GekkoScience Compac BM1384 Stickminer Official Support Thread

[VirosaGITS]

My understanding is that neither the "21 Inc Bitcoin Computer" nor the "Antminer R1" are released and shipping yet. While the prevailing sentiment is that they will both largely be "closed" in terms of configuration, how does anybody KNOW this?

It seems until we get actual documents, and/or a shipping product, it's kinda speculative as to what can and can't be done. I don't have a specific interest in either product, but it seems a bit premature to decide what they can and can't do at this time.

Do not know why you think we are speculating on the R1? This from the Bitmain Website.

The AntRouter R1 is a wireless networking device containing a bitcoin mining chip. The R1 is preconfigured to mine on AntPool’s solo mode, meaning that as long as the device is powered, you are competing for a chance to solve a block and win about the entire block reward of 25 bitcoins(after deducting the 1% fee).

 Shipping starts from Oct. 8 after receiving full payment.

Rich

From the Bitcoin site with their short listing up for that 10 unit test sale... from memory it was 5.5gh, the price seemed like a placehold but it was 0.02BTC. I'm not sure what the real price is? And the aforementioned detail that i posted earlier are from Q&A with Bitmain's CEO.

[1714691832]

1714691832

[1714691832]

1714691832

[HerbPean]

Quick question, My compact reset overtime (10 sec without hashing detected).

During the night it reseted 4 times, I use a diff of 16.

I will lower it to 8.

[VirosaGITS]

Quick question, My compact reset overtime (10 sec without hashing detected).

During the night it reseted 4 times, I use a diff of 16.

I will lower it to 8.

What is the question?
Also which miner software is that? I never noticed the gekko cgminer do that.

A Diff of 8 on the stock stick setting should give you 2.something shares per 10 seconds.

[sidehack]

Herbpean - what's your errors look like? Even shares that aren't registered as over diff, if they're not error shares, should clear the reset timer. We have sticks running at diff 512 that don't reset because they're submitting valid shares regardless of the pool vardiff. If a stick submitted no shares to cgminer (not the pool, cgminer), or submitted only error shares, for a period of ten seconds, it would reset. If everything's working properly (chip has good voltage, stays cool etc) that should be a very unlikely condition.

[HerbPean]

Herbpean - what's your errors look like? Even shares that aren't registered as over diff, if they're not error shares, should clear the reset timer. We have sticks running at diff 512 that don't reset because they're submitting valid shares regardless of the pool vardiff. If a stick submitted no shares to cgminer (not the pool, cgminer), or submitted only error shares, for a period of ten seconds, it would reset. If everything's working properly (chip has good voltage, stays cool etc) that should be a very unlikely condition.

The stick is set to 150 mhz, my guess it's the voltage. (I had to crank it up a bit) I have 0 HW as speak but it's only running since 1 hour. Maybe i'm too close from what my usb can handle in term of amp.

I restarted my cgminer with diff set to 8, no reset yet.

It would looks like this message

[2014-04-10 10:16:01] ANU 8 COMPAC X: No valid hashes for over 53 10 secs, attempting to reset
[2014-04-10 10:16:01] ANU 8 COMPAC X attempted reset got err:(0) LIBUSB_SUCCESS

EDIT: the chip is cold, I have a use fan pointing at him.
EDIT: other then the reset ... it's looks all good, getting 8.29gh pool side (ck solo pool with our group)

[Meech]

Never had to adjust the difficulty in cgminer-gekko and those difficulties seem way too low.
I'm steadily running @300mhz 7.2 v adj. @ 1.52a with 4 hashing about 16-17ghs.  They do get warm about 50c with a quiet 120mm fan on them.  No hardware errors in 3 days though and about 7 w per stick.

[Jake36]

Herbpean - what's your errors look like? Even shares that aren't registered as over diff, if they're not error shares, should clear the reset timer. We have sticks running at diff 512 that don't reset because they're submitting valid shares regardless of the pool vardiff. If a stick submitted no shares to cgminer (not the pool, cgminer), or submitted only error shares, for a period of ten seconds, it would reset. If everything's working properly (chip has good voltage, stays cool etc) that should be a very unlikely condition.

The stick is set to 150 mhz, my guess it's the voltage. (I had to crank it up a bit) I have 0 HW as speak but it's only running since 1 hour. Maybe i'm too close from what my usb can handle in term of amp.

I restarted my cgminer with diff set to 8, no reset yet.

It would looks like this message

[2014-04-10 10:16:01] ANU 8 COMPAC X: No valid hashes for over 53 10 secs, attempting to reset
[2014-04-10 10:16:01] ANU 8 COMPAC X attempted reset got err:(0) LIBUSB_SUCCESS

EDIT: the chip is cold, I have a use fan pointing at him.
EDIT: other then the reset ... it's looks all good, getting 8.29gh pool side (ck solo pool with our group)

Is this on the Pi that you set up the other day and had a little trouble with?

What kind of hub, USB 2.0 or USB 3.0? Pi's don't like 3.0 hub's and are really picky with 2.0 hub's, because some of them will back-feed power to the Pi.

[notlist3d]

Herbpean - what's your errors look like? Even shares that aren't registered as over diff, if they're not error shares, should clear the reset timer. We have sticks running at diff 512 that don't reset because they're submitting valid shares regardless of the pool vardiff. If a stick submitted no shares to cgminer (not the pool, cgminer), or submitted only error shares, for a period of ten seconds, it would reset. If everything's working properly (chip has good voltage, stays cool etc) that should be a very unlikely condition.

The stick is set to 150 mhz, my guess it's the voltage. (I had to crank it up a bit) I have 0 HW as speak but it's only running since 1 hour. Maybe i'm too close from what my usb can handle in term of amp.

I restarted my cgminer with diff set to 8, no reset yet.

It would looks like this message

[2014-04-10 10:16:01] ANU 8 COMPAC X: No valid hashes for over 53 10 secs, attempting to reset
[2014-04-10 10:16:01] ANU 8 COMPAC X attempted reset got err:(0) LIBUSB_SUCCESS

EDIT: the chip is cold, I have a use fan pointing at him.
EDIT: other then the reset ... it's looks all good, getting 8.29gh pool side (ck solo pool with our group)

Is this on the Pi that you set up the other day and had a little trouble with?

What kind of hub, USB 2.0 or USB 3.0? Pi's don't like 3.0 hub's and are really picky with 2.0 hub's, because some of them will back-feed power to the Pi.

I can attest to this.  The RPI does not work with all hubs.  Avoid 3.0 as there is a very good chance it just wont work.   On 2.0 if the powered hub sends power feedback the PI does not like it.

I have seen one guaranteed way to stop power feedback some say cutting the red wire works (do not do this if usb cord is attached to hub only if it's a cheap usb cord not connected).  I did this on one and it did work.

Other weird way I found to make some powered USB hubs work with pi is put a non-powered hub between the powered hub and the PI.  I tried this and it also worked on the hub I had.

Or you can read a few boards do share one confirmed to work with PI but i prefer to tinker so I bought cheap good powered ones and tinkered with them.

[HerbPean]

Is this on the Pi that you set up the other day and had a little trouble with?

What kind of hub, USB 2.0 or USB 3.0? Pi's don't like 3.0 hub's and are really picky with 2.0 hub's, because some of them will back-feed power to the Pi.

Directly to my PC (Win7)

[notlist3d]

Is this on the Pi that you set up the other day and had a little trouble with?

What kind of hub, USB 2.0 or USB 3.0? Pi's don't like 3.0 hub's and are really picky with 2.0 hub's, because some of them will back-feed power to the Pi.

Directly to my PC (Win7)

Then you can use any hub 3/2, etc.   Win 7 is a lot more forgiving on usb hubs.  So pretty much no worries besides how much power they have.

I have been playing with 3 compacs on freq got to 250 on regular cgminer 2.9.2.  Work rock solid.   Tomorrow I will get novak's version to be able to get higher.  Thanks for putting together a cgminer to get higher with these sticks!

I have a little bit of air from PC fan's blowing through and they seem to be doing very well.   

[HerbPean]

Is this on the Pi that you set up the other day and had a little trouble with?

What kind of hub, USB 2.0 or USB 3.0? Pi's don't like 3.0 hub's and are really picky with 2.0 hub's, because some of them will back-feed power to the Pi.

Directly to my PC (Win7)

Tomorrow I will get novak's version to be able to get higher. 
 

I want one !

[HerbPean]

This is what I got. No HW. 3 HW

15-10-02 01:32:30.158] COMPAC 0: No valid hashes for over 10 secs, attempting to reset
15-10-02 01:32:30.176] COMPAC 0 attempted reset got err:(0) LIBUSB_SUCCESS

[Jake36]

This is what I got. No HW. 3 HW

15-10-02 01:32:30.158] COMPAC 0: No valid hashes for over 10 secs, attempting to reset
15-10-02 01:32:30.176] COMPAC 0 attempted reset got err:(0) LIBUSB_SUCCESS

If it's directly plugged into the computer, sounds like windows 7 is probably set to save energy.

Check under the power options and see if it's set to actively suspend USB and if it is turn it off.

