Just wanted to add my thanks as well. Thanks giga and team!
I'd be happy to come back if/when you reopen.
|
|
|
These power cables are only intended to carry 75w each, while 8 pin connectors can carry 150w each.
This is PCIe spec, not molex. Well, each 6 pin has 3x 18 gauge awg 12v + ins, which should hold 10 or 11 amps each in theory. That means a max wattage of 360 - 400w each, but you get pretty close to the max at 600w load. If the psu cheaps out and uses 20 or 22 gauge awg you will probably run into problems. Contacts (e.g. pins) have a current rating as well, which is independent from the capacity of the wire.
|
|
|
We've had this incident once (happened overnight) with a rig that had both cables plugged. How well they were plugged, that's another question. The Molex Mini-Fit Jr is an ok connector and should handle 13A on each circuit, but sometimes it doesn't connect properly and you get arcing and overheating. And sometimes PSU manufacturer saved by using thinner cable Hi Punin, as I'm sure you know, 13A/pin is dependant on the type (and quality) of the pin used in the housing. I think it's quite possible, even likely, that the contacts used in the PSU adapters are the culprit, at least in my case.
|
|
|
One of my rigs is V1 MB based and has been running heavily OC'd (>500GH/s across 16 boards) for months. I noticed yesterday that the white plastic male connectors (minifit jr 6P) were darkening (near the contacts) due to heat damage - fortunately there was no damage to the PSU or PSU cabling. The connectors were hot to the touch, likely due to the current handling capabilities of the pins are being exceeded. I'll upload a pic when I get home
This rig is in my basement which is very, very cool. PSU is a Corsair AX850. No damage has been observed on any connectors (board or cable) either on this rig or my other two. Heatsinks are installed on every HB and 3 HSFs per rig.
I'd advise everyone to check connectors regularly, and to exercise caution when overclocking.
|
|
|
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 to Steamboat & Kano & others who helped save the Klondike Works like a champ!!!!!!!!!!! What he said. Personally, I'd given the K16 up for dead based on the technical challenges and the loss of the primary developer, combined with the Yifu chip fiasco. Major kudos to all involved in finishing this project! I'd also like to point out Steamboat's remarkable integrity and perseverance. He could have cut and run a long time ago, but he stuck it out and is holding up his end of the deal. Great job Steve!
|
|
|
That's a picture with a chip mounted on a PCB right after they got the chips in hands(can't remember how many days. 1-3 max), but BEFORE THAT they did publish a pictures with their PCB without the chip. Since you seem to not know that i will let you search the thread after the picture, but i'm sure that many can confirm this if you don't believe me. HashFast hasn't published any single pic (not even a single box like BFL did with their box of fans) and let's see how long will it take them after they get the chips in hands to publish pictures and results like KNC did.
https://www.kncminer.com/news/news-30 - August 21st.
|
|
|
I haven't noticed that problem myself, running 7 original boards with 8 "new" boards in one rig right now.. Very stable. You got lucky and I did not get lucky ordered 5 Hboards. I got one dead and 4 other shut off about 10 minutes and get very hot. Tom: I got 17 H-Boards, and not a dud amongst them - sorry about your luck. You email Dave or any of the other megabigpower email addresses?
|
|
|
I haven't noticed that problem myself, running 7 original boards with 8 "new" boards in one rig right now.. Very stable.
|
|
|
I haven't seen a single poster claiming their order has shipped today.
