Regarding a voltage underrun, the VRMs on Cube Cards are actually fairly clever. There are two VRMs per card, each sourcing voltage to 8 chips. The regulator uses three resistors and a FET to select the high and low voltages. In one configuration, all three resistors are in the series divider for feedback sampling; in high-speed configuration, the FET shorts around one resistor which changes the division ratio for the sampled output voltage and shifts it up from 1.05V to 1.15V
(More information on Cube VRMs here http://www.gekkoscience.com/webuilds/cube_oc/cube_oc.html

What I would recommend is, with the card running, measure the outputs from the VRMs (across C99-103 and across C83-87) and make sure you have about 1.05V; then switch it to high clock and see if both VRMs changed to 1.15V

If one is still reading 1.05V, measure the voltage present on R25 and R26 and see what they read. If they're both reading about the same (several volts), I'd say the FET itself is bad or not properly connected. First I'd resolder the FET (Q3/Q4) and see if that fixed it.
If one is several volts and one is low, you're not getting the signal to turn on the FET and switch the VRM to high-voltage mode. I'm not sure offhand which logic driver controls this signal but it shouldn't be too hard to trace. In the case of no signal or a bad FET, a workaround should be to remove the R9 or R18 resistor (on mine, it was 1800ohm but I've seen some that had 2100 there) and short across its pads. This would force the VRM to run in high-power mode all the time, so it would be providing more power than is required for "low clock" but would be right for "high clock".

Currently gathering resources for a first batch of these boards. The bulk are already spoken for, but I should have a few dozen available in about two weeks. If folks are interested in making cases for these, let me know and we'll work something out for getting samples/small orders to you.

The current board is about the same width as the Z750P power supply, so they can be put side-by-side. The output pin header includes lines for 3.3V standby power, an integrated 5V 3A supply (powered off 12V), current-sense signal, external fan speed input, auto-on input, power-good and current-share outputs.

I'm also trying to make available 14AWG/24" PCIe 6-pin cables with spade terminals on one end for mounting on the screw terminals.

Working on dev for a controller board that will integrate power supply health and per-supply/total power output, as well as thermostatic fan speed control and maybe some other bells and whistles. Also working on boards for the DPS800 interface.

Isn't this for AM BE100 chips? That datasheet you linked looks quite a bit like something else.

Ah, well nifty. Carry on sir. Not sure what time I'll have to assist, but I could probably (eventually) assist with resources and/or assembly.

I'm still wondering if anyone has the actual interface data and whatnot required to drive the chips. Board design can't happen until you know the actual logic and signalling required; makes no sense to start at step 2.

Not sure if anyone really cares, but it's possible. I actually worked one up about a month ago, but they need cooling improved and since they overclock natively you can really only get another 5-6GH out of them. I've had two people ask me about Cube overclocking in the last three days so I figured there might actually be demand for the information. Now that I think about it though, of course there was demand for the information. I'm sorry I didn't post this before Thanksgiving.

http://www.gekkoscience.com/webuilds/cube_oc/cube_oc.html

If I'm thinking right, the fans in those are self-regulating based on internal temperature. Anyone know a more complete pinout, like the power-good signal and current share? I'll have a couple here in a few days to play with so it'll be no problem figuring out, just wonder if anyone's seen a more complete pinout listed somewhere I haven't yet looked.

Which supply is that? Is that for the DL360?

Depending on funding availability and whatnot, I might have some ready by the end of next week or shortly after. I'll definitely keep the thread posted.

That should be enough to get it turned on and spitting out 12V.

Why, what's March? There's people that want these yesterday.

Still having trouble finding a reliable source for the Z750P power supply socket, that isn't about 6 weeks out.

Working on pinout/interface for the DPS-800 HP power supply. I've gotten information enough to turn them on, but haven't turned up an actual pinout of the signal lines yet but we'll get there.


Is there another people want boards for? I'm seeing a lot of reference to the DPS-2000, so that's a possibility so long as people understand they're gonna need to rig up fans. Suppose I could put a fan header on the board.



As of now we're including a 5V 3A output, which was requested as some standalone miner devices need a 5V line to power internal controllers. It'll break out the 3.3V standby, give option for external fan control, and supply an output current measurement signal in addition to auto-poweron, power-good detect, and current balancing. Hopefully I can locate/derive enough info on the DPS-800 and DPS-2000 to implement the same features.

The DL360 is next on my list to build boards for. Hopefully I have some better info on those sometime next week.

In the next few days I'm gonna run up some test procedures for the Z750P boards, which should make for an interesting writeup. With any luck we'll start mass production by the end of next week.

Detects my U1 and my USB BE. Well, it knows they're plugged in but doesn't know what they are (Failed to open, err -12). CGMiner 3.1.1 doesn't care, just uses the BE anyway because I tell it to explicitly. Didn't have much luck with zadig when I was messing with it a few  months ago, don't want to mess with it again.

So whatever happened to the -S option in cgminer? Is it possible to run these without the zadig driver? USB BE would work just fine without it, back in the day when you could manually specify which devices to use. Is it still possible to manually specify which devices to use, instead of assuming cgminer knows what you want?

Definitely test the load-balancing on those if you can.

Good call, I was thinking about saying the same thing.

Yeah the best I can do right now is a 51761 modified, that's what the current prototype is using. The standard spacing for those is 0.25" on the power sockets, so I gotta basically hollow out all the pastic and mount the blade sockets to the PCB on 0.20" without any support.



Correct connector for which?

From FCI I've gotten a fairly stamped response "I'm talking to a product engineer, hopefully I can update you before Dec 31st". Molex says they don't see anything in distributor inventory and tech support might be able to help me find a compatible replacement.

Anything with a s24p6 connector on a 0.1/0.2 inch pitch, signal-compatible. I'm testing with a Z750P so anything pin/signal compatible with that should work just fine.