[ANN] US/North American Bitfury sales NEW STOCK ***NOW SHIPPING***

[klondike_bar]

is there any way to prevent the creeping change in resistance that occurs in the hours (or even days!) after a pencil mod? 24hrs ago i modded both my boards (august and oct versions) to run at roughly 35GH

within 2hrs they ran at 36.5
within 4hrs they ran at 37.5
within 12 hours, 38.5
now, at 24hrs, they are both pushing up into the 39.5-40.5GH range and starting to induce errors and intermittent shutoff of chips in 1 board until i swiped off a tiny bit of the graphite to drop it back to 38.5GH

I dont mind them creeping up a bit, but its hard to tell if/when/where it will plateau. IME, 41GHash is about the point where (even with lots of airflow) chips start to turn off, eventually the whole board is left with 0-4 running chips (at about 2.4-2.5GH each)

[xstr8guy]

I just got some old style H cards today so I shutdown and inserted the cards.  And of course when I restarted, the SD card won't boot.  This has happened every single time I've shutdown and restarted.  Why?!  Now I have to re-image the card and try again. 

What am I doing wrong here?

I couldn't guess what's causing the problem. But I can answer your question.

SD cards are CHEAP. You should have a few already flashed, then it's a 30 second issue.

agreed. the kingston sdcards are realllllly realllly bad. if you blow on them wrong, they die. I picked up a few spare SDs for ~$6.

not sure if it helps, but i've been adding the 'sync' command before shutting down my bitfurys...like the old days.

Would you please post the full command line?  I'm not familiar with Linux.  Thanks!

[tom99]

is there any way to prevent the creeping change in resistance that occurs in the hours (or even days!) after a pencil mod? 24hrs ago i modded both my boards (august and oct versions) to run at roughly 35GH

within 2hrs they ran at 36.5
within 4hrs they ran at 37.5
within 12 hours, 38.5
now, at 24hrs, they are both pushing up into the 39.5-40.5GH range and starting to induce errors and intermittent shutoff of chips in 1 board until i swiped off a tiny bit of the graphite to drop it back to 38.5GH

I dont mind them creeping up a bit, but its hard to tell if/when/where it will plateau. IME, 41GHash is about the point where (even with lots of airflow) chips start to turn off, eventually the whole board is left with 0-4 running chips (at about 2.4-2.5GH each)

  better way to do it change resistor.

[klondike_bar]

is there any way to prevent the creeping change in resistance that occurs in the hours (or even days!) after a pencil mod? 24hrs ago i modded both my boards (august and oct versions) to run at roughly 35GH

within 2hrs they ran at 36.5
within 4hrs they ran at 37.5
within 12 hours, 38.5
now, at 24hrs, they are both pushing up into the 39.5-40.5GH range and starting to induce errors and intermittent shutoff of chips in 1 board until i swiped off a tiny bit of the graphite to drop it back to 38.5GH

I dont mind them creeping up a bit, but its hard to tell if/when/where it will plateau. IME, 41GHash is about the point where (even with lots of airflow) chips start to turn off, eventually the whole board is left with 0-4 running chips (at about 2.4-2.5GH each)

  better way to do it change resistor.

all i have is a terrible pencil iron, minimal soldering skills, and no resistor(s). Id rather not risk the downtime to destroy my board

[ktbken]

I've been using shutdown, but it reboots automaticly. so when it asks for a login i turn off the power supply. so far that hasn't hurt the SD card.

If you use

Code:

shutdown -h now

it will not reboot

[EsDva]

I've been using shutdown, but it reboots automaticly. so when it asks for a login i turn off the power supply. so far that hasn't hurt the SD card.

If you use

Code:

shutdown -h now

it will not reboot

just use

poweroff

[Ridicuss]

I've been using shutdown, but it reboots automaticly. so when it asks for a login i turn off the power supply. so far that hasn't hurt the SD card.

If you use

Code:

shutdown -h now

it will not reboot

just use

poweroff

Dont forget sudo before shutdown commands.

sudo shutdown -h now   --should work.

[Morblias]

For those of you looking for the v1 raspberry pi image, i uploaded it to mega: https://mega.co.nz/#!mRg1WDzJ!QYdEySHkHaOjyMJhxdkp8Xt4S6pz8VdUswP7_JyveY4

Note: This is the v1 that Dave put on dropbox in august.

[vs3]

is there any way to prevent the creeping change in resistance that occurs in the hours (or even days!) after a pencil mod? 24hrs ago i modded both my boards (august and oct versions) to run at roughly 35GH

within 2hrs they ran at 36.5
within 4hrs they ran at 37.5
within 12 hours, 38.5
now, at 24hrs, they are both pushing up into the 39.5-40.5GH range and starting to induce errors and intermittent shutoff of chips in 1 board until i swiped off a tiny bit of the graphite to drop it back to 38.5GH

I dont mind them creeping up a bit, but its hard to tell if/when/where it will plateau. IME, 41GHash is about the point where (even with lots of airflow) chips start to turn off, eventually the whole board is left with 0-4 running chips (at about 2.4-2.5GH each)

  better way to do it change resistor.

And if you do do that make sure to pick a resistor with the proper "temperature coefficient" - that is how the resistor value changes with temperature (which is pretty much what you've discovered here, although in your case there is an apparent "aging" of the pencil mod too - the carbon trace that you've added probably oxidizes and becomes more or less conductive).

