Sixerdoodle Electronics

Looks like I’ve got another board with capacitor problems. This one I’ve caught before the ultimate failure. Here’s a shot overlayed with the temperatures of the various capacitors I took with an IR sensor. Note that the right most two caps are already showing a distended top and they’re also the hottest. Can you imagine 164 Deg F?? heck the CPU is only running at 115F so it’s not airflow from the CPU heat sink warming the caps, it’s internally generated heat Certainly soon to fail…

ok, so accidentally ripped off the USB plug from the USB cable on a Magellan Roadmate GPS. Figured I’d just stick on another Mini-B cut off another cable, connect the + and – (pin 1 and 4) and be done. No go, the GPS thinks that it’s plugged into a PC and won’t navigate. Some investigation later turns up the fact that the ‘X’ pin needs to be grounded to get the Magellan Roadmate GPS to work right. (mini-usb is pin 1,2,3,x,4) and… the problem with that is trying to find a pre-existing mini-B with attached cable with the ‘X’ pin wired is dang near impossible. Ended up replacing the end with a usb mini-b with a solder tail connection like a KMBX-BP-KIT-ST30. But note that it’s a tough soldering job getting the wires in the Magellan charger soldered into the jack, including shorting X to pin 4, without shorting the whole thing out, but it is doable.

Quick and dirty interval timer for Nikon. It’s about as dirt cheap as they come, but it works. check out the video at the end of the project page for a sample time-lapse movie

One thing keeping me from updating the RecLight page is had another motherboard fail. As this is the 2nd one in same number of months, thought I’d document some of the problems / solutions.

The problem presented itself as the PC just randomly rebooting. the dreaded “Machine Check” error 25. Sometimes it would run for a couple of hours, then just bang like someone hit the reset button all of a sudden rebooting. Checked the usual things, the temps were OK, no disk errors, voltages, nothing installed recently, etc.

Finally took a good visual look at the motherboard. Guess what capacitors with bulging tops.

In fact all the 6.3V caps had distended tops. So I determined that was the likely culprit. bad 6.3V caps means bad 5V power on the motherboard. Now I’ve been running MotherBoardMonitor and it’s been reporting ‘normal’ voltages the whole time, so you think how could the board have bad power? The problem is the ripple, the A/C component on top of the normal 5VDC supply. The capacitors are meant to smooth that out. Bad caps, more ripple on what should be a DC power level. No way MBM samples the voltages fast enough to detect that ripple.

So, it’s on to “let’s replace the dead capacitors on the motherboard”. How to go about that? first, size’em and count’em. On this board, there are 5, 6.3V 3300uF caps (all bulging). Go buy some new ones, that’s the first step ’cause if you can’t find replacements, why bother taking the old ones out?

Couple of things to keep in mind when buying replacement caps

• of course make sure the ratings match. Note that most motherboard caps have a “high temperature” rating (ie 105 degrees)
• watch the physical size. You may find that physically it’s a tight fit. The replacement capacitors I found were slightly larger.

Once you’ve got replacement parts, now you’ve got to get the old one’s out without damaging the motherboard. This will likely be the hardest part. A good soldering iron is essential as well as a good quantity of solder-wick (braided copper).

One thing you’re likely up against is that the solder on the motherboard is lead-free, which means to you, it melts at a higher temperature than leaded solder. Worse, the power supply capacitors will likely have their negative (-) lead soldered into the ground-plane on the motherboard. Why’s this bad? well the ground-plane amounts to a (relatively) large amount of copper, which unfortunately here, is a very good head conductor. So not only do you have to get enough heat on the lead to melt the lead-free solder, you’re fighting the ground-plane which is sucking all the heat away from the lead.

On my soldering iron, I had to keep the temperature control turned all the way up and to have patience to allow the ground-plane to warm up too. You do NOT want to get antsy and try pulling the capacitor out of the holes. You’ll likely just rip off a trace and ruin the motherboard.

Some tips I can suggest

• Keep the soldering iron hot Hot HOT. The high heat will try to oxidize the tip, so make sure to keep it well tinned
• Use the solder-wick to remove as much of the lead-free solder that you can
• As crazy as it sounds, try re-soldering the capacitor lead, but use leaded solder
• You may want to repeat the wick, re-solder step a couple of times. Eventually you can get most of the high-temp lead free solder in the joint replaced with lower-temp leaded solder
• Then… apply a small dab of solder to the soldering iron and heat the lead/board joint well, let it soak up the heat a bit, then quickly apply the solder wick. The goal is to get all the solder out at one go because as soon as the top surface of solder is gone, you’ve lost your main path of heat transfer into the joint
• You’ll find that only the neg(-) lead on the capacitor will cause you grief. the pos(+) side has much less copper directly connected to it and so not the severe heat capacity issue.

Repeat the process for each dead cap and then replace with the appropriately sized replacements. For me, this gave the motherboard new life. Hopefully you’ll find similar luck!