The second batch of Compac stick miners should be appearing within a week or two. I'm a bit behind schedule on manufacturing for them. We're also sold out on 750W PSU boards right now, but another batch should be ready around the first of December. I'm working on an 8-chip pod miner that takes a standard CPU cooler, made with recycle parts. There's a trade-in program going to happen, where you can trade old or dead miners for these small miners. They'll run about the same efficiency curve as the Compac stickminer (0.3-0.55W/GH or so) with an expected draw around 15W for 44GH at the bottom and about 120W for 200GH at the top. Voltage will be software-adjustable. More details in this thread: https://bitcointalk.org/index.php?topic=1203190.0If that sounds like something you'd get in on, please cast a vote. I'll use the results of that poll to determine how many I want to make. We also will soon have prototypes and get started running out a batch of new server supply boards. The board will work with DPS800 and DPS1200 supplies, and any other PSU which uses the same pinout. You'll have the option of either our standard screw terminals (8 positions per rail), or 6-pin PCIe connectors (10 per board). All the favorite features will be included - external on, power good signal, 3.3V standby, and the optional 5VDC aux supply. We're also in the process of upgrading our hosting infrastructure and adding about 50KW of power, which I should have online by the first of December. We'll be dropping the standard rate to about $65 per KW-month and there is no minimum power draw. As always, VPN remote access is free. We'll probably be offering some nifty packages like discounted PSU purchases or free use of PSUs for new customers on things like the S7 or Avalon6. This is deliberately a teaser update and more details will follow because I don't have them all yet. GekkoScience is going to be pretty darn busy the rest of the year, but it should be pretty great.
|
|
|
MacEntyre didn't raise any stink about me selling a stick to TheRealSteve over in Europe to complete his collection. I'm not gonna fight anyone anywhere in the world that wants one of MacEntyre's sticks.
|
|
|
Most of the questions anyone could ask in this thread are answered in the first post. Any question that can't be answered in the first post is more than likely either the result of a misunderstanding, misconfiguration or grounds for warranty replacement.
|
|
|
Assuming you turned the voltage up, since that's one of those things you have to do to overclock it?
|
|
|
The only firmware difference between the test and production version is updating the CP2102 with data that Luke provided, so it seems likely that he can, for all practical purposes, turn his test stick into a production stick as seen from the mining software.
|
|
|
Actually, the S7 is not critical at all since it has probably none of the parts I want - or at best, very few - for this project.
The parts that come off an S1 or S2 I might prioritize higher when I've exhausted my own stock. I don't have any S2 boards to part out (all mine are museum) but I have enough S1 boards to provide caps for 50 pods. I'll burn through those, but 30x S2 boards would run power for over 100 pods. However, an S5 board provides about 5 times the value in parts I'm short on compared to an S2 board (when component cost is considered). The design has changed a bit such that the S1 and S2 now have more valuable parts than I had originally assumed so they will be prioritized a bit higher.
However. Only trading pods for the hardware required to make more pods is not sustainable.
Only about half the cost of the pod is recycle parts, so for every pod that goes out in trade for hardware I have spent a decent number of dollars which did not come back. Additionally I spent a lot of time pulling, cleaning and readying parts for assembly on something new. I know a lot of people will be paying money for pods, but a lot of other people will not. If I don't keep a practical ratio on things, I'll end up with a whole lot of parts I can't use and a lot of PCBs I'll get ten cents a pound for and will have basically paid a lot of people to take my miners. Unfortunately, I can't trade old PCBs for sandwiches at a very favorable rate, nor can I pay the rent with them.
You can't just send me parts and assume I'll send a miner back. I will not send a pod to anyone, no matter how many boards he ships to my door, without having agreed upon the terms of the transaction first. The agreement will be based on the availability of parts I already have and how much I value the parts being provided at that moment, which means that at different times different boards will have different values because I'll need them more or less than at other times. If I find that I have enough boards already on hand to build every pod I feel like putting together ever again, I'll stop taking trades entirely and only sell for money (or other things which I find interesting or worthwhile, surprise me).
|
|
|
I've got decent wick, but sometimes it needs a bit of fluxing. I also have a 75W iron though. I tend to wick the solder as soon as the chip is lifted so there's still a lot of residual heat, otherwise it still takes five minutes.
I think I found a workable socket, but I need more information on it (what pins go to what contacts) before I could design a board around it. If I can get that information in time I can mod the Compac layout and send off for a prototype board with the two other boards we're hoping to get proto PCBs of ordered tomorrow. It's not an immediate need, but if I can get it done on time it'd save shipping.
|
|
|
I'm doing some looking now on what's available. Apparently 8x8mm 56QFN is a less-than-common package because I'm finding lots of stuff for everything to either side of that.
