Ansmann POWERline 4 repair

Repairing a microprocessor controlled NiCd/NiMH charger

Similar projects worth following
Pretty standard stuff. Actually I *should* have taken some pictures, since now that it's together I don't really want to take it apart just for pictures. A fairly popular microcontroller ran AA/AAA batter charger. Made in Germany by Ansmann. That's supposed to mean quality, and expensive, and it did last me maybe close to 15 years until one day all of a sudden it let out a loud bang and a foul, acrid smell. Indeed looking at the prices of new ones I decided I'd rather try my hand at repairing the one I had, I mean a blown capacitor, how bad can that be?

So, after picking up the soldering iron and swapping in an "emergency capacitor" that didn't quite fit in the case, I hooked it up to the power outlet (more on that later), and... Nothing. A-ha! Of course the fuse would have blown I figured, checked it out, and it was indeed dead. This was a plastic 800ma fuse, and getting an exact replacement took forever, so I swapped in a 500ma glass fuse from another failed power supply.

You didn't expect that would work, did you? Yeah, wouldn't be reading the story here if it had! But no, the fuse didn't blow again. So a word about my setup: For testing, I'm using an old UPS as a power-source. Current limit and isolation (as long as it's unhooked from mains!) built into one. It's not perfect, for example not sinuisoidal, but it's good enough for most purposes. And I now have bunch of spare 800mA subminiature TR5 slow blow fuses (just in case you were wondering what kind of spare parts get...)

So no blown fuses, but I did get to observe that it repeatedly triggered the current limit on my 225W UPS, so I could see it was taking far more current than it should have, and I was dealing with something else than straight out dried and blown up mains condensator with no other damage.

I confess, I should have given up here, battery chargers are still a cheap lot, but having gotten this far I decided to just not give up easy and go all the way. I desoldered all the discrete components on the mains side and tested them up separately. The combined switching controller I couldn't easily test though, so it was first one to get replaced. I now have four (or five) spare TO-220 TOP223Y offline switching power IC's, but no, that wasn't it.

The photocoupler checked out, but just in case I got 10 pieces of PS2501-1 / NEC2501 DIP-4 optocouplers and replaced it. The switching circuit has a diode in it, which I couldn't identify, but it did check out (though unfortunately I forgot to record the junction voltage; the circuit isn't any of the TOP223Y switching IC application notes, but I assume the diode is pretty basic).

All the resistors were fine, none burned, and all the smaller caps were fine, even measured on an ESR meter. So, what was the problem? A shorted flyback transformer coil? A berserk microcontroller? Nah, none of those, I'd kind of forgotten of one component - because it was on the solder side of the board.

Sure enough, the DB107S / DF10S 1A 1000V SMD rectifier bridge on the solder side was bad, which was detectable even without de-soldering once I got around to remembering and suspecting about it. Took a while to arrive, but after replacement it's working good as new.

I'm posting this as a project mainly because though Googling brought up a lot of examples of people repairing the device, none seemed to suggest the rectifier bridge, and because the component list MIGHT help someone to figure out what spare parts to get. Apparently getting blown chargers off eBay and repairing them is a business for some; well at least I'd have the parts now :)

Link to a polish site with a few very high quality close-ups of the innards for verifying the specific model etc. They just have blown switch IC's and dried out caps. Many other Ansmann chargers have identical component choices, but again this discussion suggest to just replace the caps and switch IC, but I guess I'll be checking rectifier bridge on-circuit first on a blown fuse from now on.

