Just looked into the Restricted API access and speed changes using Awesome Miner: You need either the Professional Edition or higher to use those functions. Looks like time for me to bite the bullet and upgrade from Standard to Premium (good for 20miners per-site).

Nice thing is that Premium also unlocks remote access for monitoring/rebooting my farm at work

Have you tried powering just the controller board and while leaving the hash boards un-powered, pressing the reset for >10sec? That seems to have unbricked another person's s7 here.

From their site "patented 256 colors RGB fan".
PATENTED??? How the hell did something as common as RGB modulation of LED's in a fan or whatever get a patent??? 

No idea why it took so long, luckily my 2 warranty issues with a s5 and a s7 were handled within 2 weeks. I do see that in the Forms to fill out for service Bitmain has added several 'special instructions' to the WO form... https://cnshop.bitmain.com/workOrderGuide.htm

"Shipping way:
UPS,DHL,FEDEX,EMS, TNT, USPS, or your local post office are available.please note that EMS and USPS provide poor service in HK,we cant ensure they can delivery the parcels in time.
NOTE:
1.   Be sure mark "deliveries are made on working days only" and "Antminer" on the items description of
      the waybill.Otherwise we wont accept any delay due to missing of these info .

2.   It's necessary to package the hashboard very carefully. We can not repair if the PIN on hashboard is lost because of the shipping problem. Please claim to your logistics agent if the PIN is lost on the way.
3.  The HongKong warehouse only responsible for receiving the parcels ,we wont accept "Freight Collect",please pay
      the shipping cost in your side. Please mark the warehouse NO. 3811
Consignee: (The package will be rejected if there is no "No.3811"; The PSU can not be packed with other stuff in the same tracking number because of customs clearance)
Name:FUKWOK KWONG-No.3811
Company Name:TUNG-D INTERNATIONAL DEVELOPMENT LIMITED Warehouse No.3811
Address:Unit 104, 1/F, Sunbeam Centre, 27 Shing Yip Street, Kwun Tong, Kowloon, Hong Kong
Office Tel:+852-2750 5059 (The number is only available for logistics not for service )                
Warehouse:Tel: +852-2750 5150(The number is only available for logistics not for service )"

That looks to open a lot of loopholes for them to claim 'lost in transit'....
PIN?? as in the barcode sticker on the boards??? What would a logistics company have to do with that?

Now as to getting back a different board:
Well, ja. That is how most often any electronics servicing goes. Rarely is any board repaired and sent back. Even using low-cost labor it is usually just not worth the time to do more than verify a board is dead. So, either a re-worked board (from production problems or returned boards from other customers that were able to be fixed) is returned to the customer or if really lucky -- a brand new one is sent.

Of course given that test/replace (not repair) is the norm, the question still remains why it took so long for Bitmain to get a replacement out to them?

Now considering that ^^ was in reference to a post from Tupsu who is certainly NOT a friend of Bitmain, rather amusing aye-what?

Since it appears that this idiot is of the opinion that anyone not jumping on his honey truck to join them in the crap slinging has been paid off or whatever, now Ignored here.

What a yutz...  

Proof of no s9 sales to China? Without that you are just trolling.

Zero interest in your idea

That violates every countries National Electric Codes. For power wiring a neutral must ALWAYS be tied to a ground at the incoming power panel. That is what makes it a Neutral vs just another Hot wire...

Technically, the major distinction between Neutral and Ground wires is that Neutrals are for carrying current whereas Ground is strictly a safety wire.

What you described is commonly referred to as 'isolated power' and is restricted for use only with very sensitive test equipment. A variant of it is an isolated ground setup (usually orange power sockets) that is much like you described except the neutral is still tied to ground but with care that the Neutral/Ground bonding only happens at 1 location.

As to single phase PSU incoming power connections -- Makes no difference if L1 or L2 are Hot and the other is Neutral.

Access to free power is a really good start to making money mining.

That leads to parts 2 and 3 of the equation:
2) How much power/is it in a readily accessible area for miner maint?
   and
3) Heat removal from using said free power. Remember that miners are 100% perfect heaters -- all power fed in comes out as heat. A LOT of it!

Ouch!
I know the last 2 s7's I bought (b18) are set to Restricted as well as 1-2 others from earlier batches.

To do it I assume I'd use PuTTY to SSH in but from there...
Where is the config file located?
What editor and more to the point, what commands are used to change the permissions?

And of course, wonder if this works with Bitmain's fork of Kano's CGminer on the s9? ...

Most likely need to change the s9 miner software from Restricted API access to Privileged same as some s7's need done?

Just so ya know -- some s7's and all s9's operate in a restricted (vs Privilaged) API/SSH access mode... All operating data can be read but not all can be changed & saved outside of using Bitmain's GUI.

My s7 b18 is one of them. M's Miner Monitor simply does not see it and I believe one from an earlier batch is the same. It also has never seen the s9's and since that program has been abandoned, that leaves Awesome Miner Monitor.

Finally pointed Awesome at the s7 batch 18, yep. It sees it and reports it as restricted access.