[goxed]

is there any way to prevent the creeping change in resistance that occurs in the hours (or even days!) after a pencil mod? 24hrs ago i modded both my boards (august and oct versions) to run at roughly 35GH

within 2hrs they ran at 36.5
within 4hrs they ran at 37.5
within 12 hours, 38.5
now, at 24hrs, they are both pushing up into the 39.5-40.5GH range and starting to induce errors and intermittent shutoff of chips in 1 board until i swiped off a tiny bit of the graphite to drop it back to 38.5GH

I dont mind them creeping up a bit, but its hard to tell if/when/where it will plateau. IME, 41GHash is about the point where (even with lots of airflow) chips start to turn off, eventually the whole board is left with 0-4 running chips (at about 2.4-2.5GH each)

Graphite had negative temp coefficient or resistance, (temp, up, resistance down, vice versa). One way would be to make sure you have good airflow around the regulator, along with good heatsinks.

[buzzdave]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistance measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again. 

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

[frankenmint]

How much of a reduction in life could I expect from my miner if I OC them up too a level like 40 GH per card or approximately 640 GH/s?  I'm planning to use heat sinks on all chips and have fans on them too.  Could I possibly hope to see something that lasts a year?

[kaerf]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistance measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again. 

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

nice to see you guys selling 8 chips boards. FYI, for the one I made (https://bitcointalk.org/index.php?topic=228677.msg3259455#msg3259455), I'm running around 0.9v and pulling 21-22GH and ~23W. been too lazy to try higher voltage.

[mb300sd]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistance measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again.  

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

nice to see you guys selling 8 chips boards. FYI, for the one I made (https://bitcointalk.org/index.php?topic=228677.msg3259455#msg3259455), I'm running around 0.9v and pulling 21-22GH and ~23W. been too lazy to try higher voltage.

do i see oshpark boards? I've been placing quite a few orders there lately

[xstr8guy]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistance measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again.  

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

If we don't have a multimeter or plain just don't know what we're doing, are we better off just getting gen 1 boards?  Does the trimmer have a zero stop so we don't accidentally plug the board in with the trimmer cranked up from shipping and handling?

I'm loving the gen 1 boards that arrived today (thanks for the quick shipping!).  They're hashing above 30GH/s with no OC.  I might be happier just filling up my old M board rather than working with a new starter kit.

[kaerf]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistance measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again.  

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

nice to see you guys selling 8 chips boards. FYI, for the one I made (https://bitcointalk.org/index.php?topic=228677.msg3259455#msg3259455), I'm running around 0.9v and pulling 21-22GH and ~23W. been too lazy to try higher voltage.

do i see oshpark boards? I've been placing quite a few orders there lately

yup, they are pretty convenient and relatively cheap

[Keefe]

We are a couple days yet from getting our 16 chip boards with the trimmer on the board.  I have some 8 chip boards that are an experimental OC board that use the same trim pot and I want to set the stage for how this will work.

In order to use this trimmer, you *must* have a multimeter and know how to measure resistance.  If you don't know what I'm talking about, you aren't ready to try to OC your boards.

An easy place to take resistancevoltage measurements is to get ground off the M-board GND terminal (where you would connect direct 12V cables - the terminal screws).  The red probe would go on the top metal contact of the Pulse inductor, which is the large bulky component on the H-card.  Be sure not to also touch the caps that are nearby or you won't get a correct voltage.

While measuring the voltage, use your super-micro tweaker phillips screwdriver to *slowly* turn the trimmer clockwise for higher voltage or counter-clockwise for lower.  The trimmer has an effective range of about 180-degrees.  If you turn it down too far, you will see voltage begin to rise again.  

Don't make voltage changes quickly.  If you go higher than about .895v you better know what you are doing or you will kill your chips.  I don't even know what voltages people are getting away with on these boards.  Find a guide or post before you start mucking around.  Overclocking *will* reduce the reliability of the boards.

FTFY

We certainly don't want to connect an ohmmeter in that way!

Will you be selling any 8-chip boards? It'd be nice to eliminate the voltage regulator bottleneck on overclocking.

At what voltage would the chips be ruined if not actually mining and producing heat, such as with the chainminer process shut down? Is .895v where it just gets too hot, or where stuff starts shorting internally?

[CryptoCluster]

At what voltage would the chips be ruined if not actually mining and producing heat, such as with the chainminer process shut down? Is .895v where it just gets too hot, or where stuff starts shorting internally?

Sometimes graphite changes resistance after several hours of hashing in high temps. When voltages gets over 0.880 (measured between caps) on regular H-cards, they start to slow down. After 0.900V they slow down even further (5-15 ghs per card), until they shutdown completely. In my setup cards shutdown immediately when fired up over 0.950V. At one time I was able to measure 1.028V! The card shut down but after "repenciling" it got back to hashing as always :-)

Also there are differences between cards, I got one that runs ok. even at 0.930V.

I've got one 12 cards to experiment on. Second rig runs currently untouched.

EDIT: I have heatsinks on back of all boards in regulator area, and 120 CFM fans!

[klondike_bar]

How much of a reduction in life could I expect from my miner if I OC them up too a level like 40 GH per card or approximately 640 GH/s?  I'm planning to use heat sinks on all chips and have fans on them too.  Could I possibly hope to see something that lasts a year?

noone has been running bitfury chips more than a few months, so its unknown

[Cablez]

EDIT: I have heatsinks on back of all boards in regulator area, and 120 CFM fans!

Does the backside of the regulator even get hot? I have felt the top of the chip and its very hot but I don't feel much on the backside. Wouldn't putting a small heatsink on the chip do more?

For the hashing chips those thermal vias work well as even touching the vias themselves gets hot, even better with a heatsink.