  • 1 × TOPSwitch II TOP223Y(N) TO-220 Three Terminal Off-Line PWM Switch Power Integrations datasheet at - got some TOP223Y's off eBay, but TOP223YN should be RoHS compliant replacement still in production. This chip looks standard in many Ansmann chargers, though it doesn't look very common otherwise.
  • 1 × PS2501-1 NEC2501 DIP-4 Photocoupler Pretty standard optocoupler separating the line voltage and low voltage parts as you would find in the feedback circuit of most switch mode power supplies. The RoHS part is PS2501-1
  • 1 × 10uF 450v 105°C Aluminium Electrolytic Capacitor Line capacitor, holding the rectified line voltage. This is for a 230V model. I went with a 33uF model Rubycon WXA 18X25mm piece, which is physically the largest that will fit with some careful fitting, both for reasons of availability and durability.
  • 1 × 47uF 25V 105°C Aluminium Electrolytic Capacitor In my case the auto restart/control pin bypass capacitor appears to have gone bad, though it wasn't outwardly visible. Best to replace this one with identical capacitance, as it's used to set auto-restart timing and control loop compensation.
  • 1 × 372 Series, TR5®, Time-Lag Fuse 800mA 250V, 50A Breaking Capacity Littlefuse - see text for details.

View all 8 components

  • Finish

    BadgerBadgerMushroom04/09/2015 at 19:07 0 comments

    Having got the charger working with the earlier listed replacements, I decided to open it up for one (hopefully!) final time to get some of the pictures I attached in this project. As well, since I had it open, it was a good opportunity to replace rest of the capacitors with new ones. The backside of the charger has printing "200134" which I take to mean it was manufactured in 34th week of 2001, indicating the capacitors could be almost 15 years old - if true, it's held up pretty well, despite there being much critique about Ansmann quality over the web (perhaps it only took a turn for the worse after).

    Not to mention I'm not really sure what was the triggering fault in this one. Could have been one of the primary side caps drying out, but could just as well have been one of the diodes breaking due to power surge or similar. Regardless, where there's a bad cap there's usually several, so I set out to replace all of them, and added the electrolytic capacitors inside into the component list. I didn't want to make more substitutions (besides the line capacitor), so I put in another order for missing capacitors from eBay, and settled in for a long wait...

    And that's how I got ripped off by the Chinese on eBay for the first time :) I had put out an order to happy-shop09 (Though they're shipping in Newfrog gadgets envelopes) for a capacitor assortment supposed to have up to 10 pieces of 25 capacitors each. After about a months wait it arrived, only for me to discover about one third of the capacitors are 0.1uF 50V pieces. In particular, low voltage & high capacitance pieces were all missing. If I had been shipping just for filling for junk-box that might've been acceptable, but since I was looking for the specific values listed, and getting new ones from eBay will be another month...

    Since I want to get this one out of the table and move on to the other projects I have open, I replaced the missing 16V220uF capacitor with a 25V220uF capacitor that I had laying around. Getting it to fit required bending some of the components out of the way, but no biggie. And the replacement capacitor had an effective series resistance of about 0.44 ohms, which is definitely too high for a 25V capacitor (though datasheet seems to specify up to 1.2 ohms for these KMG miniature caps!). On the other hand it replaced a 16V220uF cap with 0.58 ohms series resistance, and as said is larger so will dissipate more power and hopefully not heat as much, so maybe it's an improvement nonetheless.

    The 35V220uF capacitor measured 0.23 ohms, and was replaced with equal piece with 0.08 ohms series resistance, and 10V1000uF measured 0.16 ohms, replaced with a new 0.05 ohms capacitor of same values. All 105 degreess C of course. My worst case ESR table gives about two times the ESR of the replacement caps for those pieces, these measuring over triple the replacement caps were definitely due to be replaced even if they hadn't totally failed yet. For the replacement pieces I measured ESR direct without use/pre-conditioning, it might have changed a bit from long shelving time.

    As a final note, and to get to the "hacking" in the hack-a-day, I had to hack a piece of plastic off the separating lip in the cover piece to fit in the physically larger line capacitor (450V10uF). If ever having to do this, be very, very careful as it's still necessary to maintain separation and safety distance between the line voltage and low voltage sides, and that plastic lip serves to enforce it. As you may be able to see from the pics, I only cut little off the corner to get the cap just fit. The insulation paper wrapping the line voltage side was originally doubled over under the circuit board, while some images and layout suggest the flap in it is meant to slip through the matching slit in the circuit board to add to that isolation so I bent it up through the slit as well.

