Pages:
Author

Topic: LightningAsic usb miners based Gridseed GC3355 Tech Support Thread - page 8. (Read 73415 times)

hero member
Activity: 672
Merit: 500

I can't answer your question but am curious about your configuration.  I cannot get my gridseed 5 chip to run either dual or single without it crashing the Pi within 30 minutes.  I've tried various compiles and what knot but no luck.  I'm also headless so I can't see if there are any console errors.  It works perfect in windows but the noise is too much for me to leave running in my room.  Any information on your setup would be great.   I used to have the Pi connected to two hubs with antminers.  Now I just have the Pi connected directly to the Gridseed. Still crashes.

Another forum member told me about this, but what I did was:

Download minepeon 0.2.4.6 the newest version, then image onto a SD CARD - ( figured this be the easiest, since it already has all the dependencies installed)

then download the already compiled version of BFGMiner, and move the bfgminer binary and replace it in /opt/minepeon/bin/ . Then either sudo nano the miner.conf file or go to the browser and type in your minepeon IP ADDRESS and edit the start config there, with:

Code:
#!/bin/bash
sleep 10
/usr/bin/screen -dmS miner sudo /opt/minepeon/bin/bfgminer --scrypt -S gridseed:all --set-device gridseed:clock=850 -c /opt/minepeon/etc/miner.conf

and you should be good to go


EDIT: check out this thread, step by step:
https://bitcointalk.org/index.php?topic=168174.3000;topicseen
newbie
Activity: 37
Merit: 0
for people using raspberry pi's and cgminer... Do you guys get lower hashrate poolside, then what it shows on the cgminer dash? Because with the TPLinks, i get 3500-3700kh/s on 10 units @850 but with cgminer on the Pi, I get 1600-2500 kh/s POOLSIDE, but the cgminer shows the same hashrate as what i am suppose to be getting

It's weird, but I also have 4 plugged into a different pi, same exact config, but it shows proper hash rate pool side and on cgminer.. I'm thinking either the Hubs or something is not fast enough to handle the load, i dont know

Any ideas?

Danny


I can't answer your question but am curious about your configuration.  I cannot get my gridseed 5 chip to run either dual or single without it crashing the Pi within 30 minutes.  I've tried various compiles and what knot but no luck.  I'm also headless so I can't see if there are any console errors.  It works perfect in windows but the noise is too much for me to leave running in my room.  Any information on your setup would be great.   I used to have the Pi connected to two hubs with antminers.  Now I just have the Pi connected directly to the Gridseed. Still crashes.
hero member
Activity: 525
Merit: 500
..yeah
hero member
Activity: 616
Merit: 500
legendary
Activity: 1456
Merit: 1000
hero member
Activity: 616
Merit: 500
Good news everyone! Stable firmware for LightningAsic controllers is underway, it should be ready for public in the coming week and it will be possible to upgrade (from v1, v2, v2.1..) by uploading it to the upload page. The firmware .bin will be released for public, but it is compatible only with LightningAsic controllers. Upgrading from factory Tp-link 703N will not work and it might brick your Tp-link, so don't even try!

It has some awesome features:

  • Stripped, custom compiled OpenWrt (controller's OS) image optimized for size and speed
  • Stability is heavily emphasized and guaranteed! No more random reboots or crashes
  • Custom compiled cpuminer OC, frequency selectable up to 1200 MHz. We have tested it for 48 hours straight at 950 MHz (405 Kh/s) and without hardware errors! (see #1)
  • Advanced miner monitoring and management
  • Option to auto power cycle miners when one of the miners is stuck, this is done software wise and it has the same effect as unplugging, and plugging in the USB cable! (see #2)
  • Up to 10 miners per controller, replacing the 32MB RAM chip with 64MB will double that figure
  • Huh Shocked

note #1: depends on miner hardware
note #2: this will only work if the USB/hub is powered by the controller, you must not supply power to the USB hub.

