I'm in the same boat.  I ordered 3 units of batch 2 and the first one out of the box had a faulty controller.  I got the other 2 running, and they're hashing away at normal rate with no modifications.

I contacted their tech support about the faulty unit, did some troubleshooting that they advised (swapping it with a board from one of the working units, using different PSU, etc), and isolated it to the controller board on that unit.  This took only a few days of troubleshooting. 

Now I'm at over a week, and I finally get an email from their support with a shipping label.  I asked for clarification, and they expect me to ship the faulty board to them before they will ship me a replacement.  A process they say will take another week.  So I'll be over 2 weeks with a dead unit, and the next difficulty looming - basically at least .166 BTC lost because of their faulty hardware and slow response.

I asked about compensation, and they said they do not offer compensation.  Given that these units are priced so as to be barely profitable, I am now in a situation where I will most certainly lose money on the deal, while Bitmain has potentially made millions off of their customers.  I find it staggering to think they can't offer simple customer service solutions that are commonplace in other industries - especially with mission critical or time sensitive applications (which bitcoin miners undoubtedly are).  I mean, at least cross ship the replacement board or offer that as an option.  But they drag their feet, and insist on a time consuming return process basically ensuring I'm out any profit that unit could have delivered had I got it mining the moment I received it.

I have another 2 units on B5 and I've requested they cancel my order if they can't make this right. That might seem like an overreaction, but I have a simple approach to those I do business with - and if they don't treat me like I would treat a customer in the same situation then it's no deal.

Hopefully we can all make some noise here and get Bitmain to change their tune.

I wonder why, its not as if someone can run off with another controller and build an Ant from it........it should just be heres another one and toss that piece of shit in the bin....they must know that about 1 in 10 of these things are faulty.

what about old S1 controllers?

There was a bounty thread somewhere around here for making old S1 working with rPi. My S1 is still profitable, and I would like to maybe upgrade them soon with S3 blades.

Bitmain mentioned that they will sell upgrade kits. 

The bounty is offered by me and some one else also promised to add additional to my bounty.
The bounty is still active and anyone can participate to win.
The link to that thread is https://bt.irlbtc.com/view/671128.0
The aim is to run multiple(>2) blades with a single control board.
You can do it on S1 or S3. Hope soon participant will join as more S1s are available at cheap price now.
On a side note: how we track our old threads open by ourselves? In other forums there is a page were we can see threads started by ourselves.
But I think I miss it in bitcointalk.

Have you tried swapping PCI power connectors? Hope you have done it.
Then if not working, Open your faulty antminer S3 and swap the cables from blades to control panel.
You can also swap the control board of another working S3 for testing as a second option.
If now the other blades are working its a control panel or cable problem.

Now, open the front side (ASIC Chip side)heatsink and check for thermal paste lease or absence.
If its so, just buy some good thermal grease and apply, the blade will be ok.
Else its a clear case of RMA. Contact bitmaintech as soon as possible via email.
I assumed you are using a proper power supply.

my miner drops down to 410-415GH. I would rather keep it running and pay a deposit so he would ship a new controller out and refund me after he receives the bad one than shut the miner down for a week or two.

It may not be a controller issues as from your description.
Are you seeing any "x" or "-" in you asic status page?
Give me your Ant S3-es fans speed, temperatures of miner and ambient temperature?.
Most probably its, the heat dissipation issue and the chip are throttling to lower clocks automatically as in GPUs.
As already advised by bitmaintech support engineers, unscrew the chip side heatsink and check the condition of thermal grease.
Remove it with a cloth and apply new thermal paste grease and it will be mining back at 441.
Use a good thermal grease which provide conductivity above 3.

I have sent the same request away as well. Ive asked for a cost of a replacement controller.

I know this will have been asked before but for some reason I'm not getting any results back.

What is the UK VAT on Average for the 2X S3 ?

Thanks

DNS is set on control board not on blades.