    And that should finish this repair, hopefully for another 15 years or so :)

  • Component images

    BadgerBadgerMushroom03/07/2015 at 21:04 0 comments

    On the first pic is an assortment of some of the parts involved. Despite the loud report of the power supply's failure, the hole at the middle of the cap (upright position near middle) is barely visible and easy to miss, although the slight outward bend of the top plate is more reliable indicator of a failed cap. After removal, from the leaked and burned electrolyte at the bottom in the second picture, the failure becomes much more obvious. It's a 10uF 450v 105C piece, about 21x13 millimeters.

    Finding a suitable replacement part is often an artform of its own, because people don't usually keep components of every possible value and size hanging around. On the far right is the first cap I tried, which I had to solder dangerous extra leads to because it was snap-in type and so big it wouldn't fit in the original location at all. The 33uF one at the bottom is the replacement I settled on in the end, in part to avoid another blown cap, although it required cutting a bit of plastic off the insides of the cover.

    At the top is also visible another nearby capacitor in the SMPS section (47uF 25V 105C); it doesn't even read on my effective series resistance meter, suggesting it's totally dried out, although there's no outward sign of failure. I should probably go through every cap on the low voltage side to make sure they haven't failed as well. The feedback opto-isolator didn't show any outward damage and the diodes checked out fine, but I replaced it just in case because I didn't want to set up a test-bed to see if it worked. For the actual switching IC there were no indications of how it should measure if it's intact, but Google searches suggested it would be almost always bad so I replaced it as well. One with short leads on the left is the original, one on right is one of the replacement parts.

    The blown fuse was a matter of its own. Being enclosed, there was no indication of failure on the outside, but of course being it's a fuse, it's easy to measure even in-circuit. It wasn't until I cut it open that I realized it should probably normally be possible to screw the top open (though I wouldn't recommend this as it might influence its integrity). As can be seen, this one is quite clearly blown.

    I already threw out the bad rectified bridge, but as usual, there was no outwards sign it was bad either, but again due to the nature of it's function, it should check out rather easily even in-circuit. The schematic I've added is from the TOPSwitch II switching IC datasheet, it's not exactly same as the batter charger, but close enough.

    It's probably worth reminding, especially if there's a fault in the circuitry, those capacitors can retain charge far after the power has been disconnected, and need to be bled empty through a power resistor before touching anything. Incidentally, with a shorted rectifier bridge like this, it seems like you could get a shock from the power-plug if the cap and fuse are replaced!

View all 2 project logs

  • 1
    Step 1

    I'm using the Ansmann POWERline 4 as example, but most of this is applicable to any switch mode power supply repair. I promised to make checking the rectifier bridge my new first step in case of blown fuse, that much is true. However, there's one thing even more important than that. I'm sure you're all tired of safety warnings/disclaimers by now, but considering there's no telling who are reading this, it should be stressed repair of line/mains voltage equipment should only be attempted by trained professionals.

    If you nevertheless decide to try it, be aware that even when disconnected from mains or protected by ground fault circuit, the mains capacitor in even a small switch mode power supply can hold enough juice to potentially kick your heat into ventricular fibrillation. In such a case there's 4 minutes of time to begin CPR or defibrillate before permant brain damage and eventual death begins to occur. So, either make sure there is someone trained and ready to perform CPR at hand, or avoid getting shocked in the first place (Better yet, both).

    To this end, be aware of the danger already when opening the case of a mains appliance, using only a well insulated screwdriver, or better yet a plastic edge with finger-guards to pry open the tabs holding the case together. Using a plastic wedge will also minimize damage to the case. In this charger, as can be seen in some of the pictures, the tabs holding it together are at the ends of the case, though sufficiently strong twist along the long sides will crack it open just as well. Working from the low-voltage "bottom" side will be safest and avoid accidentally pulling the wires from the circuit to the plug off.

    After the case is open and carefully laid off the the side. Careful as the wires will wear down from tugging and bending; if you end up doing more work on the circuit board, it will save you some grief to de-solder the wires from the circuit board once you've ensured the capacitor is empty, then re-solder them the same way around after your work is done. Having to re-solder frazzled wire-ends or worse yet fix the connections on the plug end will be a pain, and cause unnecessary risks.

