Wish I had better access to it so I could put an IR meter to use.  I have some stick-on heatsinks left over from USB BEs.

Interesting.  It was around mid-90's.  The top is off and I have two additional small fans pointed at the voltage modules.  For the heck of it I placed a magazine inside taking the air pushed by the in fans and forced it down on the BBB and controller board.  The realtime hashrate went up to 109.  I removed the magazine and it dropped to the 70's and was falling when I put the magazine back.

KnC 0:                | 108.8G/95.68Gh/s | A:127753 R:283 HW:55150 WU:1436.3/m

Just lowered it again at 5PM.  I use firefox but needed to go to IE for BertMod.  Just tried Chrome.  Suddenly I got an advertisement over my computer speakers for a phone service and there's nothing displayed!?

Anyway, with the lowering of an additional 200 millivolts there is no corresponding slew of re-enabling cores.  .....wait, spoke too soon, 5 cores just enabled.

Lowered the input voltage from 12.2v to 11.9v.  I note that I have a number of digital voltmeters but they do not agree entirely.  It had been set for 12.05v with one, then I added an inexpensive 2 wire digital display so I could tell if ever there were a voltage down issue.  The start before adjust could have been 12.2 or 12.05, but I adjusted it down from 12.2 to 11.9.  Many cores were re-enabled, then some disabled again shortly.  Now after some 10  minutes or so my total power usage as shown by BertMod is up 6 watts to 100 from 94 watts, my total cores are better on 3 of the 4 dies but worse on 1, hashrate up to 96.64GH/s avg reported by cgminer.   Must try this lowering voltage thing again.

100GH/s Mercury.  Tried to save on power supplies, bought two 360watt 12vdc supplies and a BBB 5 volt wall wart then added a 4 pin Molex and brought in the 12vdc for the intake fans, which I found are powered by the 4pin Molex on the controller board and that 12 vdc isn't common with the PCIE 12vdc.

Fired it up this morning and got lots of cores disabling.  Thought perhaps if I add the second 12v supply in parallel, did that, same errors.  Shut it down, went back to a single 12v supply and added two tantalum capacitors.

It's back up again now about an hour.  Seems to be settling in between 75-78GH/s.

http://eligius.st/~wizkid057/newstats/userstats.php/18XUsmKnuPmGdNaPrtpwSr11iAXFUdFUys

Okay, I'll need to spring for a decent power supply.  If it turns out the hardware errors an disabled cores don't improve, I'll go back to this supply and put the new in my main machine who's supply fan sounds just a little loose as the home cools with the season.

I have yet to try the 0.96 firmware.  Also, I have the where with all to adjust the 12V in so I might try lowering that reducing the power and see if any cores enable.

I like USPS, after that I'll go UPS, and have tried to avoid DHL like the plague.  A previous delivery went bad some time back.  That had been flown to Columbus, GA, then driven up here once, a note left, then subsequently there was a report the driver had returned 2 additional times.  After he didn't show the second I set up a video camera covering  the driveway.  He never showed and reported he had.  I won the argument with customer service but no way would I have chosen DHL ever again.

Then yesterday I find they couldn't deliver to my PO box.  Not a good experience with Customer Service, not antisocial discourse but just that they couldn't deliver to a PO box and their delivery window as all day.  

I expected the DHL driver today sometime between 8a and 9p.  He called at 8:45AM and said he wasn't far away.  He delivered at 9:15AM.  Tall black guy he looks like he could play for the Knicks.  This isn't a neighborhood he'd feel comfortable in, hell I don't and I'm Scotch/Irish/German/Dutch, well less uncomfortable after 20 years here off and on.  The box was in perfect shape and although the fan had come loose I understand that almost always happens.

Not lagging by much, 261GH/s, 237GH/s, 275GH/s.  You're using 0.96?  I don't have those cgminer stats reported with 0.95.

My Mercury hasn't hit 100GH/s but seems to have settled in around mid-90's according to cgminer.  Likely not the ASIC module fault as I tried going with a cheap 360 watt switching supply of a kind I'm having good success with powering a blade.  I went to 2 of the supplies in parallel even tho the  two supplies are only drawing 143 watts at the wall and bertmod reports DC/DC power 94watts.  I'll go back to one 360 watt supply and add some tantalum capacitors then after letting it run for a while I'll upgrade to 0.96.  Although I've heard that if one's system isn't doing well on 0.95, going to 0.96 isn't a help.  More than a third hardware errors.

 KnC 0:                | 104.0G/94.74Gh/s | A:91699 R:323 HW:33892 WU:1412.1/m

ASIC Board Info
4 Temperature sensor: 34.0 C

Die ID Cores ON Cores OFF %
0 44 4 91.7
1 39 9 81.2
2 35 13 72.9
3 38 10 79.2
DC/DC ID ON/OFF Status Input Voltage Output Voltage Output Current
0 ON OK 11.7 V 0.747 V 33.6 A (25.1 W)
1 No DC/DC detected
2 ON OK 11.9 V 0.748 V 31.7 A (23.7 W)
3 No DC/DC detected
4 ON OK 11.7 V 0.741 V 30.1 A (22.3 W)
5 No DC/DC detected
6 No DC/DC detected
7 ON OK 11.8 V 0.738 V 32.2 A (23.8 W)
 


If mine were running as well as your slowest, it would be 118.5GH/s.

Someone else mentioned this, I'll repeat, it may be DHL has less a problem moving it along if you call and talk to a customer service representative.

Just saw your post.

Well, after 36 minutes I have this:

KnC 0:                | 92.74G/90.61Gh/s | A:34415 R:179 HW:14505 WU:1378.3/m

Hardware errors 14505 while accepted 34415.

Shutting down.

Firmware 0.95 and no I have not upgraded.  Temp in the room is cool, 70.2°F (think that's 21.2°C) and the miner is
 HW status:

ASIC slot #1: -
ASIC slot #2: -
ASIC slot #3: -
ASIC slot #4: -
ASIC slot #5: 30.5 ℃
ASIC slot #6: -

About the same.  Nothing visibly wrong on the ASIC board.  Didn't check the controller board or BBB.  Soon after the restart the 12v switching supply said the ASIC and fans were using 156 watts but as the cores shut down the power dropped and presently it's just 132 watts.

 KnC 0:                | 95.96G/85.33Gh/s | A:10513 R:26 HW:4586 WU:1333.5/m

Oh well, I'm going to shut it down and inspect the board with a loupe and check out its thermal paste or tape.

KnC 0:                | 71.37G/231.7Kh/s | A:17372 R:4 HW:8355 WU:   0.0/m

bummer

eligius is showing 2013/10/18 14:48:45: 675 seconds: 34741.51

34.7GH/s

spoke too soon as the cores are disabling again

[2000-01-01 00:21:05] KnC: core 4-82 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:06] Accepted 0515e1e9 Diff 50/31 KnC 0 pool 1
 [2000-01-01 00:21:06] KnC: core 4-118 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:09] Accepted 05be9c55 Diff 44/31 KnC 0 pool 1
 [2000-01-01 00:21:09] KnC: core 4-52 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:09] KnC: core 4-105 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:11] Accepted 07cf965b Diff 32/31 KnC 0 pool 1
 [2000-01-01 00:21:12] KnC: core 4-69 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:13] Accepted 0093f6d2 Diff 442/31 KnC 0 pool 1
 [2000-01-01 00:21:13] KnC: core 4-121 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:14] Accepted 02d2db15 Diff 90/31 KnC 0 pool 1
 [2000-01-01 00:21:16] KnC: core 4-10 was enabled back from disabled state
 [2000-01-01 00:21:17] Accepted 0178e748 Diff 173/31 KnC 0 pool 1
 [2000-01-01 00:21:17] Accepted 03679585 Diff 75/31 KnC 0 pool 1
 [2000-01-01 00:21:17] KnC: core 4-16 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:19] KnC: core 4-88 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:19] Accepted 069d0bbd Diff 38/31 KnC 0 pool 1
 [2000-01-01 00:21:19] KnC: core 4-79 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:23] Accepted 05a1604d Diff 45/31 KnC 0 pool 1
 [2000-01-01 00:21:23] KnC: core 4-96 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:23] KnC: core 4-92 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:23] KnC: core 4-142 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:25] Accepted 07baca69 Diff 33/31 KnC 0 pool 1
 [2000-01-01 00:21:29] Accepted 04ee09c4 Diff 51/31 KnC 0 pool 1
 [2000-01-01 00:21:30] Accepted 0210f426 Diff 123/31 KnC 0 pool 1
 [2000-01-01 00:21:30] KnC: core 4-131 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:31] KnC: core 4-135 was disabled due to 10 HW errors in a row
 [2000-01-01 00:21:31] KnC: core 4-138 was disab