Am I missing something?
|
|
|
Jupiter on Giga's pool - 538GH/s avg (~24hr)
.95 fw
|
|
|
Update:Good news everyone! We practically finished packing all of 700's last week. About half of them were also shipped, and the remaining ones will be going out on monday or tuesday with whatever else we'll have packed by then. My team is confident that we'll have all remaining orders out of the door within 2 weeks, and if we really manage to pull it off I would then reopen sales for immediate shipment of v1.2 H-cards and hopefully also our new stuff! (more about this later when I get my hands on them) For those of you receiving your first units, there should've been an instructions sheet for setting up your miner inside your box. If yours came without instructions you can download it from HERE. We're also preparing a datasheet for the chip (it's currently in russian only). And last but not least I was given permission to release our M-board and H-card schematics to the public!Download in zipped eagle .sch format here:M-Board v2.3H-Card v1.2 Hi Punin, Any changes of 3D CAD files (e.g. STEP/IGES) being made available?
|
|
|
just asking so don't take my head off ..I know zip
but have you upgraded to .95 and if so does your cooling seem to improve on that as well (can't hurt imho)
anyway curious
Searing
the one miner that was roasting and I cooled it and ran .93 to get it stable at 520Gh was running 770watts on a 850psu so.... I put .95 on it and it lowered everything. 300 less watts, 490Gh and 10 less degrees C on all boards... but, soon it dropped to 470Gh so I switched back to .93 and had crap luck since. I've done a lot of experimenting on this miner the past few hours and it is very odd to see that sometimes one module will be stuck turning off and turning on 12-24 cores and then sometimes another module will do it and the other wont... but atleast one does. It is only the back right and front right that take turns doing this. I put more cooling on those both (propped it right against an window a/c unit) and no luck with .95 or .94 worse is that going back to .93 will get the dreaded FPGA lock up loops now... so it seems I can't run the one firmware that was able to deal with whatever wack-a-core issue that is with this miner. so I am back to .95 and it has dropped down to ~400Gh stable which is a joke. KNC lives in a fantasy world where they just attack one issue and think they cured cancer each time. I'll probably try to get .93 running again, it might just be luck of the roll with restarts i think i am a lucky one , i put .95 on and i got 550ghs Same here, I'd been having SERIOUS problems but .95 has finally given me the product I ordered.
|
|
|
Probably a stupid question: Is anyone's Jupiter actually working reliably at 450+GH/s?
|
|
|
Okay, so what the hell is the L/P for SSH access? No, it's not admin/admin.
EDIT: never mind L: root P: whatever_you've_set_for_webui
|
|
|
Also there will be some compensation for trade-in and chip order customers.
If you could provide some more details on the plans for this it would be greatly appreciated. Thanks Yifu will let you fuck his ass. Doubtful, that would likely(?) damage his Chi.
|
|
|
(SNIP)
If any of you have received one of these, I suggest you ask KFC to provide their source modifications to the GPL licensed cgminer source code and they're obliged to provide it within a reasonable timeframe.
I've already asked Knc support for a timeline for source code distribution. When/if I hear back I will let everyone know.
|
|
|
According to the Euro thread, they have a different style connector so the boards can't be placed in backwards.
I just read that, apparently a pot as well.. A little disappointing, a second regulator would allow these boards to perform superbly - and would only require minor board mods (i.e. copy/paste of existing DC/DC section and some changes to the 12V plane). Well, maybe in H-Board R3
|
|
|
Did new V1 Hboard had add trim pot too or only for new Hboard?
Snip New H-boards (called V2.2) work only with the new M-boards (V3). Snip We can all send Pulse & Arrow some hate mail due to the rug they pulled out from under us, leaving us scrambling to buy inductors at any cost in order to get them to the fab in time to not hold up production! Snip All the best, Dave Hi Dave; Can you provide any details on the v2.2 h-boards? Dual regulator perhaps?
|
|
|
How do we configure our miners? Can you plug in a monitor, mouse and keyboard setup to them directly, or are settings accessed via LAN?
Exactly what I was wondering. KNC said it would be "stand alone". And what pool would we join to get running? Oh boy, this'll be fun. pretty sure its all done by lan, ssh login to the beaglebone then edit settings form there. should be quite simple really. Use any pool you like! seems like everyone knows about ssh but me, how do you access a linux machine, using ssh from a windows (XP) machine? Putty
|
|
|
|