    Next use the insulated screwdriver, a wooden chopstick or what you have to lift the green insulating paper off the back of the circuit board and tape it out of the way. Use a multimeter to measure the capacitor to make sure it's empty. In this case the capacitor is on the as-of-yet hidden flip side of the circuit board, so if you can't without a shadow of doubt locate the capacitor's leads, measure over the + and - side of the rectifier bridge which should be connected to the capacitor. After measuring it empty, make it double-sure by shorting the leads of the capacitor with the insulated screwdriver (ONLY if the capacitor measures empty, otherwise damage to equipment may occur).

    The generic switch mode power supply FAQ presents many useful tips, including what to do if the capacitor is still holding a charge. NEVER work on the circuit before you've conclusively measured any big capacitors as empty AND double-checked by shorting them.

  • 2
    Step 2

    We still aren't getting to the actual rectifier bridge. If the fuse is still intact, then some less catastrophic failure has occurred. So I would (and did) check the fuse first. If the device is completely dead with no sounds or lights when it was plugged in, you have a good reason to expect the fuse to be blown. As noted in the pictures, this miniature fuse can apparently be screwed open, but that may affect its integrity, so better just measure across the logs.

    A fuse is designed to completely cut off the power in case of a short/overload, so once you've identified the fuse it's a simple matter of measuring continuity over its legs. It should either conduct or not. If it doesn't, then there's been some reason for the failure, and you should seek to repair that failure before changing the fuse and re-trying, or you'll just blow another fuse.

    The rectifier bridge is a similarly simple component in function, except it's made out of diodes. As the intent is to prevent reverse polarity voltage, conducting in reverse or not conducting in the expected direction should be sure signs of failure even in-circuit. Defining "reverse direction" requires the usual mental gymnastics. The leg marked with + is where the positive leg of the mains cap will rest, so this marking means "positive voltage will not pass back from this leg".

    Thus, with positive lead of multimeter at + marked leg, neither mains leg should conduct to it, and with negative lead at it both mains legs should conduct to it. With negative lead at - marked leg neither should conduct, and with positive lead at - again both need to conduct. With positive lead at - and negative at + it will also conduct, but with positive at + and negative at - it won't conduct, or otherwise the cap would be constantly shorted. The mains legs do not conduct to each other, again otherwise it'd blow fuse instantly.

    The diode forward voltage drop across each conducting path was just below 600mV for me on the replacement rectifier bridge, it can slightly vary depending on chip, but should be about same for each path (Except for - to + pin, as those have 2x2 diodes between them). There may be some transient readings from the caps, but they should even out quickly.

  • 3
    Step 3

    Finding out if the TOPSwitch II on an Ansmann charger (or similar) needs replacement can be tricky; the suggested test bench for determining if it works isn't readily available, and there doesn't seem to be any sources on simple checks that could be performed. However, Googling up on other repairs indicates when the capacitors go bad, the TOPSwitchII IC almost always needs replacement as well.

    Since I do happen to have intact and one broken TOPSwitch II IC, though, I can tell that for my intact TOP223Y off-circuit the following holds: From source (positive, middle leg/plate when facing the printing on the component) to control (left leg) is 530 millivolts diode voltage drop, and from source to drain (right leg) is 520 millivolts. All other combinations are blocked on normal multimeter in diode test mode. The exact values will of course depend on calibration of the multimeter and this being an IC, possibly on its revision.

    For the IC removed from the blown power supply, source to control is 49 millivolts voltage drop and 26 ohm resistance in both directions, and blocked every other way. The new IC exhibits no resistance. Presumably the multimeter is reading on the body diode of a MOSFET (which conducts in the opposite direction from normal current flow) on both legs, so reading no connection, resistance, let alone conduction in reverse direction of the diode should be pretty surefire way to detect a blown IC.

    The other components on the board will, however, render in-circuit measurements dubious, so the IC needs to be isolated first in any case, at which point it may be easier to just replace it. Especially if everything else on board checks out, but it still won't work. With the replaced components (diodes and elco's blocking) the chip does measure as described for me even in-circuit, but with different multimeters and possibly damaged components this isn't certain.

View all 3 instructions

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates