Quote from: NotFuzzyWarm on August 26, 2014, 02:01:00 AM

And I might as well add that obviously under legal obligations set forth in NDA I cannot publicly confirm nor deny my participation in said chat room, nor can/will private pm's between me and Josh be publicly released. All dry technical stuff anyway...

My overall forensics role of v1-6 (more like 0.1-0.6) hash boards and eventual minor technical advice role to try and aid AMT's customers has been made clear numerous times in the Forum.

What can be inferred from said leaked chat room posts I have no control over.

If I'm not mistaken, if one's able to infer something beyond a reasonable double, then it becomes proof.

https://bitcointalk.org/index.php?topic=569769.msg7610324#msg7610324

Quote from: NotFuzzyWarm on July 01, 2014, 12:21:08 AM

Just tossing this out there for general information:
http://www.edn.com/electronics-blogs/the-workbench/4421218/PCB-layout-tips-for-thermal-vias

http://circuitcalculator.com/wordpress/2006/03/12/pcb-via-calculator/

http://m.eet.com/media/1200036/XLamp_PCB_Thermal.pdf

Those are examples of standard industry practice reference items for thermal via design. Even if not explicitly specified in 'build-to-print' any CM should:
a. Knowing the thermal needs for the A1 assumed maximum thermal transfer is needed meaning; heavy-plated thermal vias at a minimum preferably filled as well, and a full complement of thermal contact bumps on the bottom of the board that are high enough to go past the gap created by green screen layer to make firm mechanical contact with the main heat sink. btw, that layer creates a 3mil air gap. If there are a few bumps but not a full complement - that gap just got increased to 8-10mil. Thermal compound is not going to make up for that...

b. Flagged the above concerns and asked their customer to verify what is needed to meet industry-standard practice for working with high-power packages. If the customer is unsure then follow Best Practice and overkill the issue.

Then there is the matter of basic Ohms law... If the value of 1 component in a divider network is changed while leaving all others the same - the result (output) from that network is vastly different. To keep the network balanced all components need to be scaled by the same percentage.. Even if a customer signs off on the ECN I still would have insisted on a signed declaration spelling out the result of said change just to CMA.

I will leave it up to the Peanut Gallery to figure how this applies to us here Tongue

I will say that a power chip missing over 50% of its expected cooling path combined with getting hit (very briefly) with excessive core voltage is not long for this world... These are all items that any CM with a design team should have found in initial design review and later design change reviews and known how to address properly.

All this of course in no way gets AMT off the hook for this mess. No matter what they are the ones that have to deal with it. It is just a glimpse into what the production problems were. Their horrible lack of communication skills is a whole 'nother matter and well worth the tomes that have been written here.

Question: How much longer do you desire to hide behind a man due to an NDA that is on record in wanting to burn down a home while 6 people and a dog are inside?

http://s15.postimg.org/lo5naqwnv/zipkin_threat.png

Here's an image of the dog, Trixie, that Joshua Zipkin wants to kill via burning down a house: