GekkoScience Compac BM1384 Stickminer Official Support Thread

[notlist3d]

Sadly, i do not see how to measure amperage to figure out what the maximum clock i should set it to with a y-splitter on an USB3.0 hub. I only have a multi meter.

Get a usb tester: http://www.amazon.com/DROK-Multimeter-Capacity-Charging-Alignment/dp/B00J3JSEG6

I just ordered 4 sets of Y cables so I can really start playing with my sticks.  Highest freq. I can get on hub right now is 262.5 without exceeding the amp rating for each port.

Yeah i know about those, but i don't want to pay 20$(I'm in Canada) to test around with my 20$ toy. I through i could check the amp with the first and 4th pin, but that didn't work out. And i don't know enough about circuits to guess how to do it without frying stuff.

I would suggest taking a look at ebay then.  If money and not timeline is the big concern you can get very low priced ones (Here is one example: http://www.ebay.com/itm/USB-Charger-Doctor-Voltage-Current-Meter-Mobile-Battery-Tester-Power-Detector-NE-/131214096046?hash=item1e8cf836ae:g:MtUAAOSwmrlUvcj9 )

But you can get it cheap just takes a while to get from China.  I'm not sure on Canada shipping but I imagine it still will be a few bucks total for one if your willing to wait.

Ah, thanks. I thought those only read Voltage. I got one for 3$ free shipping. 1 month ETA so i'll just wait until i do it before doing crazy clock on this thing.

Not a problem that is part of what I love about this compac group is the sharing of info.  I got Y cables off ebay thanks to phil for cheap that I waited on, and worked awesome.

I also got those Y cables off the link Phil gave. As soon as i receive that meter, i guess i'll try to OC it to the max. I wonder how high these things can go. I'm wondering what the bottleneck would be, the stick's own USB port?

Most likely I think it would be the power provide by the usb hub.  If you had a modified hub with more power I predict that would be highest "OC".

I could be wrong though.  But I'm guessing that is most common.

[VirosaGITS]

Sadly, i do not see how to measure amperage to figure out what the maximum clock i should set it to with a y-splitter on an USB3.0 hub. I only have a multi meter.

Get a usb tester: http://www.amazon.com/DROK-Multimeter-Capacity-Charging-Alignment/dp/B00J3JSEG6

I just ordered 4 sets of Y cables so I can really start playing with my sticks.  Highest freq. I can get on hub right now is 262.5 without exceeding the amp rating for each port.

Yeah i know about those, but i don't want to pay 20$(I'm in Canada) to test around with my 20$ toy. I through i could check the amp with the first and 4th pin, but that didn't work out. And i don't know enough about circuits to guess how to do it without frying stuff.

I would suggest taking a look at ebay then.  If money and not timeline is the big concern you can get very low priced ones (Here is one example: http://www.ebay.com/itm/USB-Charger-Doctor-Voltage-Current-Meter-Mobile-Battery-Tester-Power-Detector-NE-/131214096046?hash=item1e8cf836ae:g:MtUAAOSwmrlUvcj9 )

But you can get it cheap just takes a while to get from China.  I'm not sure on Canada shipping but I imagine it still will be a few bucks total for one if your willing to wait.

Ah, thanks. I thought those only read Voltage. I got one for 3$ free shipping. 1 month ETA so i'll just wait until i do it before doing crazy clock on this thing.

Not a problem that is part of what I love about this compac group is the sharing of info.  I got Y cables off ebay thanks to phil for cheap that I waited on, and worked awesome.

I also got those Y cables off the link Phil gave. As soon as i receive that meter, i guess i'll try to OC it to the max. I wonder how high these things can go. I'm wondering what the bottleneck would be, the stick's own USB port?

Most likely I think it would be the power provide by the usb hub.  If you had a modified hub with more power I predict that would be highest "OC".

I could be wrong though.  But I'm guessing that is most common.

I see, but i don't know. Sidehack's OP says a USB3.0 should support 900mA. And now we have two and maybe we can even daisy chain the splitters to get even more? I'm not sure how much current the splitter themselves can support.

I should be able to reach the 1.5A limit easily, going higher for short time might be fun to do, but i would not risk it 24/7.

[sidehack]

Actually it's novak's OP. I know cgminer can have issues with USB3 but some folks have tested 2.0 hubs with 3.0 power availability with good results. I don't have anything USB3 to play with.

The power bottleneck in any system is more likely to be your hub or cooling than anything on the Compac itself. If the stick is provided with stable power over 4V, it'll push enough current to melt your ASIC without adequate cooling. Voltage drop within the hub, and burst current requirements causing input voltage ripple, can result in instability of the stick. With my modified hub and a modified stick (increased voltage range, extra caps on the buck circuit) we got 488MHz out of it and it was probably drawing close to 4A through the jack. At that point the input power was dropping too low and the core voltage was unstable. For most people, the power bottleneck will be the USB hub.

