|
1101
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 28, 2013, 11:19:53 AM
|
What's that pic ... the new boards? and what's that pic ? ^^  Microwaved chips  I think the string project makes use of the internal voltage regulator of the bitfury chips. A few months back I had read that these chips had this feature where they could be connected in serial to obviate the need for a separate buck regulator. |
|
|
|
|
1102
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 28, 2013, 05:58:03 AM
|
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?  |
|
|
|
|
1106
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 27, 2013, 01:59:45 AM
|
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. Yes, it gets quite hot. I believe they are also constructed do dissipate the heat to the board, but I did not do any research in that matter. In my opinion the regulator heatsink is much more important than chip heatsinks. +1, the most important task before serious overvolting of the h-boards (0.8V+) is to first stick heatsinks on the back side of the board under the regulator. If possible, stick a heatsink on top of the regulator and the inductor. This way you can take them as high as 1V. The board will dissipate close to 70Watts at 1V. If you're seriously at the point where you're heatsinking the coil, I'd suggest just getting a part with a higher saturation current as there are pin/footpring-compatible replacements in the same series. This version has sat. current of 55A: http://www.digikey.ca/product-detail/en/PA0513.441NLT/553-2076-1-ND/3687452. FYI the stock part is a PA0513.441NLT which has a sat. current of 35A. I don't know what other impacts this would have on the performance of the regulator, so don't try this unless you're willing to experiment. Thanks for the link. I have to get back to the data sheet of the regulator and check the specs of the inductor. |
|
|
|
|
1108
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 27, 2013, 01:56:44 AM
|
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. Yes, it gets quite hot. I believe they are also constructed do dissipate the heat to the board, but I did not do any research in that matter. In my opinion the regulator heatsink is much more important than chip heatsinks. +1, the most important task before serious overvolting of the h-boards (0.8V+) is to first stick heatsinks on the back side of the board under the regulator. If possible, stick a heatsink on top of the regulator and the inductor. This way you can take them as high as 1V. The board will dissipate close to 70Watts at 1V. Can you recommend a source for suitable heatsinks? a) it's copper b) it's low profile c) tried and tested on my cards |
|
|
|
|
1109
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 26, 2013, 12:28:59 PM
|
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. Yes, it gets quite hot. I believe they are also constructed do dissipate the heat to the board, but I did not do any research in that matter. In my opinion the regulator heatsink is much more important than chip heatsinks. +1, the most important task before serious overvolting of the h-boards (0.8V+) is to first stick heatsinks on the back side of the board under the regulator. If possible, stick a heatsink on top of the regulator and the inductor. This way you can take them as high as 1V. The board will dissipate close to 70Watts at 1V. |
|
|
|
|
1111
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 25, 2013, 02:43:16 AM
|
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. |
|
|
|
|
1112
|
Bitcoin / Mining support / Re: [GUIDE] BitFury Miner Support/Tuning
|
on: October 24, 2013, 11:25:59 PM
|
I NEED AN IMAGE FOR THE SD-CARD FOR V1 MBOARD FROM AUGUST.
Every link on the forums says "account has been disabled due to high traffic". This applies to all "drop box" links. I NEED AN IMAGE FOR V1 MBOARD SDCARD, BOTH MINE DIED.
Please and thank you...
I will try to get it to you in 30 minutes  |
|
|
|
|
1115
|
Bitcoin / Mining speculation / Re: $12366, first data point.
|
on: October 24, 2013, 02:52:40 PM
|
$12366 is the amount of money a consumer must spend on average today to have delivered to their home within 2 days the 534 Gh/s currently required to mine 1 bitcoin per day (24 hours).
This number was arrived at by using ebay prices for obtaining 60 Gh/s Butterfly units, the cheapest source I could find ($1400 each) for immediate delivery of hash.
I do not have any historical data for cost of immediate delivery of hash vs amount required to mine 1 btc/day, so I consider this the first data point. It is also important to note that every day more hash comes online, meaning that this datapoint is only really valid until the next diff increase, and is only an average of the cost across the entire 2048 diff bracket we are currently in.
While manufacturers can certainly generate hashing units at a far lower cost, I am looking at only what end user consumers must pay.
If others have information on previous datapoints it would be of some value for plotting the trends and someday determining the direction and velocity of the changes, and forecasting future prices of bitcoins.
Unless the manufacturers have paid off their NRE costs, you may have to amortize it. |
|
|
|
|
1116
|
Alternate cryptocurrencies / Altcoin Discussion / Re: Litecoin had it?
|
on: October 24, 2013, 02:47:01 PM
|
Your right, I dont believe we need alternate cryptocurrencies, Bitcoin works well.
Maybe Alternative Coins section would be better place for this thread
I didn't say we didn't need them, I just thought that in terms of market price, an increase in Bitcoin confidence naturally means that other currencies should go down. After death of Atlantis, LTC did not have any new application, though it's well supported by miners. |
|
|
|
|
1119
|
Bitcoin / Hardware / Re: [ANN] US/North American Bitfury sales now open ***full prototype pics***
|
on: October 24, 2013, 12:49:49 PM
|
Exactly same thing happened. I desoldered the diode, but GPIO functionality is gone. The rPI cannot detect any h-Card. Did your GPIO functionality return? I guess I will order a new PI.
Before you do, measure the 12V at the end of the M-board. I was measuring 4V and the pi also didn't see any H-card. I replaced the psu and all was ok again. (I didn't have a blown diode on the rpi) Which points on the M-board exactly? Thanks My 12V rails measure 11.98V and the PSU power usage is 550W at the wall on a rPI that works. With the rPI with broken GPIO, the 12V rails measure 12.2V, power usage is 70W, and rPI does not detect a single chip / h-board. The rPI boots up and networking works. |
|
|
|
|
1120
|
Economy / Service Announcements / Re: [ANN] Bitfury ASIC sales in EU and Europe
|
on: October 24, 2013, 12:05:02 PM
|
My Raspberry PI got shorted out, a diode on the backside burnt out. I had upgraded from v1 Mboard over to v2.3 Mboard. It lasted just 5 minutes or so before burning out. (I had oriented it correctly, HDMI port facing upwards, etc etc. It mined for a few mins as well ). Now back to v1 Mboard. I think the PI does not like being fed power through the pins, and rather expects a micro-usb plug.
I had read sometime back that people complained of red-hot smoking rPI after plugging it in on the new Mboard. Anyone else concur?
I was wondering, mind you out of nothing, I'm not an electronic engineer, do you have fans attached to your mboard? Can fans attached to the mboard help preventing this "absorbing" such a spike? spiccioli. ps. don't be too rude when answering...  I think fans probably will not help with transients. Basically the component that blew out is this (D17). It makes sense, I had just powered cycled the M-board, when this guy probably sensed transient voltage, and started destroying itself. After a few minutes magic white smoke. http://www.littelfuse.com/products/tvs-diodes/surface-mount/smbj.aspxSurface Mount TVS Diode tvs diode Web definitions A transient-voltage-suppression diode is an electronic component used to protect sensitive electronics from voltage spikes induced on connected wires. The device operates by shunting excess current when the induced voltage exceeds the avalanche breakdown potential.
Apparently it smokes up when it senses transient power from the 5V out lines on the GPIO headers. To be on the safe side, I recommend de-soldering D17, the black TVS diode on the back of PI, near the micro-USB plug and green fuse. I have removed it from my other working rPI, and it seems to be pretty happy since last one day. |
|
|
|
|
Show Posts
