Topic: [Guide] Dogie's Comprehensive ASICMiner Blade Setup - page 49. (Read 580761 times)

If you get the model of PSU listed, they have a direct plug into the back plane... otherwise, you must continue to use the green connectors.

It looks like you don't have to power them individually however, just 2 connectors to the back plane

So do the back panels allow you to power them without the molex cables and from normal power supplies?

They are very small.... Good luck man, sorry to hear about the trouble.

Ahhhh - yes, that looks like the problem. Unfortunately, they still will not turn on after I fixed the yellow and black wires. I am going to try switching the super small fuse that it came with to see if that helps.

I'm looking at mine right now to compare.... Did you swap the yellow and the black? Comparing them to mine looks like they may be on the wrong sides of the green plug. Yellow for +

Hi, I cannot get two of the new blades to power up (see pic below). Any advise on what I am doing wrong? I am using a 750w Thermaltake PSU that should be more than enough power to run both boards. I have only tried to plug in one at a time so far. There appears to be a brief flash of lights the first time I plug it in and then nothing. Thanks,

{update: this post should act as a warning on how not to wire your blades. You will need to replace the fuse if you do this.}


screen shot download

Just got my V.2 blade this morning. Got everything hooked up and configured. Dogie your guide has helped me get this far and for this I thank you and have sent you a small tip..

Problem I am running into with the blade is that it keeps restarting right around 2 minutes 15 seconds. Does this sound like a power issue to anyone? 750w supply dual rail.. I used a dual molex into PCIe splitter and attached the two yellow wires on the positive screws of the green plug and two black on the negative side screws.

EDIT: I have attached each side of the molex-PCIe power splitter to two different power "chains." Honestly I don't know if the power is being pulled from two different rails or not. Has anyone used the PCIe power, like with a PCI express power extension cable as the splice straight to one PCIe power line?

Update: All is well, I guess it was just having a network/configuration problem? Seems that these blades will just keep restarting themselves if they think something is wrong... I went ahead and rebuilt BFGminer on my pi, and added libmicrohttpd-dev this time around. This runs an internal getwork server when the --http-port xxxx option is used. Pointed the blade to my pi and now all my USB sticks along with my blade are hashing together on one worker with a diff of 16 and over 19GH!

take 2 mm thick wires and solder it to your wires from PSU, then connect those wires to green connector. dont worry about 0000xxxxxxxx once you will connect your blade  properly and restart it will be gone.

Be careful do not touch board with screwdriver.

Hey guys I just got one of the older blades the first revision (red fuse) and I had it hashing away at 10.7gh right out of the box, chip read all 0 and everything was alright. I got a multimeter and started tuning all the voltages to 1.2 so that I could change the clock setting to high and mine at 13gh like everyone else.  After adjusting about 4 chips on the 5th on it started ticking, the blade. I'm not sure what exactly is ticking, It sounds like its coming from the green power connector, but its ticking away now. Still accepting shares, but ticking. So I finished tuning all the chips to 1.2v, carefully making sure with the voltage meter. And when I went to the ipadress:8000/main I now have 0000xxxxxxxx00000000000000000000

So does this mean that my PSU isn't providing enough power or that those chips can't operate at that voltage and need more or less power?

It should be noted that I'm using a small crappy diablotek psu that is only 350w, it is also powering a small htpc that is idling while I test the blade. However the +12v connection (i'm assuming the single strand with 2 molex and and one sata plug) is rated for 20A so I should be ok I thought? Especially given that it used to work without ticking or throwing x's in the chip row, before adjusting the voltages.

Yes, a larger name brand PSU is on the way. But in the event that isn't the issue, what could be causing this??

Thanks,
Greg

OK! So for anyone else experiencing this issue.. It is resolved! (I also had the reset/dropout/low-hash issue, also resolved!)

I wasn't doing anything wrong. My PSU was faulty. Maybe it had been gathering dust too long (was from my PC since the last time I upgraded it)
Got a new PSU, just a little 350W'er, plugged it in, VRRROOOM! blades powered up no hassle at all!

Put in my 2 pools in and I had some hash, but it was slow, and resetting etc.. I was using 50btc and BitMinter as my pools... with a bit of trial and error I worked out 50btc was unstable
so I now use only BitMinter and I have blades hashing away at 10Ghs on LOW without hassle. (No Proxy)
So on that issue I'd say it's most likely the pool you're connecting to doesn't play nice with the blades miner (well, it was in my case anyway!)

Hope this helps the next poor stressed out soul that his a similar issue!

The new blades werent shipped out with some sort of guide?

Make sure cable is thick enough (1,5 mm2 should be ok) to conduct 10A.

No, I'm the guy who wants definitive proof you can or you can't.  Please quote the post where I claimed you -- in fact --  could supply it with 24Vdc without potential damage.

Again: What's the P/N of the DC/DC Regulator? 

wow where can u buy this so cheap?

i think i can do this setup??

link where they sell this is owsome

Are you sending two rails of 12V power to the board?  As shown in the OP?  Do you have a meter and can confirm somewhere in the neighborhood of 11.5V - 12V at the connector?

I got one of the new blades, web interface starts but Chip shows all "x" and it doesn't hash at all.

I already tried 2 different power supplies. Any ideas? 

You're the retard who is telling people its okay to put 24V through their blades, knowing that there is a good chance it will destroy them. You can fuck off now, you contribute LESS than nothing, no one wants you here, eat my ignore list.

Because I'm not the moron who made a statement without having anything other than some arbitrary bullshit they pulled out of their ass as a reason.

Again: what's the P/N of the DC/DC Regulator?  

I have received a new modular blade and OC'd it to 1.16V but it seems that the blade keeps restarting approximately at 45 minutes mark. Something I noticed that has rate keeps rising while I have brought voltage down from 1.20 to 1.16V. It keeps creeping up to 13000MHs while my understanding is that it shouldn't.

Voltage on connector is sitting at 11.55V with a 850W Corsair professional series PSU. Only one blade and 4 fans are connected to this power supply.

At times I have seen a couple of X on the chips as well but even if there are no X blade will still restart. X's are not frequent but I guess it could also be not frequent because blade restarts.

4 X 120mm fans are cooling the blade but they are not high CFM, just average 8 bucks case fans.

Heatsink is cool to touch from sides but a little hot if I touch it at the bottom.

Should I lower the voltage? Maybe it's getting too hot?

Is there any point attaching a temperature probe to monitor temperature of the heatsink?

Any help is appreciated.

EDIT: I was keeping a close eye on the blade. Running it on Low clock was giving 10GH and almost exactly around 59 minutes mark the blade restarted. It also changed the clock to High upon restart.

yes, if you have different proxy_miner.exe run on one computer (on other port) pointing to different pools.