ROCKMINER ASIC miner official thread

[TrekRider]

hmmmm, maybe u guys should build a case like S3 from bitma*n...
am sorry ,but it helps in reducing temp.
and maybe some discount at the price..
i just hoping that u guys also can selling miners like hotcake

Perhaps R4 will have it

After getting screwed on 3 RK-Boxes I will not purchase from Rockminer again and urge others to look elsewhere.

So, do you get paid to troll?  Every single post that you have made is negative. Welcome to my ignore list.

No, but I do expect to get what I paid for and with the RK-box I did not.  

I'll go cry now since you no longer want to read my posts...

P.S. - welcome to my dumbass list

[CanaryInTheMine]

PSA:

Please, Don't feed the trolls!

[TrekRider]

PSA:

Please, Don't feed the trolls!

Like you are?

The fact of the matter is that the harware errors and amount of rejects on the RK-boxes is completely absurd and they do not perform at the advertised hashrate.  This is a FACT.  Prove me otherwise, there are many others with the same issues.

[kcal63]

Well I have only had my RK-box for 4 days and am still working through the kinks (Shout-out to Thomas Salvatore of Minersource for all the help -- Dude you rock!)
First prob was low hashrate, update of pi image fixed that to a reasonable level, now hashing away at 505 GHs, then it was prob updating firmware (what a weird process they chose for this a Win PC to putty into a pi to download linux files from github and process them to communicate with the RK box).
Next had a prob with one blade constantly dropping out (after more advice from Thomas and lots of head scratching and desperation checks) turned out to be a faulty USB to mini USB wire (one of the five that came with my RK).
Now trying to get updated to latest pi image (they just released the v1.3.0 image, get it while its hot ) and get firmware updated again (fingers crossed on that one) hoping it will help with the error rate (currently running at just below 10% pretty steadily) and maybe even give control of fan speed?
(I am told that the latest firmware restores the frequency setting option that I lost in the last update I installed).
 
Yeah its a bit more work than I anticipated with the unit but so far not a disaster by any means (getting 500 GHs for $650 is not bad).
Oh, anyone else a bit disconcerted when they opened the shipping box for their RK and found just the unit, usb hub and five wires? No invoice, no instructions nothing? That threw me for a minute but then I turned to my friendly vendor at Minersource and they have been great about answering all my questions (even the dumb ones )
My point being that it is far too soon to get angry and give up on this unit, it is still new (and I must say the reconfig into the R-3 was a great idea, I may build a new frame and reconfig my RK into an R-3 myself) and new devices have issues to resolve, grow up and fix it or return it but please post info on the board here for all of us to share, not rants.

[Phosphorous]

So I can only get about 440-450Gh using the special version of CGMiner to my PC directly. What is so special about the Pi image that allows the units to perform at 500+?

Also, is anyone else seeing the problem that the blades start at 80Gh and then over the course of an hours or so move up to full speed at about 110Gh each? With my box, it seems like 1 blade starts around 105-110, then the other 3 start at 80 and work their way up.

[mstrongbow]

Thanks to a very cool member here on the forums I now have my RK-Box miner hashing over 500 Gh/s!
Rockweb is reporting between 122 and 129 Gh/s per blade. 12.6% reject rate but I am on version 1.2.1 and will upgrade asap.

Now I can work on the RK-Box miner 3d printed stacking clips and USB hub enclosure mounts (protect the circuitry and clip mount to the miner)

Happy mining all!

[himarmy]

I had run two of R3box for 3 days and it seems hashing rate is around 460 GH/s each. (rockminerWeb V0.91)

But when I check from the pool I found a lot of duplicate and stale shares as shown in picture below (more than 50% of hashing rate from these two rockminers) Please noted that before running these two rockminers, I never had these duplicate and stale share from 3 x antminer s1.

http://upic.me/show/51863458

http://upic.me/show/51863529

Is there something I missing during setup or how to fix this?

Thanks in advance.