While your in the power settings, set the computer sleep to never.

[HerbPean]

This is what I got. No HW. 3 HW

15-10-02 01:32:30.158] COMPAC 0: No valid hashes for over 10 secs, attempting to reset
15-10-02 01:32:30.176] COMPAC 0 attempted reset got err:(0) LIBUSB_SUCCESS

If it's directly plugged into the computer, sounds like windows 7 is probably set to save energy.

Check under the power options and see if it's set to actively suspend USB and if it is turn it off.

While your in the power settings, set the computer sleep to never.

Will do thanks !

[edonkey]

While they might not win any races, these sticks are really solid performers!

This past weekend I got my two sticks up to 306 MHz and 16 GH/s each (running on Y cables with the voltage tweaked up as needed). But I was still seeing a few HW errors on one stick (the weak one of the two). So I tweaked up the voltage on that stick one more time.

So now I've been running for a couple of days with zero HW errors. It's more stable than when they were running out of the box with the stock 150 frequency (8 GH/s) and no voltage adjustment.

Note that my voltage adjustments were not scientific. I've got a meter but I didn't use it. I basically increased the frequency, and when I started to see HW errors I just made very small adjustments to the pot until they stopped. It was all by feel, but it seems to have worked well.

Thanks again to sidehack and novak for making such a cool product!

[notlist3d]

Thanks Novak for the cgminer version also a thanks to sidehack for all his work.   I went from 4.9.2 to it and am able to clock it higher.  In 4.9.2 I went to 250 and it seemed after that I was not able to go higher.

But the version for compacs from Novak is letting me go higher.  Slowly have 3 that I have been pumping up a little by little currently playing with 300 freq.  It's been fun to see how well they have worked. 

[notlist3d]

Thanks Novak for the cgminer version also a thanks to sidehack for all his work.   I went from 4.9.2 to it and am able to clock it higher.  In 4.9.2 I went to 250 and it seemed after that I was not able to go higher.

But the version for compacs from Novak is letting me go higher.  Slowly have 3 that I have been pumping up a little by little currently playing with 300 freq.  It's been fun to see how well they have worked. 

Really got to play with it for a good bit.   Was able to do up to 300 freq with no problems with 3 compacs and Y cables.   The hub seemed to start to have trouble over 300 with 3 compac and Y splitters.   Tomorrow I might try less then three on hub as 2 seemed to worked great still at 325.

But I ended up going for 275 getting around 15 GH which I am very happy with.  Have some pc fans so not a ton but a little air flowing through the compacs.   These are very fun to play with to say the least.

[sidehack]

Did you see we jacked one up a bit yesterday and got it to 488MHz? Most of the high-power parts were actually still stock.

Definitely thank Novak for the cgminer. He spent a lot of time diving into the code to figure out what needed to be done to make everything work right, and adding a lot more frequencies to the program is super nice. The ramp-up init code he worked out is also a lifesaver - without it, high frequencies would be impossible. When the chips try to init at high frequency and high voltage, there's an initial super high current draw that isn't present with low frequency startups. His init code starts all sticks at low frequency and pushes them up to your max over several seconds, which allows the loads to stay very stable and doesn't overwhelm the regulator. As awesome as the sticks are, they wouldn't be half as useably awesome without his code.

[notlist3d]

Did you see we jacked one up a bit yesterday and got it to 488MHz? Most of the high-power parts were actually still stock.

Definitely thank Novak for the cgminer. He spent a lot of time diving into the code to figure out what needed to be done to make everything work right, and adding a lot more frequencies to the program is super nice. The ramp-up init code he worked out is also a lifesaver - without it, high frequencies would be impossible. When the chips try to init at high frequency and high voltage, there's an initial super high current draw that isn't present with low frequency startups. His init code starts all sticks at low frequency and pushes them up to your max over several seconds, which allows the loads to stay very stable and doesn't overwhelm the regulator. As awesome as the sticks are, they wouldn't be half as useably awesome without his code.

I thank you both for the sticks and amazing code.  I have to admit I did not know how amazing the compac code was till switching to it from the regular version.   For it to allow such high freq and work with no effort on my part is pretty mind blowing.   These things just are able to match your needs greatly.  To be able to go from 8 to 16 easily is a big jump.

I will have to look for that one you jacked up to 488 Went to other thread that is mind blowing considering regular cgminer started at 125 or so, and around 250 on regular cgminer stopped you from pushing it further up.  Seeing 25 or so GH out of one chip ... wow.

[philipma1957]

these sticks can run at freq 250 with a little breeze with no effort at all.

weeks on end no hardware errors and under .36 watts per gh

that is pretty good.