Maybe I can find something and throw together a modified Compac with a test socket. That'd be awful handy. Also awful tedious to test a thousand chips one at a time, but still pretty darn handy.
When I resolder a replacement, I clean off the existing solder with a good wick, repaste and then hot-air the chip. If it doesn't align quite right, poking it with the big tweezers will usually bounce it into place.
|
|
|
If belly pad contacts don't get into the VDD corners, I guess a 56QFN socket could work.
I never messed with the BM1382 but that package looks like it sucked quite a bit. BM1384 is fairly straightforward and BM1385 appears to be a stock footprint. Anything with perimeter pins like QFN is going to be a lot easier to work with than BGA or a bump die package because you can more easily verify alignment by looking at the solder rather than just hoping it was right. Belly pad floatation also helps a bit. I haven't worked directly with much BGA so I don't know how well those self-align, but they tend to be larger packages as well so the weight could prevent some of that movement.
|
|
|
Heatsink isn't part of the base cost. Someone wants 50 boards with 50 heatsinks, that someone will also be paying for 50 heatsinks. One S5 board only gets you the 8-chip PCB and nothing else. You provide your own cooler, power, computer running cgminer et cetera.
I do not have a socket to test the chips standalone. Since it's, as far as I know, a custom footprint (I don't know any other QFN with corner pads like this one) I'm not sure where to even look for a socket.
My assumption is at least 24 viable chips per 30-chip board. More than that is a bonus.
|
|
|
There have been a lot of people mentioning they can't figure out how to set the clock and whatever they type in, it goes to default instead. The new version where you just type in the actual MHz should be pretty foolproof and awful dang handy.
|
|
|
If someone corners the market by hoarding pods from a bucket of trade-ins, within a few weeks I'll have turned those trade-ins into between two and three times as many pods.
Also, the only people that should profit from this venture are not GekkoScience, but GekkoScience and its customers. If we're the only ones coming out ahead, that means the product is not worth buying.
|
|
|
There's a "what's the best stickminer" thread on Hardware, and somebody mentioned he might have some fun news regarding a scrypt stickminer in the near future. That's the only public posting I've seen but I also haven't really been looking. Guy might have his own thread in an altcoin board. I don't know.
|
|
|
As many people as are popping up with boards, I'll probably look to making more like 200 than 50.
I guess what I was thinking for a cap was limiting the number of free pods that could be gotten for S5 trades; say, past the cap an S5 board would pay for half a pod instead? That does seem kinda crappy though. I think I'll leave it wide open.
Regarding "extra pods" if folks don't buy them all up, that's not really a problem. Heck I'll just mine on them myself.
|
|
|
So, who thinks I should put a cap on how many S5 boards a person can trade for pods? What should that cap be? Five? Ten? I know right now only 8 survey respondents have selected the straight trade option, but I'm also told that the market value for dead S5 boards has been increasing since this thread appeared because people are buying in anticipation of the opportunity to trade.
Should I worry about a cap, in order to prevent some folks from being taken advantage of by a few opportunistic individuals? Or should I not care and let the market fend for itself? I don't like being party to someone getting screwed, but at the same time it's not my job to be the dick police.
Also, y'all will be pleased to note that I made absolutely no progress on this or any other project today.
|
|
|
The last post in this thread was almost two years ago. Since then, ASICMiner has gone quite belly-up and pretty much stopped doing business entirely about eight months ago.
So it's pretty unlikely anyone will be distributing their products anywhere.
|
|
|
There are also test pads on the back if you don't want to risk shorting something out.
Also, rumor has it someone (not me) is working on a scrypt stickminer that should be interesting.
|
|
|
Dang. That is a sweet stickminer.
|
|
|
No I haven't, for the same reasons I've been saying since about July that I didn't want to work with their chip - I don't want to mess with BGA, or scrypt, or 50W ASICs.
|
|
|
The Compac has been tested to work (enumerating as a U3) on an R1, if I'm remembering right. A Compac2 with 1385 would require specific software so not backward-compatible. I didn't think about trying that angle with Bitmain though.
I am more optimistic about Innosilicon selling chips, but since they thus far have no interest in small quantities or samples it seems getting anything even for testing is months away which puts production more months away.
Also, all this should probably be discussed in a dev thread rather than threadjacking this guy's question.
|
|
|
|