Great news, it's re-enabling the cores while on eligius:
 [P]ool management ettings [D]isplay options [Q]uit
 KnC 0:                | 100.1G/91.20Gh/s | A:16070 R:4 HW:5982 WU:1364.5/m
--------------------------------------------------------------------------------

 [2000-01-01 00:19:42] Accepted 07f66382 Diff 32/31 KnC 0 pool 1
 [2000-01-01 00:19:43] KnC: core 4-128 was enabled back from disabled state
 [2000-01-01 00:19:44] Accepted 03f5e7d0 Diff 64/31 KnC 0 pool 1
 [2000-01-01 00:19:44] Accepted 00ae90b6 Diff 375/31 KnC 0 pool 1
 [2000-01-01 00:19:46] Accepted 07e82ec4 Diff 32/31 KnC 0 pool 1
 [2000-01-01 00:19:48] Accepted 0668ea61 Diff 39/31 KnC 0 pool 1
 [2000-01-01 00:19:48] Accepted 0447df3a Diff 59/31 KnC 0 pool 1
 [2000-01-01 00:19:50] Accepted 05a262fc Diff 45/31 KnC 0 pool 1
 [2000-01-01 00:19:52] Accepted 0068831d Diff 627/31 KnC 0 pool 1
 [2000-01-01 00:19:52] Accepted 045a118d Diff 58/31 KnC 0 pool 1
 [2000-01-01 00:19:55] Accepted 07efbecb Diff 32/31 KnC 0 pool 1
 [2000-01-01 00:19:55] KnC: core 4-57 was enabled back from disabled state
 [2000-01-01 00:19:59] KnC: core 4-75 was enabled back from disabled state
 [2000-01-01 00:19:59] Accepted 01ee87f8 Diff 132/31 KnC 0 pool 1
 [2000-01-01 00:20:00] KnC: core 4-139 was enabled back from disabled state
 [2000-01-01 00:20:01] Accepted 011fabea Diff 227/31 KnC 0 pool 1
 [2000-01-01 00:20:02] KnC: core 4-89 was enabled back from disabled state
 [2000-01-01 00:20:02] KnC: core 4-49 was enabled back from disabled state
 [2000-01-01 00:20:03] Accepted 059cf6ac Diff 45/31 KnC 0 pool 1
 [2000-01-01 00:20:05] KnC: core 4-5 was enabled back from disabled state
 [2000-01-01 00:20:05] KnC: core 4-64 was enabled back from disabled state
 [2000-01-01 00:20:07] KnC: core 4-77 was enabled back from disabled state
 [2000-01-01 00:20:07] Accepted 07315ea2 Diff 35/31 KnC 0 pool 1
 [2000-01-01 00:20:08] KnC: core 4-50 was enabled back from disabled state
 [2000-01-01 00:20:08] KnC: core 4-85 was enabled back from disabled state
 [2000-01-01 00:20:08] KnC: core 4-4 was enabled back from disabled state
 [2000-01-01 00:20:08] KnC: core 4-121 was enabled back from disabled state
 [2000-01-01 00:20:09] KnC: core 4-107 was enabled back from disabled state
 [2000-01-01 00:20:09] KnC: core 4-117 was enabled back from disabled state

Yes.  Arrived this morning.  Had no testing stats to show on Slush because the password was incorrect.  Now it's here and many cores disabled while getting warmed up.  It's running 0.95.

Changing to eligius.

KnC 0:                | 100.6G/79.68Gh/s | A:6026 R:3 HW:3364 WU:1338.5/m