BTC or dual mining is removed in the firmware for a multitude of reasons, BTC mining on GC3355 is unstable and wasteful.

If this doesn't want to make you pull out the LA controller out of the trash, I don't know what will...
For questions or feature requests, please PM me, not Jack.   Donations (1AMsjqzXQpRunxUmtn3xzQ5cMdhV7fmet2) are gladly accepted, and will be put towards improving the firmware even more.
sr. member
Activity: 458
Merit: 250
beast at work
for people using raspberry pi's and cgminer... Do you guys get lower hashrate poolside, then what it shows on the cgminer dash? Because with the TPLinks, i get 3500-3700kh/s on 10 units @850 but with cgminer on the Pi, I get 1600-2500 kh/s POOLSIDE, but the cgminer shows the same hashrate as what i am suppose to be getting

It's weird, but I also have 4 plugged into a different pi, same exact config, but it shows proper hash rate pool side and on cgminer.. I'm thinking either the Hubs or something is not fast enough to handle the load, i dont know

Any ideas?

Danny



hero member
Activity: 672
Merit: 500
for people using raspberry pi's and cgminer... Do you guys get lower hashrate poolside, then what it shows on the cgminer dash? Because with the TPLinks, i get 3500-3700kh/s on 10 units @850 but with cgminer on the Pi, I get 1600-2500 kh/s POOLSIDE, but the cgminer shows the same hashrate as what i am suppose to be getting

It's weird, but I also have 4 plugged into a different pi, same exact config, but it shows proper hash rate pool side and on cgminer.. I'm thinking either the Hubs or something is not fast enough to handle the load, i dont know

Any ideas?

Danny


The hashrate cgminer displays is nothing like your 'actual' hashrate, by far. If you aren't accepting any shares, the hashrate will remain the same. You have to look at the accepted difficulty and time running. Lightningasic dashboard displays hashrate calculated in that manner, or how pools would calculate it.

What i mean is,

TP -LINK shows:
3500 kh/s (PoolSide)
3300 kh/s (Webui)

CGMiner shows:
1600 kh/s (Poolside)
3300 kh/s (SSH Screen)

(thinking maybe a problem with RPi) so i plug in 3 gridseed units with the same config file and i get different results with lesser units on the Pi

another RPi: (3 units only)
1080 Kh/s (SSH Screen)
1000 kh/s (PoolSide)

hero member
Activity: 616
Merit: 500
for people using raspberry pi's and cgminer... Do you guys get lower hashrate poolside, then what it shows on the cgminer dash? Because with the TPLinks, i get 3500-3700kh/s on 10 units @850 but with cgminer on the Pi, I get 1600-2500 kh/s POOLSIDE, but the cgminer shows the same hashrate as what i am suppose to be getting

It's weird, but I also have 4 plugged into a different pi, same exact config, but it shows proper hash rate pool side and on cgminer.. I'm thinking either the Hubs or something is not fast enough to handle the load, i dont know

Any ideas?

Danny


The hashrate cgminer displays is nothing like your 'actual' hashrate, by far. If you aren't accepting any shares, the hashrate will remain the same. You have to look at the accepted difficulty and time running. Lightningasic dashboard displays hashrate calculated in that manner, or how pools would calculate it.
hero member
Activity: 672
Merit: 500
for people using raspberry pi's and cgminer... Do you guys get lower hashrate poolside, then what it shows on the cgminer dash? Because with the TPLinks, i get 3500-3700kh/s on 10 units @850 but with cgminer on the Pi, I get 1600-2500 kh/s POOLSIDE, but the cgminer shows the same hashrate as what i am suppose to be getting

It's weird, but I also have 4 plugged into a different pi, same exact config, but it shows proper hash rate pool side and on cgminer.. I'm thinking either the Hubs or something is not fast enough to handle the load, i dont know

Any ideas?

