http://www.rcgroups.com/forums/showthread.php?t=1882529

Yeah the pictures Crover posted elsewhere looked like just homemade cables wedged into place. I'm working on (should have had done weeks ago, silly real jobs) actual finished boards to tie into these supplies and provide additional capacitive filtering, external-trigger power-on and fan speed control.

For just the basics though, it only requires two pins grounded to fire it up. 60+A output though, so be very careful with stuff-wedged-into-place wiring.

Could be worse. I'm still waiting on a (non-BTC-related) package that was supposed to be delivered on the 5th, claims to have been delivered on the 9th, and I still haven't actually seen. Without said package's contents I can't finish a project we were supposed to have released today at the latest. Course we also couldn't do everything because another batch of parts that came in were 30% destroyed after spending 6 weeks in the nebulous world of international shipping, so... yeah, USPS ain't doing very well right now.

But at least SSB is on the ball.

I should have a completed PCB design in the next couple days, need to find a reliable/affordable source for the blade terminals. I'll definitely keep you posted.

Also definitely looking forward to the next GB of these guys, I just barely missed the last one.

A fuse is a fuse - as long as it meets the current rating it'll be fine. Cubes and 49-port hubs use automotive fuses. I've used automotive fuses on Blades. As long as you can get it to solder on reliably, it'll work.

Nothing, yet. Just monitoring temperatures and refining the design; been running two high-clock Cubes for 10 hours straight now and nothing on the board is even warm. Obviously the final product is going to be more enclosed/polished.

I think the picture Crover posted looks a fair bit more dodgy. Looks like they just crimped some 1/4 quick-connect tabs and shoved them in place.

Good to hear that worked. Yeah testing for power on these blades is sorta deceptive - you have to check on the downhill side of the fuse for 12V or you won't know if what's on the board is actually going anywhere.

If you measured 12V across the fuse, then it means the fuse is blown. You're getting power into the Blade's inputs, but not past the fuse. Sparks when testing voltage at the fan header could definitely cause the fuse to blow. Check the DC resistance of the fuse with no power applied - if it reads high (open-circuit) then it's blown.

The soldered-on fuses for the green-board blades are 8A SMD fuses in a 6125 package. Looks like Digikey is out of stock, but a 10A would also work (http://www.digikey.com/product-detail/en/2410SFV10.0FM%2F125-2/2410SFV10.0FM%2F125CT-ND/2766065) - that's what the blue-board blades came with by default.

If you want to test this, you can solder a 10A automotive blade fuse across the contacts in parallel with the existing fuse and if your blade lights up, that was the problem.

If you mean something like this, I haven't quite finished inventing it yet. This is a prototype for driving two Cubes; the final will have screw terminals for hooking up whatever leads you need. If anyone's interested in something like this guy, let me know and it'll spur development a bit faster.

As a note, the 750W Dell supplies run about 90% efficient (http://www.plugloadsolutions.com/psu_reports/DELL%20INC_Z750P-00_750W_SO-81_80+_Report.pdf)

Try this:

Cube Page:
IP: 192.168.0.201
Mask: 255.255.255.0
Gateway: 192.168.0.1
Web Port: 8000
Primary DNS: 192.168.0.1
Secondary DNS: 8.8.8.8
Pool Ports: 8332,8332
Pool addresses: 192.168.0.14,192.168.0.14
Miner user:pass :,???:

command prompt:
>mining_proxy.exe -o stratum.btcguild.com -p 3333


These settings are based on your page, and assuming:
   a) your network is on 192.168.0.x subnet
   b) your computer running mining_proxy.exe is at IP 192.168.0.14

Your user:pass on the config page need to be formatted correctly; make sure to repeat it, with a comma inbetween.

I've got one that did the same thing, haven't narrowed down the source yet but I'll post about it when I do.

Fuse is good? Does the relay click over?

Did you replace it with a three-wire fan? These guys need a tach signal to let them know they for sure have a fan running before they'll initialize fully.

PM sent

If I understand it, they're already working on the next gen chips. Took them a while to get started on it so no delivery until at least February. At least someone's making decent mid-entry hardware that's not built-for-n00bs or $4000.

Nope, voltage is fixed at one of two discrete levels dependent on clock selection.

Boy what happened to all the people a month or two ago that pimped cex at every available opportunity? Told everyone ever how stupid it was to buy their own equipment, just mine on cex? Couldn't find a hardware thread anywhere that didn't have half a dozen referral links scattered about.

Honestly, not really sure. I know a lot of folks would say no because of voltage differences, but it really depends on how the supplies load-balance. I know with server supplies there's no issue putting several in parallel because that's what they were built to do; ATX might not be the case. But people do double up supplies for GPU rigs and stuff, that in a less direct way couples multiple 12V rails to the same level and it works. So I guess... it's worth a shot.

Your cube and pool server aren't in the same subnet, so 255.255.255.0 won't really work - might need to change the Cube's IP to 192.168.0.x subnet

mining_proxy.exe -o stratumpool -p stratumport -gp getworkport

stratumpool needs to be the URL for your pool's stratum node (us1.ghash.io or whatever)
stratumport needs to be the port they do stratum on (probably 3333?)
getworkport needs to match the Cube's "Pool Ports" setting; the proxy defaults to 8332 if you don't specify this option.

Yeah they decided to go cheaper on the green-board V2 blades. The fuse is soldered, so no socket. No reset or test/debug headers, and it's only got one output cap on the bank VRMs instead of two. Stuff like that.

And what's wrong with server supplies? For equivalent power and efficiency they're usually cheaper than ATX, just take a bit of tweaking.