[Billbags]

I had run two of R3box for 3 days and it seems hashing rate is around 460 GH/s each. (rockminerWeb V0.91)

But when I check from the pool I found a lot of duplicate and stale shares as shown in picture below (more than 50% of hashing rate from these two rockminers) Please noted that before running these two rockminers, I never had these duplicate and stale share from 3 x antminer s1.

http://upic.me/show/51863458

http://upic.me/show/51863529

Is there something I missing during setup or how to fix this?

Thanks in advance.

Update to new 1.3.1 pi sd card image(use a model B 512 pi made in UK with blue audio jack and use a freshly formatted class 10 sd card) Also, pi needs good clean power (5v - 1.5a)

Try setting freq to 320 at bottom of pool page in rockweb(this will cut rejects in half or better) but will lower ghs a little.

I use 1 pi for each miner so I'm not sure about 1 pi for 2 miners

Set low difficulty on ghash workers page either to 64, 128, 256 or 512 for that worker(gets rid of low diff shares)

Check plug-in pcb on each blade to make sure it's seated properly.

I purchased 5 high quality monoprice usb to mini usb cables($3 a piece) to replace cheap ones sent with unit.

Always use high quality single rail 12v power supplies with more amps than u need. I use Corsair and have never had a problem. (cx750 for this build)

One member here claimed his unit had a very excessive amount of thermal paste that was effecting chip performance before he cleaned chips up and 're-applied a higher quality thermal compound correctly.

We have seen duplicate shares before with Avalon 1 and Knc miners(it usually gets fixed through firmware updates)

Duplicate shares theoretically shouldn't hurt your ghs. I don't really agree but there are so many threads from the past years proving otherwise.
These are new chips with evolving software and Rockminer seems to be continuously giving us updates as they make them.
Duplicate shares involves the pools also, so we may see improvements there also.

[sushi]

YAY!!!  R4?  (Reseller from HKG tweeted it as well, so this must be the new one)  Congrad!!!



https://twitter.com/RockMinerInc

[rockxie]

YAY!!!  R4?  (Reseller from HKG tweeted it as well, so this must be the new one)  Congrad!!!



https://twitter.com/RockMinerInc

Yes,R4, better heat,two fans.people can put it vertically.

[rockxie]

Mining Everyone: Project R100 (ROCKMINER's 100 Experience Hot Spots Over the World).
Show your ROCKINER BOXs to others, we reward your sharing.
http://www.rockminer.com/r100en.php

[Trends]

rockxie,


Have the bugs been sorted out on the R4 which are affecting RK(R2) and the R3?

1) proper PWM fan speed control
2) Random hash rates dropping on individual blades for no apparent reason
3) Reject rate error (from 17% to 5%) although it doesn't appear on the pool I'm using

If not, maybe you should before putting on the market.

item #2 has been the biggest concern.
I have already refunded 2 customers on their RK boxes due to lack of miner stability so I'm now running 4 of your RK's and
this blade problem happens ever couple of hours to 1 to 2 blades on random boxes so it's not just one off (each RK has it's own pi now on V1.3 of your upgrade).
Not everyone can monitor 24/7 to ensure the hash rates stay up

I'm certainly not going to attempt to resell anymore of your product with the lack of stability I've experienced with these four units and would not even consider joining your R100 program and deal with all the customers issues.

Please provide a clean stable reliable product and I'm sure you will sell all you can build.

Trends


UPDATE

It seems item #2 is being caused by the pi's rebooting. Why do the pi's reboot on their own?
  

[sushi]

Trends....

Not attacking you or challenging you, but you might want to look into this as a possibility. 

I can see you might be frustrated with low hash rate but you might want to look into the whole setup conditions.  Sometimes the cause could be other than the actual Rocket or R3 miner itself...

I wouldn't mention other mining gear since this is the RockMiner's thread but there are other machines that use Rpi and does the same things if the setup is not proper.

Raspberry Pi is known to reboot on it's own when the following condition exist:

Primary suspect is power supply and actually that's the secondary as well.