[VirosaGITS]

Actually it's novak's OP. I know cgminer can have issues with USB3 but some folks have tested 2.0 hubs with 3.0 power availability with good results. I don't have anything USB3 to play with.

The power bottleneck in any system is more likely to be your hub or cooling than anything on the Compac itself. If the stick is provided with stable power over 4V, it'll push enough current to melt your ASIC without adequate cooling. Voltage drop within the hub, and burst current requirements causing input voltage ripple, can result in instability of the stick. With my modified hub and a modified stick (increased voltage range, extra caps on the buck circuit) we got 488MHz out of it and it was probably drawing close to 4A through the jack. At that point the input power was dropping too low and the core voltage was unstable. For most people, the power bottleneck will be the USB hub.

Woops, sorry Novak.

Sound good, that means it would not be particularly hard to draw power from 4 jacks and feed it to one usb stick. And i'm not sure i want to get to 4V, but can the turn-knob even go over 0.8V, by doing more than a 360degree turn?

[notlist3d]

Actually it's novak's OP. I know cgminer can have issues with USB3 but some folks have tested 2.0 hubs with 3.0 power availability with good results. I don't have anything USB3 to play with.

The power bottleneck in any system is more likely to be your hub or cooling than anything on the Compac itself. If the stick is provided with stable power over 4V, it'll push enough current to melt your ASIC without adequate cooling. Voltage drop within the hub, and burst current requirements causing input voltage ripple, can result in instability of the stick. With my modified hub and a modified stick (increased voltage range, extra caps on the buck circuit) we got 488MHz out of it and it was probably drawing close to 4A through the jack. At that point the input power was dropping too low and the core voltage was unstable. For most people, the power bottleneck will be the USB hub.

Woops, sorry Novak.

Sound good, that means it would not be particularly hard to draw power from 4 jacks and feed it to one usb stick. And i'm not sure i want to get to 4V, but can the turn-knob even go over 0.8V, by doing more than a 360degree turn?

Nope that is something I thank sidehack for.   It has a range where you could turn circles on the voltage knob but with the set range it wont go out of it.

1st or 2nd page has a good picture of it showing range.

[VirosaGITS]

Actually it's novak's OP. I know cgminer can have issues with USB3 but some folks have tested 2.0 hubs with 3.0 power availability with good results. I don't have anything USB3 to play with.

The power bottleneck in any system is more likely to be your hub or cooling than anything on the Compac itself. If the stick is provided with stable power over 4V, it'll push enough current to melt your ASIC without adequate cooling. Voltage drop within the hub, and burst current requirements causing input voltage ripple, can result in instability of the stick. With my modified hub and a modified stick (increased voltage range, extra caps on the buck circuit) we got 488MHz out of it and it was probably drawing close to 4A through the jack. At that point the input power was dropping too low and the core voltage was unstable. For most people, the power bottleneck will be the USB hub.

Woops, sorry Novak.

Sound good, that means it would not be particularly hard to draw power from 4 jacks and feed it to one usb stick. And i'm not sure i want to get to 4V, but can the turn-knob even go over 0.8V, by doing more than a 360degree turn?

Nope that is something I thank sidehack for.   It has a range where you could turn circles on the voltage knob but with the set range it wont go out of it.

1st or 2nd page has a good picture of it showing range.

OP has a picture that show up until 0.8 volt, which seem fairly low and reasonable and i doubt i would melt my stick with that with the airflow i have available. Unless it get knocked off somehow, that would be too bad. I guess we could use those thermal switch thing i forgot the name.

But they were recommended for the S5 when people were panicking about the firmware shutting down fans but not hashing when they lost internet connection.

So if the max voltage is 0.8 then i'll oc to the max with 0.8

[sidehack]

If the stick is provided with stable power over 4V

This means input. Crappy hub will drop voltage when current increases. If it goes below about 4V the core voltage buck resets. You are right, you don't want to get to 4V because it means your hub is probably overheating.

[chiguireitor]

... Unless it get knocked off somehow, that would be too bad. I guess we could use those thermal switch thing i forgot the name.

But they were recommended for the S5 when people were panicking about the firmware shutting down fans but not hashing when they lost internet connection ....

Thermocouple is the word you're looking for.

[Landy1264]

Has anyone noticed some "sweetspots"? My sticks runs @150 MHz with a few HW errors (about 3-5/day) but with the same setup @160 MHz with no errors for days!

[Luke-Jr]

Has anyone noticed some "sweetspots"? My sticks runs @150 MHz with a few HW errors (about 3-5/day) but with the same setup @160 MHz with no errors for days!

