ASICMINER: Erupter Blades. Review, comments, photos, and discussion!

[SebastianJu]

I would like to know what happens if one chip on the pcb is broken. I heard from another asicboard that is developing now that the asics are chained, so one broken one will make all chips useless. Is it the same with asicminer-boards or are they constructed another way?

It depends how broken they are. If it's just a normal dysfunctional nonce calculation scenario, the last chip on our board could run till the end of the world even if the other 31 chips are broken. And I believe it is the same with chained design.

But if one of the chips are seriously damaged. Our board does not work with chips having inner short-cuts, while chained design does not work with chips having inner disconnections. Both could be solved by a set of wire jumpers and an equivalent resistor though, but they introduce extra complexity to the board and do not always work.

Thanks... i wonder if such chip cant be replaced simply too.

[1715090623]

1715090623

[1715090623]

1715090623

[1715090623]

1715090623

[1715090623]

1715090623

[rammy2k2]

In the package is the PSU connector included too ? I mean, i will have the wires from PSU to connect to the board, will it be the connector in the package too ?

[rammy2k2]

my molex is getting really really hot ... any ideas ? i have 4 blacks , 2 by 2, and 1 yellow each hole...

[rammy2k2]

so i got this ugly surprise ...

[Caesium]

http://en.wikipedia.org/wiki/Molex_connector

"There are three typical pin sizes: 1.57 mm (0.062 in), 2.36 mm (0.093 in), and 2.13 mm (0.084 in). The 1.57 mm pin can carry 5 A of current, while the 2.36 mm can carry 8.5 A. Because the pins have a large contact surface area and fit tightly, these connectors are typically used for power."


These blades require 10A and thus even the meatiest molex can not power them. Please get these wired up properly people, before someone starts a fire :/


Edit: attempt to be more helpful:

There's a dozen posts telling you how to wire these up but there's really only a couple of things you need to get right in my opinion.
1 - don't exceed the amps that the PSU rail can supply. I was powering four of these puppies off a 750W Corsair that could supply 62A on one rail for a week without a single issue. Single rail PSU guarantees that every yellow wire coming out of that box is backed up by 62A (though that doesn't mean that one WIRE can supply 62A)

2 - chop off all the plugs and just wire the yellows and blacks straight from the PSU into the blade connector. This removes points of failure and eliminates any weak points (generally, the plugs are the weak points). I had mine wired up using 2x18AWG cores. That is two yellows, and two blacks, each 18AWG, into each blade. I chopped off the PCI-e connectors to do this.
Over short runs, 18AWG should do 16A (apparently) and by running two cables to each blade I'm only asking each pair to do 5A - well within my comfortable margin. And no connectors to worry about!

Disclaimer: not an electrician but my 4 blades ran just fine and didn't melt anything by doing the above (moved them on now but as far as I'm aware they're still fine).

[rammy2k2]

everything is ok now, blade running for 8h+, i took wires from 2 PCI-e rails and wires are not even warm 

[Phil21]

everything is ok now, blade running for 8h+, i took wires from 2 PCI-e rails and wires are not even warm 

Yes.  PCIe rails are the best bet.

Remember to use 4 conductors (pins) to hit the rated current specs (even de-rated for bundling, 10A is fine over two pair), and keep distances short.  I've found trailer hitch wire (4 conductor 16ga) to work well and is readily found and not messy if you need to go a bit further.  Also use both screwdown terminals on the blade, of course.

[josh_nc]

I would like to know what happens if one chip on the pcb is broken. I heard from another asicboard that is developing now that the asics are chained, so one broken one will make all chips useless. Is it the same with asicminer-boards or are they constructed another way?

It depends how broken they are. If it's just a normal dysfunctional nonce calculation scenario, the last chip on our board could run till the end of the world even if the other 31 chips are broken. And I believe it is the same with chained design.

But if one of the chips are seriously damaged. Our board does not work with chips having inner short-cuts, while chained design does not work with chips having inner disconnections. Both could be solved by a set of wire jumpers and an equivalent resistor though, but they introduce extra complexity to the board and do not always work.

Friedcat:

I got my 2 blades last night and I am having problems. I also sent you an email. The voltage on all 8 capacitors on both blades are all around 8mV instead of 1.5V. I tried everything I could think of to try and resolve this. It is strange that the green lights flicker from time to time and when I connect the ethernet cable to the router my wireless recognizes the blades and post the ip and macid but when I run the ethernet to my pc it will not connect to the web. Any idea what could be causing this?

J

[radiumsoup]

...and when I connect the ethernet cable to the router my wireless recognizes the blades and post the ip and macid but when I run the ethernet to my pc it will not connect to the web. Any idea what could be causing this?

J

I don't know about the voltage stuff or der Blinkinlights, but as for the networking, you should be plugging the blades into the router, not into your PC (unless you're performing NAT with your PC via Internet Connection Sharing or something similar, which is not recommended at all, really.) Keep it simple: plug your blades into the router, or at minimum an ethernet switch that is then plugged into one of the router's LAN ports. Your PC should still be able to see the blades on the network that way, as long as you don't have two routers or are doing fancy things with multiple subnets (even accidentally).

[rammy2k2]

ok ... so after 1 week uptime .. my blade restarted ... 

[rammy2k2]

ok ... so after 1 week uptime .. my blade restarted ... 

sorry, it didnt, it was BTCguild fault .. phew 

[mrb]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

[dogie]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

[Unacceptable]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

Overpriced Yes,retarded NO !!!!!..........at least they didn't do the preorder crap..............................

[dogie]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

Overpriced Yes,retarded NO !!!!!..........at least they didn't do the preorder crap..............................

I said retraded not retarded.

[Unacceptable]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

Overpriced Yes,retarded NO !!!!!..........at least they didn't do the preorder crap..............................

I said retraded not retarded.

I get dyslexic........... 

[organofcorti]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

Overpriced Yes,retarded NO !!!!!..........at least they didn't do the preorder crap..............................

I said retraded not retarded.

I get dyslexic........... 

Don't feel bad. dogie also wrote "oh great profit" instead of "oh great prophet", which is a choice insight into his mind

[Unacceptable]

C'mon people, it will make BTCs back. At most in a year, but 6-7 months seems more precise timeframe.

Difficulty won't rise exponentialy forever. In a year it can be at 300M, which will make AM still profitable (energy-wise).

Difficulty does not have to rise exponentially forever. It just has to rise 35% month-over-month, for merely 12 months, and the buyers will never make their money back: https://bitcointalk.org/index.php?topic=178275.msg1889860#msg1889860

For the record, 35% per month for 12 months would bring the network to ~2000 Thash/s total, which half of it is already being built or planned by various entities (100TH-Mine, BFL, ASICMINER, OEM Avalon Chip Customers, etc).

As I suspected/predicted, difficulty continued to increase significantly. Those who bid 75 BTC for a 12.8 Gh/s blade in late April 2013 never made their BTC back. They spent 75 BTC to mine ~40 BTC so far. They will never mine more than ~45 BTC in the rest of the lifetime of their equipment. Holding BTC would have been smarter.

I guess these blades have at least sentimental value, being some of the first ASICs.

Yes oh great profit tell us what next. Blades were one of the most retraded ASICs ever.

Overpriced Yes,retarded NO !!!!!..........at least they didn't do the preorder crap..............................

I said retraded not retarded.

I get dyslexic........... 

Don't feel bad. dogie also wrote "oh great profit" instead of "oh great prophet", which is a choice insight into his mind

   Good catch !!!!! I feel better now 

I hope to see these 10-13GH miners come down even lower,say 1-2 BTC    Come on friedcat 

[dogie]

I get dyslexic........... 

Don't feel bad. dogie also wrote "oh great profit" instead of "oh great prophet", which is a choice insight into his mind

That was also on purpose, to add sarcasm that he's not a god but someone who took a 50/50 guess and then necro'ed a 5 month old thread to gloat about an outcome we already knew and agreed upon.

[organofcorti]

I get dyslexic........... 

Don't feel bad. dogie also wrote "oh great profit" instead of "oh great prophet", which is a choice insight into his mind

That was also on purpose, to add sarcasm that he's not a god but someone who took a 50/50 guess and then necro'ed a 5 month old thread to gloat about an outcome we already knew and agreed upon.

I suck at sarcasm.