Gotta change that and the website for Awesome has some info on it http://www.awesomeminer.com/help/minerapi.aspx
but.....
a) does the privilege hack work for the s9?
b) what 'ux is Bitmain using? What editor is in it? What are the edit commands...
c) What is the file to be changed and where is it located?

Ant tools can set the voltage on s9's?
Even though I got the v0.16 zip from Bitmain, the program shows v0.14 in the Title bar and the pull down for Miner Type only goes up to s7.

 So it works for the s9 as well? Using what miner type -- s7?
Also, do the settings stick through a hard reboot or does Voltage go back to the default?

Curious minds want to know...

With BTC currently = ! 610 USD that may take awhile... Eeep!

Or... they are fulfilling order for mega customers like https://bitcointalk.org/index.php?topic=1566654.msg15721138#msg15721138 and https://bitcointalk.org/index.php?topic=1566654.msg15764113#msg15764113.

Not to mention of course probably Hashnest.

"Several thousand s9's' is a helluva lot of chips and with TSMC no doubt having a monthly production limit set for Bitmain's boutique ASIC's, well you get the picture...

On BitFury, ja it is looking like their chip making divisions foray into 16nm will not make it to the Public. But BF as a whole... They do a helluva lot of other things including blockchain analytics so bottom line is probably still pretty good for them. The losses from 16nm chip making and of course related to that, any containerized farms they sold based on it will just be write off's to them and backers/investors.

As for BTC being a commercial phase and becoming more so -- Satoshi or whoever came up with it designed that in from day-1. What it ultimately boils down to is that BTC block-found payout halving will always happen until finally all income is derived solely from Tx fees vs miner's finding blocks. The idea of course is that folks will have to pay reasonable Tx fees to use BTC -- like the CC card / banks/ PayPal. et al charge.

Anyone looked at what the typical tx fees per block are currently? Hint: Not much vs the 12.5BTC miners bounties... On CK pool typical best additional payout from Tx fee's that CK collects & distributes to us is around 1BTC.

Once Bitmine.ch finally paid attention to what had already been well documented in the A1 chip dev thread about heatsinking and the importance of a decent Vcore regulator design -- like makers of the Dragons and Dragon clones did -- ja the their A1 rigs worked well. Far to-little-much-too-late.

If they had had competent VR circuit and PCB designers along with paying attention to thermals they could have finally started filling orders months earlier. The set of hash boards that Joshua Zipkin (now aka Joshua Alexander) owner of AMT (Bitmine's only Authorized USA dealer and using their PCB's) finally sent me in June 2014 for forensics covered Bitmine's revisions from late Jan up to June. All were riddled with the exact same major design flaws.

Good question. Anyone have a window air conditioner to put one in front of? I'd think that feeding in say 60F or less air temp or hooking the s9 to a high CFM blower, say 500-600CFM  could let you boost a fair bit, maybe even 675. At the higher clocks (with cold inlet air) it will become more of a problem with just how much power the buck regulators on the hash boards can provide. To even hazard a guess on that need to know what FET's they use, maybe a pn for the rather hefty inductor as well and base the max current they can handle on that.

Of course if Bitmain ever gets around to releasing the data sheets on their BM1387 chip then we'd have some of these answers. The sheet for the s7's BM'85 lists clock speeds almost up to 1GHz along with suggested Vcore to use...

The surrounding components is a pretty important point -- mainly the decoupling caps around the chips and the bypass used for each node when powering a string. It may look like, "well just put some (random value) caps here and..." BUT -- the actual values and more importantly,  where they are located and what type they are (material used) is of vital importance. You use different ASIC's, you will need to find a new set of optimum values for the caps or at a minimum, verify how the different ASIC's work with a compromise set of values.

In many ways the power planes feeding ASIC's are much like RF circutry and must be treated as such or else you get some mighty weird and very localized dips/spikes in the 'DC' power feed to them.
ref:  http://powerelectronics.com/power-electronics-systems/five-things-every-engineer-should-know-about-pdn

Hmm, good to know. My B1 has no problem hitting 14THs and evena bit higher -- provided I have an ambient of <80F. At a higher ambient I have to lower clock to 600MHz or just a step or 2 higher at most otherwise chip temp on board #3 hits 105C or higher...

Tell ya what, I'll be getting 2 more when Bitmain has more s9's for sale so how about I have one drop shipped to ya to tinker with?

One thing I'd like to be able to tweak is the starting voltage. On my batch-1 and one batch-3 it usually takes a couple reboots to get all the cards firing on all cylinders. On the b1 it is card-3 (the usually hot one) but oddly enough on the b3 it is the middle card that needs a few restarts to get up to speed.

The hotter board caused by the lower airflow there is not 'defective engineering'. It was an accepted design choice. Since the boards in that location do work acceptably well then form-factor overrides perfection.

Perfection is not only unneeded for most things, it is also usually very expensive to build. Part of Engineering is learning the very gray area between, "Well, it works" and "it works perfectly in all ways". Most often the final product will fall somewhere between those two points. For consumer goods the roadblock on the path to perfection is usually Marketing ideas of what that (perfection) is and cost.

Now if we start seeing failures of that board more often than the others, then you have a case.