Yes and no. Yes there is a power on LED and no I didn't try to power it up. I honestly didn't think to try. I was concerned with energizing it just long enough to get the measurements. I 'think' I read it won't bench power on out of the video chain but I'll have to confirm that.
No power led @5v with diode as is. So damn curious what it is about these boards/design that causes this exact same failure in a large number of these things when mains are removed or applied. Beginning to wonder if there's a design flaw in the layering of the boards that pops up after heating and cooling and causes a localized short in the area of that diode pad. smh Where to next?
While waiting for the next step I tested every SMD cap on both sides of the board for a short to ground. A few didn't have a ground polar side and I assumed these were for circuit series filtering?...I did find one cap that was infinite on both sides. I've attached a rather blurry pic but it should do. The white on it is from me...I used a grease pencil to mark it while I continued testing.
This cap is directly underneath what I assume is a processing IC. Part number below. XILINX SPARTAN XC3S700A FGG484AGQ1045 D4171258A
Unfortunately it's a BGA. I can remove it with a heat gun but am not equipped to install a replacement, (assuming it is the problem due to an internal short and one can be found), or to re-ball this one. I was going to remove the cap and test for continuity at the pads but figured I'd best wait for you to give me the nod...that and I really don't want to try and keep track of the poles on the little blighter.
I went to the manuf site xilinx.com and read the datasheet (well, tried to). This is a fairly complex programmable FPGA that does IO, DC switching, and a slew of other data handling chores. From what I could deduce it would have to be programmed by Mitsubishi to conform to the 3DC-1000. Even if I got my hands on a direct replacement it would be devoid of their programming and useless. Hopefully the problem is the cap I pointed out is shorted internally and not getting a 'false' ground on both poles from a short internal to the XILINX. And if it does prove to be the cap...what value is it? It would be foolish to assume an adjacent one is the same value I know but without a SM would there be any other alternative? Time to pull out the Blue ESR and play some more.
Last Edit: Mar 2, 2015 16:05:30 GMT -5 by justgreg
My bad....false alarm. I tested the heck out of that cap repeatedly earlier and got a tone on both sides every time. I just retested it and this time it's behaving normally. This set of leads has probes that are relatively stubby and unfortunately my pin probe leads are at a friends house. Crap. Thought I was onto something. Back to the drawing board.
Well here is where I am at. First you found a blown fuse and verified a short to ground. Using heat testing you found a hot diode. Lifting that diode removed the short and voltages both in and out of regulator were verified as good. Powering device no longer indicates a short, at least not before or after the primary regulator. Testing the non connected pad to the 2 ic's was to verify the pad was in fact a ground plane. It was also testing the 2 ic's for a short condition. These tests conclude that the diode is being used as a surge suppressor. So the device should work properly even with the diode lifted. Since you said it does not then I would first ask if you tried the power button to over ride the auto feature to see if it would turn on. It seems odd, because if it took a surge and the diode when short and the fuse then blew then the regulator or cap should of been next to fail. Since they are not, then why does it now not power up. Had it damaged another part we would most likely still be showing a short on the 5v line. Maybe try powering it up again and test at those test points that are marked on the front and back of the board and see what you do and do not have.
Thank you for defining all the tests we did already. I was wondering about the diagnostic nature of them.
Just so I understand...am I applying full 5v with the diode soldered back in place and the missing fuse bridged?... or should I take a single strand of ubber thin wire and make a makeshift fuse? I salvaged some tiny arsed SMD fuses from an old PC CD-drive the other day but have no idea their rating as it wasn't marked on the board. I'm not even confident I can get one back in without smoking it but there's only one way to find out.