Your power supply may have been ok originaly and has failed or malfunctioned.  There are many PSUs in the market that wattage and AMP specification wise, it seem to have enough but there are only so many of them are tested to be reliable on the Bitcoin or Scrypt Coin mining operations as none of these PSUs are intended to be used for the mining rigs.

rockxie,


Have the bugs been sorted out on the R4 which are affecting RK(R2) and the R3?

1) proper PWM fan speed control
2) Random hash rates dropping on individual blades for no apparent reason
3) Reject rate error (from 17% to 5%) although it doesn't appear on the pool I'm using

If not, maybe you should before putting on the market.

item #2 has been the biggest concern.
I have already refunded 2 customers on their RK boxes due to lack of miner stability so I'm now running 4 of your RK's and
this blade problem happens ever couple of hours to 1 to 2 blades on random boxes so it's not just one off (each RK has it's own pi now on V1.3 of your upgrade).
Not everyone can monitor 24/7 to ensure the hash rates stay up

I'm certainly not going to attempt to resell anymore of your product with the lack of stability I've experienced with these four units and would not even consider joining your R100 program and deal with all the customers issues.

Please provide a clean stable reliable product and I'm sure you will sell all you can build.

Trends


UPDATE

It seems item #2 is being caused by the pi's rebooting. Why do the pi's reboot on their own?
  

[dogie]

I've been paying special attention to my R3 because I had to solder a resistor back on after I damaged it (and fuck me its hard to solder with paws). Its not flickered at all and Pi is fine. Pis generally are rock solid, so if thats happening its something specific to your situation.

[Trends]

sushi

power supplies are not the issue.
Each box has individual 700W HP server power supplies that's 4 in use now plus the 2 customers 750W corsairs (don't have a model #) for a total of 6 faulty PS's, don't think so.
I now changed to yet another pool to see if that is any help.


But thanks for the advice. I want these to work as much as rockxie does.

Still no fan PWM!

PS I'm not the only one that is having this issue!

[ProudMiner]

You can give power to the Raspberry not from the USB, but from the pins, this way you can eliminate USB cable issues.




Another possibility is that it is overheating. Buy heatsinks.

sushi

power supplies are not the issue.
Each box has individual 700W HP server power supplies that's 4 in use now plus the 2 customers 750W corsairs (don't have a model #) for a total of 6 faulty PS's, don't think so.
I now changed to yet another pool to see if that is any help.


But thanks for the advice. I want these to work as much as rockxie does.

Still no fan PWM!

PS I'm not the only one that is having this issue!

[sushi]

Raspberry Pi is still very much hardware in development just like Beagle Bone Black.  

I still love Antminer S2, SP10, KNC Super Jupitor and CoinTerra (the design and engineering that goes into these two machines are amazing) but all 4 used developmental hardware and some of the problems were caused by the defective developmental boardrather than the mining gear itself.

I still think it is related to the clean power vs dirty power, but.....  If the PSU and how you power the Pi is fine, then the 3rd cause can be the Raspberry Pi is over heating.  (These are common Raspberry Pi trouble shooting...)  Normally Raspberry Pi does not need heatsink as it will take the heat, however, the ambient temperature (mining room temperature could be high and compound on top of it) happned to be a bit too high in the mining room, the heat on the Raspberry Pi could be causing the Pi to lock up and reboot.

You might be able to use this command after 40-60 minutes into Rocket Box operation or R3 operation to mesure the temperature reading of the CPU from the Raspberry Pi  

Code:

/opt/vc/bin/vcgencmd measure_temp

If this is not too high, then the debugging the Pi and checking the logfile.  What you might be looking for is the logs from just before the reboot.  Please add persistent-logs to /boot/cmdline.txt in order for the log to survive the reboot.  You may see the log start to skip consistant log sequence or temperature slowly rising.

If the OS crash on Raspberry Pi is happening, you might want to create a clean install of the Pi and see if it will solve the problem.  Would be nice to use the RockWeb Web Gui, but something similer can be used such as MinePeon with the Pi to compare if that version would reboot or not.

