OK, here's the summary of the surgery I performed on my HL-R5067W.
I basically used the same materials as tito (#8-32x1/4" bolt and pronged tee nut at the end). But I knew going into this that I'd need something at the end of the tee nut to get more uniform pressure across the entire DNIe BGA chip. If a failure is going to occur on a BGA, it's usually due to warpage through various thermal cycles, thus creating significant stress on the corner edges of the package. This makes the corner BGA balls on the substrate tend to crack or even pop off in some cases. Because there's a wire bonded mold cap on the top of the DNIe chip, it's impossible to get pressure across the entire DNIe package by simply bolting something in the center (just over the black mold cap). The corner balls under the green substrate are still exposed. Best case scenario would be to use a "top hat" heat spreader with a cavity that goes over the mold cap, with the edges directly touching the outer edges of the substrate. Something like this.
I had a square Cu heat sink that I was hoping to use, but it turns out it was too big. So I had to improvise, and needed something just to prove that some kind of pressure on the DNIe chip would fix my problem. Best thing I could find around the house was a half dollar coin that extended beyond the diameter of the tee nut. NOT the best long term solution (especially since it's round), but good enough for a first try I guess.
Here are the photos of my "teardown."
All back covers off:
Gutted DLP after removing analog / digital board cage:
Digital board housing after removing cage:
Digital board exposed, with DNIe chip in bottom right hand corner:
My first crack at the clamp using a half dollar on the end:
Outside view of installed clamp (you can barely make out the coin inside the housing):
Side view of digital board after clamp installed. Tight fit!
I had a slight scare when putting everything back together. TV appeared dead. Turns out I just had a connector in the wrong spot. But I must say... after installing the clamp, the TV was without ANY issues for the first couple of hours that I had it on.
But...if problems resurface in the future, I'll suspect it's because there's not sufficient pressure to the entire DNIe chip. The problem is we don't know exactly how much
pressure is enough, and whether it's truly uniform pressure across the whole chip. But as a quick fix, I'm pretty happy with the solution that was figured out by Jason1976 and the subsequent improvements made by others on the forum. I'll continue to monitor my TV's performance and let everyone know how it works over the next few days. In all, I was only out 2 bucks for the clamp / bolt fix. Well, $2.50 if you count the half dollar coin I used...