Yet there would be signs the tape had been removed and replaced. There was quite a bit of additional USPS Priority Mail tape but there was no signs of repackaging that jumped out at me. What I find curious is the Bitmain box dimensions difference. Was about to measure the boxes and looked at the second first, this time more closely, and the bottom had been opened and resealed.
#1 room temp 79°, 442.64GH/s(avg) HW 151, 15hrs 49min
#2 room temp 83°, 428.83GH/s(avg) HW 39, 245 (it just made a liar of me), 15hrs 53 min.
not saying it happened, but you would be amazed how easily tape comes off on something that sits in a freezer, then reattaches when warmed.
but either way bitmain mines these before selling them. I think they actually mine the whole stock load as they are building out the next batch. like build batch 7 mine them all and not sell them till batch 8 is ready, mine batch 8's till batch 9 is ready to replace them.
As in an earlier post, I thought the unit probably came up slow in testing and was sent to troubleshooting where new paste was applied and the screws tightened grossly tight on the bad hashing board thus bringing it up to "passing". That would have been at the factory. In industry we don't trash a new unit just because it isn't up to snuff. As a matter of fact that troubleshooting of new failed units was one of my responsibilities at the last place I worked. Of the two boards on that unit, the good hashing board had thermal paste that was less new, more dried, than on the questionable board. That it may have been previously sold by the reseller and returned then shipped again is a possibility given what I see now as the condition of the plastic tape. When my two S3+'s arrive I'll know if the box comes with clear tape or just stapled. The USPO sometimes puts clear tape if a box isn't sealed with tape. This has priority mail tape but the plastic tape under it had been opened and not by me.
A manufacturing facility needs to run new units as there's what's called an infant mortality rate. Odd term to use in electronics but it's true. Testing doesn't always run long enough to test but the failure rate drops sharply.
I just finished ripping #2 apart again and have just fired it up. Both are now running in the cooler room side by side same height from the floor, same distance from the wall, both being hit by air from a 20" fan. I've /etc/init.d/cgminer restart on both units.
The reason I ripped #2 down again is that the grossly overtightened screws had 2 that had tops partly stripped. I had put those on the lower corners when putting the heatsink back on and couldn't tighten very well. I later noticed that the controller board used the same screws. I just swapped those two lower corner screws for two from the controller board. At the same time I swapped the two fans taking care to make sure the air flow direction is correct.
#1 50min14sec, 443.59GH/s, Hw 6
#2 49min52sec 439.11GH/s, HW 0
------------
#1 2hr6min 445.00 GH/s (avg) HW 18
#2 2hr5min 439.97 GH/s (avg) HW 0