But some faulty Pi are known to crash at about 50-53 degrees Chip/PCB temperature (125F and up) even though the CPU chip should be able to take the abuse all the way up to the mid 70's degree C.  (This means if your room ambient temperature is about or above 35 C degrees, the chip could reach this danger zone...)

Or simply hook up the Miners to Windows or OSx cgminer.  

or the one of the USB cables somewhere could be faulty for the hashing card to under perform or not hashing as not being detected but, that should not cause the Pi to reboot on its own.  If you have one of the micro USD cable to connect your cellphone to the computer to copy your photos and videos, try to use that cable and test the Rocket Miner or R3's hub to the computer connection.  Other USB cables to test the hashing card to the Hub connections.


P.S:  Fan PWM is nice to have, but it is no good to you, if your PI is keep crashing...  Gotta take care of that first before worry about PWM.

[Trends]

You can give power to the Raspberry not from the USB, but from the pins, this way you can eliminate USB cable issues.

[Another possibility is that it is overheating. Buy heatsinks.



Thanks for the reply ProudMiner.
I have wired as shown in your pic's and I'll try the heatsinks if we determine that heat is shutting down the pi's

Trends

[Trends]

Raspberry Pi is still very much hardware in development just like Beagle Bone Black.  

I still love Antminer S2, SP10, KNC Super Jupitor and CoinTerra (the design and engineering that goes into these two machines are amazing) but all 4 used developmental hardware and some of the problems were caused by the defective developmental boardrather than the mining gear itself.

I still think it is related to the clean power vs dirty power, but.....  If the PSU and how you power the Pi is fine, then the 3rd cause can be the Raspberry Pi is over heating.  (These are common Raspberry Pi trouble shooting...)  Normally Raspberry Pi does not need heatsink as it will take the heat, however, the ambient temperature (mining room temperature could be high and compound on top of it) happned to be a bit too high in the mining room, the heat on the Raspberry Pi could be causing the Pi to lock up and reboot.

You might be able to use this command after 40-60 minutes into Rocket Box operation or R3 operation to mesure the temperature reading of the CPU from the Raspberry Pi  

Code:

/opt/vc/bin/vcgencmd measure_temp

If this is not too high, then the debugging the Pi and checking the logfile.  What you might be looking for is the logs from just before the reboot.  Please add persistent-logs to /boot/cmdline.txt in order for the log to survive the reboot.  You may see the log start to skip consistant log sequence or temperature slowly rising.

If the OS crash on Raspberry Pi is happening, you might want to create a clean install of the Pi and see if it will solve the problem.  Would be nice to use the RockWeb Web Gui,



 but something similer can be used such as MinePeon with the Pi to compare if that version would reboot or not.

But some faulty Pi are known to crash at about 50-53 degrees Chip/PCB temperature (125F and up) even though the CPU chip should be able to take the abuse all the way up to the mid 70's degree C.  (This means if your room ambient temperature is about or above 35 C degrees, the chip could reach this danger zone...)

Or simply hook up the Miners to Windows or OSx cgminer.  

or the one of the USB cables somewhere could be faulty for the hashing card to under perform or not hashing as not being detected but, that should not cause the Pi to reboot on its own.  If you have one of the micro USD cable to connect your cellphone to the computer to copy your photos and videos, try to use that cable and test the Rocket Miner or R3's hub to the computer connection.  Other USB cables to test the hashing card to the Hub connections.


P.S:  Fan PWM is nice to have, but it is no good to you, if your PI is keep crashing...  Gotta take care of that first before worry about PWM.

All great ideas sushi and I'll try the logging and commands tomorrow (it's to late tonight).
Thanks you for the reply and advice!

Trends

[forzendiablo]

cheap rbox for EU - http://www.ebay.co.uk/itm/111404652500?ssPageName=STRK:MESELX:IT&_trksid=p3984.m1555.l2649