Eric Wallin's Internet Homepage

Video Projects: Instant DVD 2.0

As documented elsewhere on my web pages, my 2 week old problematic DVD Xpress spontaneously died after having (what I convinced myself were) heat-related issues. Call me crazy, but I took the dead unit back to the store and forked over an extra $70 in exchange for yet another ADS product: the USB Instant DVD 2.0. Ulead Movie Factory 2.0 comes bundled with the new unit, otherwise I might have opted for a simple replacement of the DVD Xpress.

Well, the IDVD2 was working really well up till a couple of months ago, with maybe 50 or so hours of capture under its belt, whereupon it began to display exactly the same early symptoms as my previously owned flakey-then-dead ADS DVD Xpress: after an extended capture, the flickering capture LED would go out and the video and audio would start looping the same last couple of seconds of material. The Windows "ba-bunk" sound of a USB disconnect would also often be heard. For the DVD Xpress this behavior was definitely aggravated by heat, as I could capture indefinitely (before the unit completely broke) if I removed the PCB from the case and aimed a fan at the chips.

I have been very careful with the IDVD2, keeping it plugged into a separage power strip so that it is only powered-up when actually doing video capture. This may have contributed to its longer life vs the DVD Xpress, but after opening the IDVD2 case, it was clear that it had inherited the same poor thermal design of its little brother.

To demonstrate this to yourself: open the unit, plug it in, and do some capturing. While it is capturing, place your finger on each of the chips highlighted in Figure 1. For me, they were too hot to keep my finger on them for an extended period of time. I can only imagine how hot they get when the PWB is completely enclosed in the case and they receive effectively zero air-flow.





Figure 1. Chips that get excessively hot during capture are highlighted.


Heatsinks

Heat, besides being an instant electronics killer in some situations, will also degrade service life and digital timing. I believe the issue here is timing, but have not attempted to prove it. Regardless, the heat inside the case needs to be dealt with, and there are two thermal issues here: hot chips and lack of airflow. The hot chips need heatsinks, the application of which is shown in the next few figures.





Figure 2. Your basic thermal crusaders: heatsinks and "hardman" (who names this stuff?) epoxy.


The heatsink material is stock stuff I found sitting around. They are cut so that they fit on top of the ICs; any excess height should be trimmed so that they clear the top of the enclosure before they are glued on. Otherwise you might stress the PWB by trimming already glued heatsinks and break something. The glue I used to affix them was a generic overnight epoxy. This is a permanent bond, you will never remove the heatsink without destroying the IC and the board, so be careful!





Figure 3. Heatsinks cut to size: check placement and height restrictions before gluing.




Figure 4. A better view of the heatsinks that had to be trimmed due to height restrictions.


When gluing the heatsinks, mix the epoxy really well and use just enough to cover the chip. Rub the heatsink around a bit after placing it in position - this will force out any excess. Epoxy is non-conductive, so don't worry about it getting on the IC pins.

Set the assembly aside to cure overnight on a level surface, and check it a couple of times before it sets-up. Like a dummy, I placed mine on the computer monitor, only to discover that the heatsinks had very slowly slid to one side over the course of maybe an hour. Good thing the evil anal side made me check up on it before the epoxy solidified!

Fan attempt #1 - The ill-fated hidden fan

Ok, now for getting some air over those heatsinks. I really wanted to find a fan solution that would fit inside the case and not show. I scrounged a worn-out fan from my junk box and affixed as shown below. Sealing clay, such as Mortite or Frost King, was used to hold it up off of the PWB. This is a 12V fan, and runs more slowly off of the internal 5V sourced by the IDVD2 AC adapter.





Figure 5. First fan attempt: the ill-fated hidden fan.




Figure 6. Ill-fated hidden fan detail: clearance with case top.


Alas, the hidden fan didn't have much of a duct to work with, and so my thermal problems with the IDVD2 came back after a 1:15 capture. Time to call in the big guns.

Fan attempt #2: Doing it right

So I hit all of obvious places that carry computer stuff: CompUSA, RadioShack, BestBuy, etc. I wanted a small (but not too small) 12V ball-bearing fan that would run at 5V and draw reasonably low current. An included fan guard would be a plus.

CompUSA seemed to have the best deal with their SKU #280152, a fan / guard / CPU heatsink combo for $15. Bought one, took it home, took it apart, and discovered the advertised "ball-bearing" fan was in fact a sleeve bearing fan - damn you CompUSA! Oh well, the 12V fan would be operating at a substantially reduced voltage, and only intermittently, so it should be good for some years. Still, I'm a bit pissed at CompUSA for lying to me. And you wouldn't see this unless you disassembled the thing, which seems extra dishonest.





Figure 7. Fan / guard / CPU heatsink combo for $15 from CompUSA. Note blatant failure of "truth in advertising".


I hooked the fan up to a variable power supply and noticed the following: the fan starts spinning at around 2.5V; it starts reliably at 5V and draws 68mA there. Good stuff.

Ok, time to adapt the case to hold the fan. I marked out a large circle and the screw hole pattern on the top silver piece of the case (the screw hole pattern is a 2" square, can't remember the large circle diameter). Drilled the large hole out with a small drill bit (see Figure 8), snipped it out with a diagonal cutter, then finished it with a curved coarse file and some medium sandpaper. This took maybe 15 minutes and turned out pretty good IMO.





Figure 8. Adapting the case to take the fan.




Figure 9. Adapting complete.


I soldered a couple of solid wires to the PWB after probing it with a DMM to see where I could find +5V and ground. An obvious ground point pickup is located in the large empty area to the left where some big component would go, probably for its solder tabs or something. +5V can be found at one pad of the also-missing L25 (see figure 10 for details). The solid wires plug into the fan connector, so you aren't ruining things if the IDVD2 completely dies and you want to salvage the fan for the computer uses it was intended for.





Figure 10. Fan / PWB electrical connection.




Figure 11. Fan mounted and in action!


I attached the fan with the same screws that originally affixed it to the heatsink. The guard is held on via the same screws. If the plastic top strips out then use machine screws here.

With the fan operating I can feel significant air coming out of the various holes in the case, which is great.

I've captured now for a couple hours and the case never gets hot at all. No USB disconnects, no box crashes.

Who knows what the future will bring? All I know is, for now:

My IDVD2 is capturing reliably again! Woo-hoo!

Back home