There was once a time when most portable electronic devices used AA or AAA batteries. Many users would opt to use Carbon Zinc or Alkaline disposable cells with these products, which proved to be costly financially and environmentally. An alternative to these would be to utilize rechargeable batteries – initially Ni-Cd which had its own environmental problems, and then later Ni-MH.
These cells were initially poorly received, due to problems with lower voltage, slow charge times, self discharge rates so high they could not be stored charged for too long and perceived memory effect issues. These rechargeable cells have improved dramatically with the invention of the low-self discharge Ni-MH cell, which solved one of its major shortcomings. This also helped partially solve the issue of slow charge times, as you could pre-charge the cells and store them charged without them quickly losing charge on their own.
But charge time is still a relevant issue, especially if you have many cells to charge or many devices, which is why I had made it a point to invest in a fast charger. While fast chargers are generally harder on the cells and can shorten their lifetimes, most cells now have enough cycle life that even under such strenuous duty, they do not fail before their jackets are completely disintegrated and contacts rusted. I haven’t lost enough capacity on any of my LSD Ni-MH cells for it to be a problem!
It was a big shock to me when my current 15-minute fast charger suddenly stopped working. My next best replacement wasn’t in great shape either, and the thought of turning back to 15-hour chargers was painful. So I set out to examine and try to repair the charger …
How did I end up with the Varta Power Play Charger?
Normally, a repair story doesn’t really concern itself with how the product came to be in my possession, but in this case, I’ve made an exception as it’s a fairly interesting story.
I had adopted 15 minute charging in its earlier incarnation, a system known as I-C3. This system pre-dated LSD Ni-MHs and was a patented technology by Rayovac, rebranded as Varta at retail in Australia. The cells were special, featuring internal cut-off pressure switches and black resistance bands at the negative end of the cell which were sensed by the charger to determine the presence of an I-C3 capable cell.
Some time in 2006, I purchased a set of four I-C3 cells and associated charger for about AU$85. It made me pretty happy, as the charger was capable of individual channel charging, and met its 15 minute charge rate with the special cells. It could also recharge ordinary cells at about 1-hour rate.
But this joy was shortlived, when after 16 months, the unit exploded and got some gunk really close to my eye. It also smoked for a good few minutes …
I took it outside, let it cool down (as it was quite hot) and took it apart.
The gunk was on the inside, and had a pretty bad smell. It shouldn’t have surprised me that it was a capacitor failing – venting catastrophically in my face while plugging the unit in.
The unit technically had a 12 month warranty, which would have meant that I was out of luck. But despite this, I proceeded to contact Rayovac Varta Remington Products Australia on their e-mail address, expressing my dismay. They graciously sent me a satchel to return the item for their investigations and sent me a replacement of a newer sort – the Power Play system using the Ready2use cells. These did not feature any special cell-sensing device while still providing 15 minute charges. I received my replacement unit in November 2007 and I was pretty stoked at its performance.
The Failure and Teardown
I suppose I should count myself lucky that I didn’t go blind and that I even got a replacement charger. But it seems the replacement has decided to kick the bucket about 7 years in.
Just the other day, I was charging a set of four batteries when it suddenly stopped charging. No LEDs, no fan. Silence. Except for a persistent periodic ticking from the switchmode power supply that ceased when it was disconnected. I took this as a sign that the charger unit itself had shorted itself out, and was triggering the over-current protection of the power supply. Good thing it had a functioning OCP otherwise there might have been smoke or fire!
Most people would probably be pretty happy with 7 years of service, but I couldn’t bear to let it go.
I tested the switchmode power supply with a different charger that accepts an identical input, and it seemed to be working, so the fault must lie in this unit. The unit is held together with three screws at the rear. Once those are removed, the rear cover can be carefully removed to reveal the innards.
The unit is somewhat dusty from use, but its construction deserves some mention. The top portion contains all the main circuitry, and is constructed from two single-sided PCBs stacked one on top of another with plastic spacers. This stack is secured to the chassis by two screws – take care not to lose the spacers when taking it apart. The spacers themselves seem to have shrunken over time from the heat, so it can take some prying.
The negative contacts and indicator LEDs reside on a separate PCB wedged into the chassis and filled with foam. This is connected to the main PCB by flexible ribbon, which you should avoid bending. There is really only passives on this PCB, but the soldering is a little variable in quality. Surface mount components seem to be held by a red adhesive prior to soldering.