Danny
sr. member
Activity: 518
Merit: 250
I am a meat Popsicle
So stepped things up dual mining SHA and Scypt to 800Mhz, and I am getting 6.4GH/s and 340Mh/s avg per gridseed, they get mildly warm but nothing that i would be concerned about.
Anyone try dual mining under windows at 850Mhz yet, I dont want to push them too far when mining sha alongside scrypt as the sha asics can get fairly warm.
member
Activity: 86
Merit: 10
Strange. Did you have you hotplug setting set to 10 and wait 10-20 seconds for it to come back? It doesn't happen right away.

Also, the Gridseed miner will NOT disappear from device manager when disabled. It just shows as disabled (downward pointing arrow, I think).

You 5m cable could also be the problem, just replace it and try again. Easy to find out.

Sorry no hotplug is set to 0 in the bat file I use to start cgminer. I'll change it to 10 and try your bat file again next time it fails.

Update: ran again using hotplug 10, but still the same problem when it failed, had to physically replug the USB again.
full member
Activity: 238
Merit: 100
So I'm running my Gridseed for about 18 hours and I notice that cgminer is now showing "OFF" where it normally says the khashes rate, and obviously mining has ground to a complete halt - red and green lights are still flashing away however. This has happened twice now in a day. Is this common, for Gridseeds to need babysitting? I had to power cycle the Gridseed, then plug/unplug USB 2 or 3 times restarting cgminer 2 or 3 times to get it back into action.

(some people mention not to plug the USB cable in until after cgminer has started.. but for me it is the opposite which works - i.e. need to plug it in before starting cgminer)

Incidentally, this problem coincides with me switching to a 5m USB cable (was using 1m before)... I wonder if that could be the cause. I never got this "OFF" error on the 1m cable but then again I'd only been using it for 2 days anyway.
Try the fix I posted here: https://bitcointalksearch.org/topic/m.5650273

Thanks but for me this did not work. I ran the bat file 3 times, but it just wasn't helping. I was viewing device manager at the time of running the bat file expecting to see the virtual com port temporarily disappear as a result, but it did not. Odd? Here is the output:


Again, I had to physically remove the USB plug from the PC and then reinsert it. I did NOT have to power down the 12v Gridseed power supply - I had previously thought that was a necessity but evidently it is not.

Here is a screenshot from when CGMiner died:


I still think the length of my USB cable (5m!) could be a factor in these crashes. From Sunday I will switch back to 1m and see if I still get teh failz.

Strange. Did you have you hotplug setting set to 10 and wait 10-20 seconds for it to come back? It doesn't happen right away.

Also, the Gridseed miner will NOT disappear from device manager when disabled. It just shows as disabled (downward pointing arrow, I think).

You 5m cable could also be the problem, just replace it and try again. Easy to find out.
member
Activity: 86
Merit: 10
Another crash this morning. Even rebooting the PC didn't bring Gridseed back to life! Only physically unplugging and replugging the USB cable got it working again. Sad
member
Activity: 86
Merit: 10
5m is probably too long if you are running a bunch of them off a powered hub.

I'm only running 1 unit, no hub, directly connected to laptop.

FWIW the Wikipedia USB page states:
Quote
The USB 1.1 Standard specifies that a standard cable can have a maximum length of 5 meters with devices operating at Full Speed (12 Mbit/s), and a maximum length of 3 meters with devices operating at Low Speed (1.5 Mbit/s). USB 2.0 provides for a maximum cable length of 5 meters for devices running at Hi Speed (480 Mbit/s).

I assume Gridseed is USB 2.0? But if its using "Hi Speed" mode or not who knows...
member
Activity: 84
Merit: 14
So I'm running my Gridseed for about 18 hours and I notice that cgminer is now showing "OFF" where it normally says the khashes rate, and obviously mining has ground to a complete halt - red and green lights are still flashing away however. This has happened twice now in a day. Is this common, for Gridseeds to need babysitting? I had to power cycle the Gridseed, then plug/unplug USB 2 or 3 times restarting cgminer 2 or 3 times to get it back into action.

