --set-device
// must be set before probing the device

I think this means use set options before specify com port.  

Poolside hashrate (24hr avg) is down to 1,514 kH/s
I'll run: --set zus0:ignore_golden_nonce=1 --set zus:freq=383 --set zus:cores=8 --set zus:chips=6 -S zus:\\.\COMxx

and see if any improvement after a couple of hours

Edit: I quit it again.  Seems like it is running at stock frequencies.  Could be my build.