The fan itself is a 50mm x 10mm XinRuiLian XFan unit just fitted into place by a friction fit.
Thick copper flat-wire straps are used to connect the positive terminals to the PCB. On the rear, you can see a Samsung F9454B 8-bit CMOS microcontroller which forms the heart of the solution. Each of the charging channels seem to be controlled by a pair of AnaChip AF4362N N-channel MOSFETs in SOIC8 format. I didn’t take a close look at U3, but I suspect it’s some switching power controller IC.
Pulling the whole stack out of the chassis took a little prying, but we can see one calibration pot on the bottom PCB. We also see a few capacitors – the large bulk capacitor shows discoloured glue which is expected due to the high temperatures. The capacitor is a G-Luxon 16v 1000uF unit rated at 105 degrees C. From my bad-cap days, I am suspicious of the capacitor, despite it not being bulged yet.
There’s a few more components on the underside of the bottom PCB. I didn’t examine the majority of them but I already see a suspect component – can you spot it?
[… pause for dramatic effect …]
If you said Q16, that’s right! MOSFET Q16 looks pretty toasty! A quick check shows it’s a through short on all legs! Yikes! No wonder it’s shorted the input supply to ground – hopefully it didn’t damage anything else with the high current pulses it must have been causing during the over-current trip.
A closer look at the plastic on the package makes it look pretty grotty, and it wasn’t entirely clear what the part number is. After a bit of squinting, I think I made out APM3030P[x] where [x] represents one unknown letter.
A bit of digging seems to suggest it’s an Anpec part, but there’s no direct data for it. A replacement is available in the form of an STMicroelectronics STD30PF03LT4 P-channel 30v 0.028ohm 24A DPAK StripFETII Power MOSFET.
In an ideal world, I would just go and grab some of these ST packages and call it a day. Unfortunately, the ST devices weren’t available from my preferred supplier, element14. Instead, I had to go hunting for a replacement with similar specifications. Of the products they had in stock for Australia, I decided the Fairchild FDD4685 would be a good candidate, with 40V, 32A, 0.027ohm rating, at a cost of AU$2.83 each excl. GST.
For good measure, I’ll also get rid of the G-Luxon capacitor, replacing it with something more reputable. I opted to put in a Rubycon 1000uF 16v ZLJ capacitor, at a cost of AU$0.34 each excl. GST. It is rated for 10,000h service life at 105 degrees C, which should help it last in such demanding environments.
Total cost of parts, AU$3.49 including GST. With my Tenma hot air rework station, I set to work desoldering and replacing the cap and MOSFET with no drama. I tested the G-Luxon capacitor, and to my surprise, found it to be good for capacitance and ESR! I suppose if you’re doing the repair yourself, you probably won’t need to replace the capacitor after all. After the repair, it looks like this:
Part of the cause of failure may have been due to sustained overheating or thermal cycling. After removing the bad MOSFET, it was noted that the red adhesive used to tack down components in an SMD process was actually covering some of the heatsink pad area which would have increased thermal resistance. By cleaning that up, and tinning the pad, we can ensure a better thermal conductivity. More solder was also added to the joint to improve current carrying and heat transportation as well. It might fail again in the future, but we can try to give it a little help …
After reassembling the unit, I tested it by charging up 4 sets of 4 AA’s. To my relief, the unit fired up and worked like new again! If it didn’t, my next suspicion would have been the two freewheel rectification diodes mounted just below the capacitor, which would probably be a few more dollars of parts to be replaced.
While I haven’t tracked down the ultimate cause of the failure, it might just be down to a lack of heatsink and the cumulative effect of overheating over time, or an unlucky component. As the design is dictated by the unit, there’s not much that can be done, but spending more for a higher specification MOSFET with lower RDSon might help reduce heat dissipation.
It’s rather lucky that nothing else was damaged, and the repair itself proved to be cheap (AU$3.49 of parts). Unfortunately, if you don’t have a hot air rework station or electronics experience, diagnosing such failures and repairing them is often out of reach, which is unfortunate. But if you have one of these chargers, and it’s dead … you might have the same problem. If you have the tools and experience, now you might be able to fix it too!