(some people mention not to plug the USB cable in until after cgminer has started.. but for me it is the opposite which works - i.e. need to plug it in before starting cgminer)

Incidentally, this problem coincides with me switching to a 5m USB cable (was using 1m before)... I wonder if that could be the cause. I never got this "OFF" error on the 1m cable but then again I'd only been using it for 2 days anyway.
Try the fix I posted here: https://bitcointalksearch.org/topic/m.5650273

Thanks but for me this did not work. I ran the bat file 3 times, but it just wasn't helping. I was viewing device manager at the time of running the bat file expecting to see the virtual com port temporarily disappear as a result, but it did not. Odd? Here is the output:


Again, I had to physically remove the USB plug from the PC and then reinsert it. I did NOT have to power down the 12v Gridseed power supply - I had previously thought that was a necessity but evidently it is not.

Here is a screenshot from when CGMiner died:


I still think the length of my USB cable (5m!) could be a factor in these crashes. From Sunday I will switch back to 1m and see if I still get teh failz.

5m is probably too long if you are running a bunch of them off a powered hub.
member
Activity: 86
Merit: 10
So I'm running my Gridseed for about 18 hours and I notice that cgminer is now showing "OFF" where it normally says the khashes rate, and obviously mining has ground to a complete halt - red and green lights are still flashing away however. This has happened twice now in a day. Is this common, for Gridseeds to need babysitting? I had to power cycle the Gridseed, then plug/unplug USB 2 or 3 times restarting cgminer 2 or 3 times to get it back into action.

(some people mention not to plug the USB cable in until after cgminer has started.. but for me it is the opposite which works - i.e. need to plug it in before starting cgminer)

Incidentally, this problem coincides with me switching to a 5m USB cable (was using 1m before)... I wonder if that could be the cause. I never got this "OFF" error on the 1m cable but then again I'd only been using it for 2 days anyway.
Try the fix I posted here: https://bitcointalksearch.org/topic/m.5650273

Thanks but for me this did not work. I ran the bat file 3 times, but it just wasn't helping. I was viewing device manager at the time of running the bat file expecting to see the virtual com port temporarily disappear as a result, but it did not. Odd? Here is the output:


Again, I had to physically remove the USB plug from the PC and then reinsert it. I did NOT have to power down the 12v Gridseed power supply - I had previously thought that was a necessity but evidently it is not.

Here is a screenshot from when CGMiner died:


I still think the length of my USB cable (5m!) could be a factor in these crashes. From Sunday I will switch back to 1m and see if I still get teh failz.
member
Activity: 74
Merit: 10
So I have a few of the newer gridseeds with the STMicroelectronics com port inside. Is there a cgminer that works with these newer ones or the modified cpuminer is the only one right now?
legendary
Activity: 1050
Merit: 1001
whats the difference between the larger red wiibox controller and the smaller controller ?
full member
Activity: 238
Merit: 100
So I'm running my Gridseed for about 18 hours and I notice that cgminer is now showing "OFF" where it normally says the khashes rate, and obviously mining has ground to a complete halt - red and green lights are still flashing away however. This has happened twice now in a day. Is this common, for Gridseeds to need babysitting? I had to power cycle the Gridseed, then plug/unplug USB 2 or 3 times restarting cgminer 2 or 3 times to get it back into action.

(some people mention not to plug the USB cable in until after cgminer has started.. but for me it is the opposite which works - i.e. need to plug it in before starting cgminer)

Incidentally, this problem coincides with me switching to a 5m USB cable (was using 1m before)... I wonder if that could be the cause. I never got this "OFF" error on the 1m cable but then again I'd only been using it for 2 days anyway.
Try the fix I posted here: https://bitcointalksearch.org/topic/m.5650273
Pages:
Jump to: