GekkoScience NewPac / Terminus R606 (BM1387) Official Support Thread

[zsmith]

I am running mine on 49 port hub, 10 port siopar 2.1a per port.  when they start all good.  then next day they all stop.  I am using raspberry pi, and uddo machine. it happens on both.. So that might be issues that you are talking about can't go 24 hours without all stopping.  was running them at 250mhz with powered 100v fan.

Not sure what to do to fix this.

I have not had any issues with these new-pacs but I have never ran them longer than 6 hours because my standard procedure is to use a script to restart cgminer every couple of hours.

Here is a similar code to what I use, but written for your Raspberry Pi.  You still have to edit it slightly to fit your particulars.

Save this code as mine.sh, edit the line starting with lxterminal with the full path to your cgminer folder and then put in the commands you usually use for your pool, mark as an executing program and run.  It goes in a loop so to exit, just use control C.

Code:

#!/bin/sh
echo "Restarting cgminer every 6 hours"
while true
do
 lxterminal -e "/home/USERNAME/cgminer/cgminer -o POOL_URL -u USER -p PW --suggest-diff 128"
 sleep 21600
 pkill -f cgminer
 echo "Restarting cgminer..."
 sleep 4
done

[1715326329]

1715326329

[1715326329]

1715326329

[House Xylander]

Ubuntu machine and cgminer running solid.

However, I came home to one of the NewPacs in a loop showing "0 chips found" and trying to restart

Tried changing hubs, solo instance, different computers all to no avail.

I have all of my NewPacs running at 200MHz and 860mV with ample cooling.

Is there anything else I can try or test for before sending it in for repair?

I have not had any issues with these new-pacs but I have never ran them longer than 6 hours because my standard procedure is to use a script to restart cgminer every couple of hours.

Here is a similar code to what I use, but written for your Raspberry Pi.  You still have to edit it slightly to fit your particulars.

Save this code as mine.sh, edit the line starting with lxterminal with the full path to your cgminer folder and then put in the commands you usually use for your pool, mark as an executing program and run.  It goes in a loop so to exit, just use control C.

[...]

Interesting, what do I need to change to run on Ubuntu aside from username and file location?

[zsmith]

Interesting, what do I need to change to run on Ubuntu aside from username and file location?

Change lxterminal to gnome-terminal.  I run Ubuntu myself, I just changed it to lxterminal for him because he mentioned using the raspberry pi.

[House Xylander]

Change lxterminal to gnome-terminal.  I run Ubuntu myself, I just changed it to lxterminal for him because he mentioned using the raspberry pi.

Thank you, I'm still learning my way around Linux but getting there.

I made the changes and saved to the desktop. When I double-click or right-click "Open with Run Software" it spins for a bit then nothing.

This is a fresh install so I may be missing something that it is calling for.

[minefarmbuy]

Ubuntu machine and cgminer running solid.

However, I came home to one of the NewPacs in a loop showing "0 chips found" and trying to restart

Tried changing hubs, solo instance, different computers all to no avail.

I have all of my NewPacs running at 200MHz and 860mV with ample cooling.

Is there anything else I can try or test for before sending it in for repair?

I had the same issue with a 2pac recently. Took a few attempts to to start hashing again. But essentially when I dropped the clock on it, it started right up. Let it warm up and then upped it again. Maybe worth a shot before sending back.

[House Xylander]

I had the same issue with a 2pac recently. Took a few attempts to to start hashing again. But essentially when I dropped the clock on it, it started right up. Let it warm up and then upped it again. Maybe worth a shot before sending back.

will do, thank you

I left the voltage at 860mV and tried all the way down to 50MHz with no success.

[zsmith]

Thank you, I'm still learning my way around Linux but getting there.

I made the changes and saved to the desktop. When I double-click or right-click "Open with Run Software" it spins for a bit then nothing.

This is a fresh install so I may be missing something that it is calling for.

Open terminal and run the following command from the directory you saved your script to (moving it to the home folder first is easiest, or navigate to your desktop folder):

Code:

./mine.sh

You can also add the script to your startup applications once you've tested that it works.

[House Xylander]

Open terminal and run the following command from the directory you saved your script to (moving it to the home folder first is easiest, or navigate to your desktop folder):

Code:

./mine.sh

You can also add the script to your startup applications once you've tested that it works.

Awesome! I got it to work on the first try. I did see that the first window mentions that the "-e" is deprecated. Is there something else that should be used like "--" in front of the gnome-terminal?

[zsmith]

Awesome! I got it to work on the first try. I did see that the first window mentions that the "-e" is depricated. Is there something else that should be used like "--" in front of the gnome-terminal?

Oh yeah they depreciated the "-e" for some reason in 18.04 LTS. I don't know why. I started getting this error message when I upgraded from 16.04. It will still work, but might not work on future versions of Ubuntu (18.04 is supported for 10 years). The current manual still shows the "-e" option but doesn't give any alternative, other than "--command=", which is also depreciated. What the error message suggests does not work and it doesn't work without the "-e".  I assume this is a bug in gnome-terminal that will be fixed or an alternative solution will be found if and when they actually remove it.

[rockmoney]

Unfortunately, I am on Windows 10.1 and can't help much with running these sticks on linux or other OS's. I restart my PC and cgminer at least once every 24 hours, otherwise my sticks will generally start experiencing performance degradation and may even go zombie (similar to the problems a couple of others seem to posting about). I remember a program from back in the day when USB miners were your only option to mine Bitcoin, called "cgwatcher". The program was made solely to restart cgminer at intervals set by the user. Having said that, I would presume the culprit has something to do with cgminer.

The solution to the problem seems to be the same for everyone - either manually restart cgminer and/or run a script to do it automatically for you. Also, running them at higher frequencies only seems to make matters worse, and may cause working sticks to suddenly go bad. As someone else previously stated, I have also noticed that these NewPac sticks run MUCH cooler than the 2PAC sticks (although I use the same powerful fan for cooling regardless).


Now to be a bit hypocritical. After seeing others run these NewPac sticks at such high frequencies, I began running some very closely monitored tests. The results are simply amazing! Using two Plugable USB3.0 7-Port 60 Watt USB Hubs (two sticks in each hub), I was able to achieve a hashrate of over 100Gh/s per stick. In order to make all four sticks run with stability, it took many voltage adjustments to get it right. All x4 sticks ran smoothly until I purposely shut cgminer down to change the frequencies back to stock. Here is a screenshot of my results:



- click on image to enlarge -

[gt_addict]

What sort of cooling did you use to get them running at that speed and stable?

[House Xylander]

What sort of cooling did you use to get them running at that speed and stable?

Yea, and what kind of temps are you seeing with said cooling. What are your ambient, pcb, chips and post air temps?

You could get 1Ths for under 150 watts on 8 sticks at 500MHz, lol. It's about 300MHz higher and hotter for sure than I would feel comfortable leaving alone in the house.

[zsmith]

I remember a program from back in the day when USB miners were your only option to mine Bitcoin, called "cgwatcher". The program was made solely to restart cgminer at intervals set by the user. Having said that, I would presume the culprit has something to do with cgminer.

The solution to the problem seems to be the same for everyone - either manually restart cgminer and/or run a script to do it automatically for you.

I agree that the culprit isn't the NewPacs or Sidehack, as I have running a restart script since the days of .333 G/h USB miners.  I still have my old .bat restart script for Windows archived. If anyone wants me to post it, I will.

Restarting loses only seconds, where if it goes offline you may lose minutes or hours.

[vh]

A new build is available (9d66fd3):

Windows binary: cgminer-4.11.1-windows-gekko-9d66fd3.7z

  - Fixed datetime display.

  - Closed gaps in data transmission to reduce line power fluxations.
    - Spin off dedicated work tx thread.
    - Switch from blocking get_work to get_queued.
    - Send null work to keep asic busy during network glitches and when pool is reacting to block events.

Reminder best practice update process: keep old working (or favorite) build readily available.

• move the existing "cgminer" folder to "cgminer.yyyymmdd"
• clone repository to create a new cgminer folder as noted from first post instructions.

[sidehack]

I've had this build running on my sketchiest tester with the sketchiest sticks (it's been a billboard example of the kind of instability and low hashrate folks are complaining about), without AB enabled, for the last 8 hours and I'm seeing 99% target hashrate and no zombies. Try it out.

[House Xylander]

Thank you for fixing the datetime display. I was having flashbacks every time I glanced over and saw that it was 1970-12-02.

is there a way to have cgminer display each device's hash rate or another way to monitor them aside from individual sessions?

Update: Went back to the previous release as I started getting the "missing nonces" and hashrate was lower after about  2 hours of running it.

Not sure what is happening but I know that since I switched to a Linux machine I had not seen any sign of the nonces issue.

I really thought it was a windows 10 thing now I have no idea unless something was left out in the new release by chance.

[sidehack]

cgminer does display each device's hash rate. What are you seeing?

[House Xylander]

cgminer does display each device's hash rate. What are you seeing?

Only data I see are the accepted shares and new block found announcement mostly. The average hash rates of all the sticks flash at the bottom every few seconds.

I have not figured out how to remote into the linux computer yet or I would upload a screen shot.

[nuanicaj]

A new build is available (9d66fd3):

Windows binary: cgminer-4.11.1-windows-gekko-9d66fd3.7z

  - Fixed datetime display.

  - Closed gaps in data transmission to reduce line power fluxations.
    - Spin off dedicated work tx thread.
    - Switch from blocking get_work to get_queued.
    - Send null work to keep asic busy during network glitches and when pool is reacting to block events.

Reminder best practice update process: keep old working (or favorite) build readily available.

• move the existing "cgminer" folder to "cgminer.yyyymmdd"
• clone repository to create a new cgminer folder as noted from first post instructions.

Linux update?

[sidehack]

Did you install all the dependencies, specifically libncurses5-dev?