Probably not. I'd rather put effort into making a nice open standard design than put effort into rebuilding within someone else's lame mechanical constraints. The idea has previously been considered and rejected.

Naw, Gridseeds are a different breed of cat. Not much on 'em or in 'em that would benefit my design more than about a dollar's savings.

The options lined out in the poll above are for the things that I guarantee I can work with, so I'll probably stick with that list for now.

That is darn sexy, sir.

Yes the return address is correct.

Also, with the V0.5 there is no longer a "danger zone" on the pot. If you circle past the top-end voltage it automatically drops back down to 550mV. All 360° of rotation are within the safe operating voltage of the ASIC.

Actually, all you have to do is grab a screwdriver, pull the hashboard off the heatsink, stuff it in a padded envelope, and mail it to me as cheap and slow as you want. Looks like a pair of S5 boards in a padded envelope would ship first-class within the US for less than $5. Shipping for the miner itself would need to be paid, but for most US customers a single miner, possibly even with a cooler, would be about $10.

Seriously, I don't really care what condition the boards arrive in so long as they're not shattered. I'm pulling parts. I don't want heatsinks or fans or controllers or anything else that make shipping difficult or costly.

Something like the original-version BTCGarden BE200 blades are probably too long to fit comfortably in an envelope, but I have no problem with one being chopped in half and sent as two pieces provided the power components are intact which gives you about a two-inch window of "right in the middle" to hack away. S2 boards would actually be the opposite - all their power parts are in that middle space, so you could chop off about four inches off both ends and it'd still be acceptable. Though I could probably find a use for all those 470uF caps if boards were delivered intact.

toptek - Maybe I could do a raffle. Maybe 0.01-0.02BTC gets a "ticket" and the drawn winner gets a free pod. Two boards sent in gets you two pods, so unless you want one of them I can do two drawings. Maybe I'll do a third-place winner gets a free Compac. Something like that will have to wait until I've got at least prototypes built, so there's a while to figure out the details.

Keep the S1/S3 heatsinks and fans. The goal we've been working toward for the last six months is being able to build new hashboards for those old chassis. I'm still working toward that goal.

I'll have to check S3 boards for parts I'm looking for, but since most of those probably work and you can get $70 for a complete S3 I didn't figure folks would be willing to trade.

I did a half-board power measure (four VRMs disabled, four with adjustable voltage) during initial testing and the board-level power consumption dropped to a hair below 0.8W/GH board-level according to my measurements, compared to about 1W/GH unmodified. With my power that's still viable for a while yet, especially since the shop leaks air like a seive and it still got balls cold with 40KW of miners running last winter so I need the heat.

Those TPS were definitely stressed. Every other BE200 design I've seen had two ASICs per, but they were also clocked higher. I have had to replace a few blown schfifty-threes. One of my test Tubes is down a node from a roasted buck chip, and I think Novak's space heater Tube is short one as well.

If I have time coming up I might try a modified cgminer build and see what happens. For now I'm satisfied with their performance. Anything I was going to do with Prismas is sidelined for a bit for more pressing projects.

That's great, but it's the opposite of what I need. I have 18-pin controllers and a 16-pin board. If I have time today I'll just do the probing and figure it out.

I was running off the standalone controller. Currently I have 'em set at 180MHz and am getting expected hashrate. Chips still didn't like to light up below 730.

When I have some time, maybe in a few weeks, I'll take a look again. What version of cgminer are you pulling from? Nothing I had handy reported HW errors correctly so was kinda useless for testing.

Oh in case anyone's wondering, about half the hashrate currently mining on the Burger (http://eligius.st/~wizkid057/newstats/userstats.php/1BURGERAXHH6Yi6LRybRJK7ybEm5m5HwTr) is my proof-of-concept. About 115GH is seven test Compacs Novak set up running at 300MHz and another 88GH is my 8-chip setup at 200MHz. Like I said it's not a proper build, it's all hacked up but the main circuitry is pretty much the same as what my board'll have all integrated.

