Topic: [ANN] Bitfury ASIC sales in EU and Europe - page 54. (Read 250482 times)

Good idea for the clarification Redacted.  At first read I was like why are they using turpentine on a pcb, then I was like .......ooh.  

Paint stripper is just a heat gun - so you're just doing some hot aire rework.
  
Personally I kind of like the current suggestion of using a belt sander / power grinder to remove a bad chip - it's, umm, unique...  

If anybody decides to use a paint stripper, first cover the undamaged area around the chip with some aluminium foil to protect it from the heat. First start out slow, so the board has a chance to heat up, and then increase the heat to take it off quickly.

Oh god, that is the worst bit.


Nice job. I think there shouldnt really be an issue running the board. I know I saw earlier in this thread there was someone who "modified" the board to disconnect a bad chip and ran it without problems.

LOL, this is why I like Russian design - it can survive nearly anything. 

If it runs and nothing gets hotter then normal, you should be ok.
Next time take a paint stripper to remove a damaged chip:)

Nice job though...

intron

The picture quality is horrible, but here goes:

The power drawn of the wall remains at roughly the same as it was before this "ordeal".

It is hard to say, but if you shorted some dangerous wires, the power consumption would rise significantly. If you have a Watt-meter, check your power consumption. The number of Watts for Bitfury should be somewhere near the current hashrate in GH/s.

And I'd like to see the picture too. 

ROFL....


Video of the sanding process, or it didn't happen....

Pictures please! 

And congrats on the good job of getting it back to work! 

1. This is not a how-to repair, I have no idea what Im doing, running on pure can-do spirit here.
2. Yes I know I fucked up and I do feel like a retard, thank you for not pointing out the obvious.

Yesterday evening I swapped out one heat sink because I felt the thermal contact was bad.. the chip was hot, but the corresponding sink was not. When I plugged the board back in and powered up there was was some smoke, a few sparks so I quickly powered down.  Chip #16 on the board had a piece missin where the sparks came out and some burn marks next to it. I had plugged the board in the wron way around. Me.. the person who made the f-ing how to plug them in picture with the instructions clearly stating to double check that the components are facing the Pi.

After some cursing I turned the board the right way around and made a trial run to see how bad things were. At this point I had 4 chips missing. After some inquries here and on irc I proceeded to cut everything leading to the damaged chip. This did not go well and I ended up with 0 chips shoving up.. after this I attempted to solder the cut SJxx traces together with a soldering iron several magnitudes too large (the only one I have at hand) ..with hands that are far from being steady enough for anything like this. Things do not go well! I manage to plant a solid ball of solder on the  SJxx traces, clumping them all together. I figure what the hell, lets try it out, it's dead. I rip off the ball of solder and proceed to carefully try and scrape the cut traces I have pushed too deep in to the board clean. This leads me to having chip #1 show up and work. Im pretty clueless at this point, but figure Im moving in the right direction and try to clean things up a bit more. It does not help, cant remember if one or naught chips were working at this point. I allready write the board off in my mind but do continue thinking about solutions and eventually come up with a plan of sanding the burnt chip and everything around it down. I protect everything I havent allready broken with electric tape and proceed to sanding the damaged part. After about 30 minutes of sanding Im not happy with the result, but do feel that it wont get any better. So I clean up all the shawings, remove the tape and fire up the board: 14 chips on the damaged board show up and hash. likeaboss.jpg

The question that now haunts me is how unsafe is it to keep running this board ?

This

It is quite good news, but I have an early October order (7500euro one) with you and I got reply to my support ticket that my order will be shipped in the end of October. So, will the October orders with 6500euro ship in November? What is going on?

I am totally with you. As a child I went amok when I lost a game. With time came reason and I learned to back off when the tide started to turn. I understood that there will be other and maybe better chances to take.





http://www.thefreedictionary.com/unit

For me a "unit" is something that does what is defined. Of what this unit consists to do that is irrelevant.

The issue is that Anduck thinks that a 'unit' should mean any amount of hardware -- even if it is not what was advertised.   

Just need a confirmation of what amount of hardware was originally advertised and it will be all over.

As far as the bet goes, I think it should have been voided rather then given to the 'yes' or 'no' side because it was poorly worded.

OMG, are you still on that bitbet trip? They advertised 400 GH/s miners. If these miners consist of one or two or ten M- and H-boards? Who the f*ck cares? The result is of relevance, and the result was achieved (over-achieved by up to 50% and more in fact).

I understand your frustration and I have to admit, the wording of this bet leaves space for interpretation, but get over it. No one who bought them is disappointed or wants to give them back and that was the core question of this bet.

Buy some bitfury miners instead, if they will be available again. Maybe you get your lost BTC back with them.

You are forgetting that these are "25 Gh/s units" as well:

http://www.bitfurystrikesback.com/product/25gh-miner-starter-kit-october-delivery/

But anyone that has bothered to read the thread knows that a "full 400 Gh/s unit" was originally supposed to be as you described - 16 H-boards in one m-board, and the starters were supposed to be a single h-board and one m-board.  The fact that they ended up sending out 400 Gh/s as "two 200+ Gh/s units" means what, exactly?

DEFINITION OF A UNIT:  An m-board, a RasPi, and 1 or more h-boards....

Hey punin.

Can you confirm what amount of hardware was originally advertised for September delivery? 

My understand is that it was 16H-cards, on 1M-board, with 1R-Pi.  As per http://www.bitfurystrikesback.com/product/400gh-miner-october-2013/

Is this correct?  I understand you had to ship additional hardware, I just want to confirm that the 'unit' was originally only supposed to include this amount of hardware.

Thanks.

Really hoping for a reply to my email sometime soon to resolve the payment issue.

Also hoping the fact the October batch just sold out doesn't screw me since the order is "on hold".