ERROR: Wrong response from the Necromancer Box

[Adam Outler]

//////////////////////////////

[Adam Outler]

//////////////////////////////////////////

[Adam Outler]

///////////////////////////////////////

[Adam Outler]

hrmph... all I have to say to that is
///////////////////////////////////////

[legija]

Four posts, and all is "///////////////////////////".

Seriously, there is a point why posting complete log and detailed explanation of what You're doing is required. It makes us much more easy to understand what customer wants to do, and what exact problem one can have.

However, You can also continue to ignore rules, it's up to You.

That applies to military personal too ! Being ex officer myself, I tend to give orders time to time.

[Adam Outler]

I was away from my box at the time I posted. I couldn't get the log. I snapped a picture on my cellular phone and read back what I could make out.

This is purely hardware related. It seems that this forum is not really hardware related, but riffbox operations. Anyway... I think I will have it working once I finish removing all the excess solder put on by the factory hand-soldering job.

////////////////////////////////////////////////////////

[Adam Outler]

Ok.. so first I had no connection to windows on this factory sealed JTAG box. It would not connect. So I busted the factory seal on the unit and reflowed the processor. It had factory installed cold joints. For GP, I also reheated every connection on the processor and FPGA.

Reflowing the over-accumulated solder on the pins created more problems which were very difficult to track down. There was so much solder that it created micro-bridges between the pins which were not easy to see because they were under the pins and flowed down onto the board.

The problem in the title "Wrong response from the Necromancer Box" was the only error I received and it was when I tried to "reset and go". Otherwise the unit would stay at "connecting to dead body".

Over the course of the 10 hour repair job, I removed about 1/2 of the solder from the board. It took about 8 inches of de-soldering braid and more flux then I have ever used on a repair before. I basically had to coat the entire board in flux, reflow, remove solder and then reflow again, allowing only the proper amount of solder to be left in place.

So here's my work... My iron does not have the heat reservoir to melt the larger joint. You can see how much solder is on those, this was pretty typical of the entire board.... Like it had been done by someone on their first day of work. This type of over soldering is a novice mistake made by most people when they start.

http://i52.tinypic.com/24yzf2u.jpg

http://i53.tinypic.com/315fiah.jpg

http://i56.tinypic.com/2ag20yd.jpg



So, it's now been corrected. This board started off with an obvious processor issue, then in the process of repair, it had a FPGA issue, this was corrected after much troubleshooting.. Now my Riffbox is better then the way it was new.

////////////////////////////////////////////////////////

[Adam Outler]

RIFFBox development team,

For extra rigidity, it may be a good idea to look at some spray-on epoxy of some sort. I have seen this on critical areas of the Samsung devices which keeps the device from using the solder joints as fixtures. We all know that solder joints are not supposed to be mechanical joints, but rather electronic connections. Samsung came up with a great solution of using a flexible epoxy after testing is completed which ensures the components are held to the board by alternate means other then reliance on the electrical joints.