Coin's way up, so let's say 0.57 right now.

That works for me, if you two want to iron out the details. I was thinking of using the PSU as partial shipping cost, but I don't really need another one. If Phil actually wants it that works out for everyone.

The current draw should be on the order of milliamps, and no more than 5VDC. Pretty much any general purpose BJT or FET should work. That's exactly how the external turnon circuit on my server PSU interface boards works.

Some of the transactions I sent out yesterday hadn't even shown up on Blockchain.info inside an hour, let alone confirmed. Ridiculous.

Also, looks like I got an order through finally. So Round 1 oughtta be getting here in a couple days.

When they realized that their miners weren't hitting 14TH reliable they probably did a bit of testing, and Batch 3 are Batch 1 miners binned for lower speeds.

I have at least one wallet client open at all times. Additionally, and this is my formal opinion - daytraders and altcoiners can piss off forever.

Seriously though, whatever happened to upping the block size limit? Wasn't that a massive stink for months?

No, I didn't want to mess with things that much since it was a loaner. If I had one of my own It'd have been completely torn down inside an hour. I didn't do anything that required applying power or actual disassembly.

The Minion was BlackArrow. Hashfast's big chip had four independent dies on one BGA base; I know they had separate power rails but I don't know about separate grounds or you probably actually could have strung them.

I've got some 40A 208V breakers powering 5x S7 off 2880s and dual Dell 750s in hosting.

I just wish the jackass day traders would save some transaction headroom for people actually wanting to use bitcoins as a means of exchange for goods and services. You know, like a currency.

Also, eff this. I'm going out to the shop and place the order a THIRD time and if I have to hang out there until lunchtime tomorrow I'll get the transaction confirmed. This is retarded. Y'all are waiting for something and you're gonna get it.


EDIT - coin being where it is, as of right now the price is back down to 0.59BTC per unit. We'll see if coin stays high.

Well apparently the network hates me. I scrapped the original order, put in a new transaction with a full order magnitude higher fee and it's still unconfirmed five hours later. So I guess that order's timed out and cancelled. I'll have to hit it again when I get to the shop in the morning. Maybe I'll step the fee up to about ten bucks this time, that should get someone's attention. If that doesn't work, we'll go freakin' fifty.

I've been out since about 4PM so I haven't checked on prices. Last I saw coin, it was still below where I priced things out initially. Guess it's gone up since then.

Power density becomes a big problem with large dies and requires complex or exotic cooling systems. High current density is going to kill peripheral parts count and regulation efficiency.

I've got an incoming transaction sent on June 6th that still hasn't confirmed, and an order with Bitmain from yesterday still pending. Blocks have been very full lately. Whatever happened to upping the cap?

The one here is in a 90F ambient. PCB temps currently 66,68,69 chip temps 97,100,98 fans 4320 and5280 at 650MHz 13.94TH after one hour.

It's got some pretty decent ADCs on there. They might have it calibrate a dynamic voltage setpoint instead of just pushing a hex value and calling it good. That's what I'd do. With that capability you could have a self-optimizing routine kinda like the Avalon4 had.

So they made sure to dual-FET both high and low sides of the buck switching, which is good considering the high currents they're seeing. The FET section is also heatsinked, I like to see that. I was a bit worried when taking the clock and core up high on my S7LN for "overclock" efficiency tests because I didn't want the transistors to smoke out.

Looks like they upgraded from PIC12F1572 to PIC16F1704 for the regulator control. Still has about a 3600-opcode program on it, probably pretty similar to the stock code in the S7 but this chip has built-in I2C so that actually makes some things a lot easier.

...or, at least I would have placed an order, if that transaction would ever confirm. This has been a bad month for me and transactions. I've got a June 1st invoice that's been pushed twice and hasn't confirmed yet, and one of the guys for this GB has a transaction in limbo somewhere. Kinda freakin' me out.

EDIT - order has been resubmitted with a fresh transaction on much higher fees. Hopefully it gets confirmed this time and we can get to business.
I've got my miner decently characterized between 720mV and 660mV and will continue the numbers tomorrow.

Probably that, or core voltage as too low. Might have had some under-spec chips on those boards.

I'm pretty sure the controller doesn't get power from the hashboards. The peripheral circuits of the hashboards (like the PIC and DPOT that set the string voltage) actually draw power from the controller, not the other way around. Worst case with a PSU common to hashboards and controller is they get powered at the same time.