First off, The few times I've had to use Factory Warranty service it all went perfectly. One was a DOA s7, promptly replaced with a new one (yes I checked it out) after they got the dead one in China. 3 other times were dead boards, all promptly replaced.

Still, that said: Their Warranty policy is my prime complaint with them.
Specifically:
1) cost. Cheapest UPS International to them for 1 hash board was around $90. The DOA s7 - $140. Ouch.
2) The Declared Value for Customs issue. This is HUGE.
I understand *why* they say to declare under US $40. It's to drastically speed up processing into China. However if the package goes a-wandering... Yer boned.

Point-2 is why I now always use Bitmain Warranty in CO for repairs - even for miners still under Factory Warranty. Yes BW in CO is strictly PAID service however even with me having a board insured for full value of $450 it still only cost $25 to ship it to them. Don't care that a typ repair cost is around $140, full value coverage is worth it.

Interesting. I have the NYA node as pool 0, the pass-thru as pool 1, and the DE node as pool 2 on both Avalon controllers (at work and GF house). All 3 dropped out on both locations. I just restarted cgminer on hers (a 741) and it's back to full speed. Interestingly, AFAIK only one of the Bitmains complained (beeping), an underclocked s7 in my office.

The ones at work: Tried cgminer restart. the 3x 721's on the controller are fine, the 741 barely over 6.1THs. GUI reboot of the 741. Same. Hard boot of the the controller and all miners on it. Same. All PSU's are APW-1600's and have always been trouble free.

Got a hunch the 721 and 741's don't play well together on same controller and have another Raspi hanging around so will try the 741 on it's own control.

 
About an hour ago the NYA, pass-thru, and de nodes went down for several minutes. When they came back up - my 4 Avalons (3x 721, 1x 741 on 1 controller) at work have now lost a total of 7-8THs! Rebooted several times, restarted cgminer several times - no go. Before the outage I was running 27-28THs with them, now down to 20-21THs. The 741 at my GF's house is also now running slow - down to 5.6THs.

I see that Canaan has also massively dropped from a couple PHs down to ~500THs. What has changed that the Avalons don't like??? All my Bitmain miners came right back up to normal speeds.

I do hope the question is an "April Fool's" joke otherwise the OP is seriously out of touch...

re: T9's
Just *where* has anyone had a complaint about the T9's aside from price/power/noise?
I've pretty much diligently followed the Bitmain T9 thread an cannot recall a single problem with them. That statement requires proof so link please?

Ja. Looks like Canaan has another batch of miners online for burn-in. 2 days ago they were down to 280THs, now back up to 2.3Phs

IMHO it is your first one: To squeeze out the highest possible yield of the 16nm chips Bitmain has seriously widened the QA window for what are "usable" vs good vs Prime chips.

Looking at speed/hash rate spread between boards I'd guess at least 3 levels are used for binning chips to be used when populating boards targeted for highest efficiency miners (s9/R4). Hopefully humorous example would be 'meh', good to very good, and finally best. Given the typ yield of 16nm that makes for probably 2-3x more 'meh' to good (say >80% of cores work) chips than best (>90% cores work at full speed) grade so....

Tie that with lapses in board assembly quality checks... On the 2 s9 boards I had fail the placement and soldering of the VRM was -- horrible. Frankly, surprised it passed at all considering the module was rotated enough to be just literally a hair breadth away from shorting right off the bat. How/why that wasn't caught by an inspection camera, who knows...

With the T9 hopefully their sacrificing eff by pushing the Vcore higher is what results in otherwise 'good' grade chips performing as very-good or better. As to the board assembly - who knows...

Even in China, selling used anything as New violates the law. Sure that only really works for Chinese Nationals as far as prosecuting goes but it IS excellent grounds for FTC to bar imports so. Bottom line is no they don't/can't. For warranty replacement them using reworked/repaired boards is allowed and in fact is very commonplace throughout most electronics industries.

As for getting dirty/dusty miners... Ever been to any large city in Asia? When it's dry they are often very dusty places. The miners are typically all burned in for at least 72hrs in open buildings. Dry out of the rain yes but sucking all that city air well you get the picture.

 

Try reading his threads....
The 1384 is the last chip from Bitmain that can easily salvaged and reused! All others after it are damn near impossible to work with due to the smaller pitch and issues on removing the heatsink adhesive.

However - he does have miners on the drawing board using NEW 16nm chips from BitFury.

a) Starting multiple threads for same subject is a no-no and just clogs the Forum. If you need to bump, just add another reply/comment to your existing thread.
b) Yes, could well be a router issue. Miners do not like internet dropouts.
c) Open a support Ticket with Bitmain.

Ja your experience with them truly sucks. Thankfully most people fair much better.

The T9 on order will make it 82 Ants I've gotten since 2014 and to date exactly 4 boards failed. 1 from an old s7, 1 from a R4 batch-6 failed after 24hrs and  2 s9 boards 1 a batch 1 the other batch 12, both failed Nov-Dec last year.

heh! Wadda ya know. Just checked and one of the blocks found on KanoPool today was from a user with 0.89TH/s! I guess the pool does still take maybe even as low as s1/3's?
But talk about a small % of payout...

How would you handle Whales dropping in to the pool and swamping out the small fry?

For all the reasons you gave I say go for it.
Yes is nice pointing my Sidehack sticks at the solo lottery but -- I can see low hash rate folks wanting a shot at pool earnings.
As you said, Kano loves my multi-THs s7, 9, A741's, etc., but forget pointing sticks at it.

Actually, not sure if KanoPool would even still like s1's or Sidehacks's upcoming Pod (>150GHs). Does it?

Well since I'm running out of s7's to replace with s9's or Avalons ( I *do* have a power limit) I broke down and decided to order a T9 for the farm at work. Ya know, just to round out me having had at least 1 of all Ants except for the s5+ (R1 and U3 don't count).

On a per-THs basis, is a fair bit cheaper than the A741's and on par for power eff. Question still to be seen is, are the T9 as reliable as the Avalons? Only time will tell....

Considering the USB is for com's only and the pod uses 12v for power: yes. If you can run cgminer on it anything will work from RasPi on up...

No it isn't. Most pools are expecting miners that are THs+ and their minimum diff's reflect that.
Use solo.ckpool.org as the name implies it is a lottery solo pool however it accepts ALL speed miners esp. sticks.

Unless you and your family are deaf, the s9 and T9 are NOT 'home miners'...
Definitely made for industrial/commercial use.

This seriously belongs in the Speculation thread....

On top of that, in the US there *are* NEC regs covering protective devices including surge suppression NFPA 70, the National Electrical Code (NEC), Section 285. A good coverage of it and more importantly - proper implementation - is http://ecmweb.com/code-basics/surge-suppression-devices

Key point for your electrician to reference:
"Installation. SPDs aren't essential to the prevention of fires, so the NEC doesn't require anyone to install them. However, incorrect installation poses a hazard. Therefore, if you install SPDs, you must install them per Art. 285. Don't install SPDs in any of the following applications [285.3]: "

Well until he pipes in, with him saying
pretty sure you aren't gonna get 12v from a USB jack