I run at 275 MHz with <0.01% hw errors...

[philipma1957]

Has anyone noticed some "sweetspots"? My sticks runs @150 MHz with a few HW errors (about 3-5/day) but with the same setup @160 MHz with no errors for days!

you realize

A) tweaking the volts will raise or lower your hw
B) quality of hub matters as well


So far every stick I have can do 350 without errors but some sticks will use more power to get to 0 hw

[Landy1264]

Has anyone noticed some "sweetspots"? My sticks runs @150 MHz with a few HW errors (about 3-5/day) but with the same setup @160 MHz with no errors for days!

you realize

A) tweaking the volts will raise or lower your hw
B) quality of hub matters as well


So far every stick I have can do 350 without errors but some sticks will use more power to get to 0 hw

I didn't change anything besides freq. More HW with lower freq @ same voltage and cooling.

[philipma1957]

Has anyone noticed some "sweetspots"? My sticks runs @150 MHz with a few HW errors (about 3-5/day) but with the same setup @160 MHz with no errors for days!

you realize

A) tweaking the volts will raise or lower your hw
B) quality of hub matters as well


So far every stick I have can do 350 without errors but some sticks will use more power to get to 0 hw

I didn't change anything besides freq. More HW with lower freq @ same voltage and cooling.

Have some fun repeat the test.
Go back to freq 150 run for 2 hours.
Then go to freq 160 run for 2 hours.
See if the results repeat..

You could have had a condition change and the test won't do the same

[chiguireitor]

I've run 400mhz for 72 hours with 0 HW. It is just a matter of fine tuning voltage until you hit the sweet spot. Btw, USB ports are ver VERY picky, and keep them cool with an attached fan to that heatsink.

[notlist3d]

I've run 400mhz for 72 hours with 0 HW. It is just a matter of fine tuning voltage until you hit the sweet spot. Btw, USB ports are ver VERY picky, and keep them cool with an attached fan to that heatsink.

I would agree if you are running OC'ed use a fan it makes a big difference compare to just heat sink.  I run my the production at around 15 GH's with no issues of heat, I did above and just seemed a little hotter but I'm sure some are fine with that I just like it cool.

I used some PC fan's in a row to push air and they do good.  Nothing special about them just have USB mounted close enough it will get a good amount of air from them.  And I have some U3's down the row that benefit from the cool fan air past the compacs.  So works quite good for me.

[chiguireitor]

I've run 400mhz for 72 hours with 0 HW. It is just a matter of fine tuning voltage until you hit the sweet spot. Btw, USB ports are ver VERY picky, and keep them cool with an attached fan to that heatsink.

I would agree if you are running OC'ed use a fan it makes a big difference compare to just heat sink.  I run my the production at around 15 GH's with no issues of heat, I did above and just seemed a little hotter but I'm sure some are fine with that I just like it cool.

I used some PC fan's in a row to push air and they do good.  Nothing special about them just have USB mounted close enough it will get a good amount of air from them.  And I have some U3's down the row that benefit from the cool fan air past the compacs.  So works quite good for me.

Yeah, the fan keeps the center of the heatsink around 43ºC, which is fairly cold.

[salaman112]

Very nice lottery miners

[sobe-it]

I had an old burnt up bfl jalapeno, so I took the heat sink fan combo and mounted it onto a GS stick. I also had some stick on heat sinks from the s1s and put them on the back side of the pcb for a little extra cooling. OC is at 400 right now and running smooth for a few hours with 0 hw errors, I will have to mod my usb hub to get any higher clocks but this baby stays cool. I usually see 22 to 28 ghs. This baby will be pushed to the max and hopefully survive, I do have plans of getting a few more and doing other odd ball cooling and OC.

[Luke-Jr]

FWIW, BFGMiner can theoretically go up beyond 1000 MHz now... try at your own risk :p

[sidehack]

By reducing the resistance of the little unmarked resistor above left of the pot, you can increase the adjustable voltage range. Our hacked-up test stick "Max" has been altered to run between about 615 and 900 millivolts (and ran stable at 488MHz before my hub wasn't good enough to keep up). In order to keep power stable at higher draws, you'll want to make sure your USB hub has very low impedance on the power rails and make sure the voltage at the Compac always stays above 4V even during brief high-current surges. The input capacitors buffer out some of that but at very high current draw (400MHz/800mV is about 12.5A but the buck circuit is good for about 17A before any parts are out of spec) the input voltage may sag and cause havoc to the power controller. Beefy input lines, preferably with a big capacitor near the socket, will help quite a bit in pushing these things farther.

That's a big friggin' heatsink. Awesome.