Right now ASICs are in tighter supply than power parts (I have power parts for about four times more boards than I have ASICs for, especially given it's an 8:1 ratio chips to VRMs) so I'm hoping some dead S5 people will get in on it.

I think the pod line will carry the Amita name. Stickminers will be versions of Compac, and pods would be versions of Amita.

I'll start working on PCB layout tomorrow. Novak's out for a couple days so we won't have any comm/control prototyping for a bit and it'd probably be two weeks at least before I had any sort of prototype PCB in hand. Mass production distribution would be, best case, about Thanksgiving. No need to send boards just yet.

It's times the entire range. Why would I build something you couldn't adjust the voltage and frequency on? That's no fun. Since it's 8 chips, it's 8 times the hashrate of a Compac at any frequency (and probably slightly better efficiency). 100MHz gets you 44GH, and you could take the voltage down to where the whole thing probably burns under 15W of DC power; 400MHz gets you 176GH at around 90W. If you had a good cooler you could probably push it to 450MHz and see just shy of 200GH but the efficiency would be probably pushing 0.55-0.6W/GH at that point.

Adjustability was noted in the first paragraph of the first post, in the sentence right after "which puts it at 8 times the hashrate".

Since this thread is already here, I might as well use it. Today I resurrected a dead S5 board but it's the early version with 16-pin cable so I had to borrow a controller to test it. All the controllers I have running around use the 18-pin. I have the 18-pin signal pinout already figured out but I haven't done anything with a 16-pin before today. Anyone know offhand what difference there is between the two?

Once I have a better estimate of parts (which will of course require more of a finished design than I currently have) the price may go down. Or it may include a cooler, in which case if you don't want a cooler the price may go down.

Also, toptek, your dead S5 board donation partially inspired this idea so you'd be getting one anyway.

I'm still working on an official source for BM1385 and other current-gen chips. I'd rather not tear apart a $600 board that still works.

I also talked to Novak about the miner-for-trade thing. We looked over some basic design ideas, approximate pricing and I put up a poll (over herefor opinions and such. I took the afternoon to hack up a bit of a proof-of-concept using mostly S5 and BTCGarden parts. Tomorrow I'll clean it up a bit, try to get some efficiency estimates and start working on a proper PCB layout.

I'm also potentially changing my mind (again) about what I want to build TypeZero power around, so I need to take a day and look at some fresh parts. While waiting for prototype parts for the pod, I'll probably rip into prototyping TypeZero power. The pod will integrate all the controls and a lot of the software so we'll be a few steps ahead of the game already.

We're gathering funds to start the second round of Compac manufacture, which should go more smoothly than the first half. All the long-lead parts have already been acquired; we just need to get some of the various components in quantity and reload the robot. I'll probably reopen the sales queue (with an estimated 14-day lead time for the first orders) Friday or so.

Seems unlikely there'd be much of use from an SP20.

Competing against the U3, yeah that's a bit of a problem. Of course, the AntRouter kinda sucks (at least until someone hacks the firmware) and so does the U3 (with its poor efficiency and known stability issues) so I figure someone might want a better option anyway, especially if it can be had for cheap or free depending on what's in the bin.

We've got enough parts running around to build a small batch just for ourselves if nothing else, but I figure if anyone else might want to get in on it I certainly am not opposed.

I need to talk to the proper owners before making any arrangements, but I have several cases of new-in-box Freezer 7 Pro (or whatever) coolers gathering dust on my shelf for over a year because some hosting customers got screwed by Technobit and never received their 50-odd HEX4M boards. Might could get a deal on those which would help people out and they'd certainly do well cooling these chips even at top-end.

Two S1 boards would get you to half price. 400MHz on the pod would get you the same hashrate as an S1 for a quarter the power. Just throwing that out there. And then if we can ever get the resources together for TypeZero boards with new chips, you'd already have a bare S1 chassis to strap 'em to.

I won't be making them with any fans. It's a board to mount a CPU cooler on, so whatever cooler you have or find or what I can find.

So, Novak and I are working on a proof-of-concept device which we think has potential. It'd be a U3-style "Pod" miner sporting 8x BM1384, which puts it at 8 times the hashrate of a Compac or approximately one eighth of an S5. The pod would have fully adjustable (in software) voltage and frequency, meaning it could be cranked up to full power or undervolted to a quiet 0.28W/GH (approximately). Power comes in through either a 12V brick or a 6-pin PCIe (both jacks would be standard). USB connection to a controller, just like any other pod. Instead of a custom heatsink there'd be screw holes to mount a standard CPU cooler, which I could probably source cheaply if you didn't already have one laying around. Cooler fan speed would be automatic and thermostatic.

Here's the thing. I talked to some of my cronies a few weeks ago about an idea like this and, at the estimated cost of materials and such, I was told it wouldn't sell. Not efficient enough, $/GH wasn't there, stuff like that.

So, what if we recycle?

We're cooking up a design that actually reuses some common components from previous-generation hashing boards. We've got boxes of this crap laying around, and the parts are still good, so why not? They're basically free, just take a bit of work to pull and clean up. And in order to get more parts in, we could trade miners for last year's boards that contain the parts.

The most important components, the hashing chips themselves, really can only come from S5 or S5+ boards. I know a lot of people probably have failed S5 boards, so in order to get some of that hashrate back would you be willing to trade a dead S5 board for an 8-chip pod?

Some additional components can be found in common power circuits like on BTCGarden AMV1/AMV2 blades, ASICMiner Tubes and to a lesser extent on the boards common to Rockminer BE200 gear, as well as AntMiner S1 and S2 boards. As the power parts don't represent near as much expense as the hashing chips would, a quota of these parts would cover half the cost of a miner, so you'd get one miner for your old boards and $25.

I'd only need the boards. Heatsinks and whatever else can be recycled or reused locally. Honestly, shipping speed doesn't matter to me either (though your miner might not go out until I have your trade boards in hand) so that can be pretty cheap.

And if you don't have boards to trade, or we have so many boards already that I have to suspend trading and sell for cash only, the price is still going to be substantially lower than it would have been otherwise.

The miners would, obviously, be built from used parts. But as with everything else we've ever built and sold, they'd be tested before shipping and go with a full warranty. If it breaks and not because you played football with it or tried to push it to 400MHz with one of those baby i3 coolers, it's probably covered by warranty.

Once I've pulled the parts I need, all the excess board will be taken in for proper recycling/reclamation rather than hauled to a landfill.

So, show of hands. How many people would get in on a program like this? I'll be honest, I just thought of it a couple days ago and we just talked it over today and I just 20 minutes ago got some semblance of a prototype approximately working with hacked-up chunks of other miners, so it's very much a plan in progress. It won't go very far without community support, so, community, how much do you support this?

I've got a two-chip test board that I can wire up to either an S5 controller or to a USB adapter. Since we don't have multi-chip code yet (soon though) it was fairly irrelevant to run it off the USB adapter but irrelevant is not impossible. Shouldn't actually be that difficult to run an S5 board then, I think, provided you had a working cgminer driver.

That cgminer stops actually is the problem. It can happen that the chips continue to do work even if cgminer stops, which since I saw it happen once on a Compac I think it's more a chip thing than cgminer. But cgminer I believe controls the fan speed, so when cgminer drops out so does the fan. What you need is something independent of cgminer to wrangle the fan based on device temperature.

Also I'm not sure how many people are still worried about it four months after the last post in this thread.

Any chip-making I get in on will probably be by cooperating with what PlanetCrypto wants to do. Right now I'm having enough fun raising enough capital in non-bastardly ways to get smaller projects off the ground without having to worry about the seven-figure price tag of an ASIC.

I do hope to continue making good miners. Thanks.