If there’s one thing I always need to do, that is to clean up and get rid of the rubbish. Almost constantly, broken things are taken apart and examined in my room, with some of them requiring only minor surgery (e.g. re-soldering, single component replacement) before heading back out. But some other things don’t fare so well and will sit around waiting for parts … only to find things didn’t go so well. Other things get donated to me just so they could be torn apart, whereas some things are just not worth fixing.
Regardless, this post focuses on three items which I’ve attributed the [WONTFIX] label to, in true bug-tracker fashion, because it’s either not worth fixing, or I just can’t fix it after all. Still a good chance to look inside some of the items if nothing else.
First of all, this device was donated to me for a teardown, so that’s probably what it’s going to get in the end. There’s no way I’d actually ever own one of these – I have nothing that uses the Lightning port myself. Anyhow, it appears to be a microSD card reader that can be used with a computer via the USB port, and with an Apple mobile device through the Lightning port to give it some additional storage and something akin to a filesystem. However, as this was a “third party” accessory, it seems likely that it succumbed to Apple’s software updates which increased the DRM security surrounding Lightning accessories, and it’s also likely that the app that one supported the device has since disappeared as it wasn’t updated for the latest version of iOS. I’m not sure, but it could be a clone of a legitimate device as well. Whatever it is, it’s got very little value to me … so lets play with it.
Plugging it in via USB, it seems to function as a “CARDER READER” with a VID of 2008, and PID of 2009 – very suspicious. A search seems to suggest MXT6208 or MW6208 chipset, and possible relation to Chipsbank as well, but nothing particularly conclusive either way.
Would you actually keep this thing just to read microSD cards? Hell no. With a Sandisk Extreme microSDXC card capable of 45MB/s read, it delivered a paltry 7.6MB/s. Nowhere near the ~28MB/s USB 2.0 is capable of, and nowhere near the ~23MB/s even a basic 25Mhz SD bus is capable of. Save your time and use something decent.
As a result, it was a no-brainer to get out the snips and eat my way through the plastic casing to reveal the insides:
Nothing too interesting on the top side, but it seems a fairly recent device – date code on the PCB of Week 43 of 2015, I presume. Seems like footprint for two different microSD card slots, or maybe two mounting positions for the one slot (one further inset) to accommodate a different shell.
The underside has a chip which has been remarked as 75089 with a date code of Week 47 of 2015. This is quite probably a USB card reader IC of sorts, with a custom firmware (?) or data stored on the flash chip above. There is an Atmel TINY13A microcontroller, which might be used to negotiate over the Lightning connector. A final chip marked 338 3959 was not identified.
Curious as to what’s on the flash chip? I was, so I dumped it. Most of the end is empty, but taking a quick look at some strings and there are a few interesting results:
The first snippet I encountered speaks of an Apple_AuthenticationCertificate. It also has the name of Universal V1.0 2015-12-16.
The second snippet seems to relate to SD command error handling/reporting. It also seems to have a (possibly Unicode) string of Generic USB 2.0 Device 20120218120009. Another string is iPODCEC*com.iStick.Protocol1 iAP Interface001. This might be a clue as to the app it was used with, or the manufacturer/designer.
The final set seems to be a fixed storage area for strings. Oh well, at least I got to peek inside.
Varta Power Play 15 Minute Charger Power Supply (SMP-75W)
Remember back in 2014 when I revived my Varta 15 minute charger by changing out a MOSFET? Well, it was doing well until just recently, it quit working again. This time, the fault was not in the charger unit – instead the power supply just refused to give any output.
If you thought I gave up on this unit, you’d be wrong. While I do have a newer charger (which I never had the time to review), I don’t like it as much and it has a habit of boiling my cells. As a result, I did my best to troubleshoot the unit – to do this required opening it up (of course).
Two security screws later, and we are in …
As expected from a quality power supply, it is covered with insulation tape and heatsinks. On the whole, I didn’t expect anything major had gone wrong – there was no smell, no char marks, and even on the power analyzer, it drew about 200-300mW when idle. Something was happening, and it definitely didn’t blow a fuse.
Peering in from the sides, nothing seemed to be hurt.
Looking from the underside, there was some flux residue, but no obviously bad joints or burnt traces. There wasn’t even any signs of heat stress. The brains of the power supply was mounted on the underside.
As working with mains switchmode supplies when live can be dangerous, and even after they’re unplugged if the capacitors are holding a charge, I first tried to diagnose the components. Probing about, the fuses were intact, the bridge rectifier was good, the output leads showed continuity, the thermistor and inductors wasn’t open, the connector was clean, the mains plug and incoming wires were good. Even a quick check showed that the windings seemed to be intact. I scratched my head, and thought it was probably a semiconductor fault of some sort.
Frustrated of probing in the dark, I decided to energize the unit and do limited probing while running on a power-limited inverter. Ultimately, the primary side DC capacitor was holding at the peak of the mains sine wave, suggesting we were good up till there. Measuring pins on IC1 randomly, I found a pin holding a steady 5v, suggesting bootstrap supply was coming in. It seemed likely that the primary side transistor may have failed open, or feedback from the secondary was being lost. As I didn’t have the tools to test the optocoupler, I decided to purchase a few spare general purpose units of equivalent spec and just “change it anyway” because it was cheap.
A quick desolder and re-solder later, the damn thing still didn’t work. By now, I was frustrated, so I thought I might tackle the primary switching MOSFET. However, due to the way it was built, it could not be deconstructed in a harm-free way, and by the time I had it in bits, I wasn’t confident it was going to go back together in a safe way (and plus, the MOSFET used was not a part I had easy access to). As a result, I decided I had enough and just tossed it out. Now, I just run the charger unit from my bench power supply … so all is not lost.
Powertraveller PowerMonkey Discovery 3500mAh Power Bank
This was a fairly expensive and basic power bank which I had obtained owing to having some gift voucher awarded to me that needed to be spent. It’s been a good unit for manual operation and low ripple, but suffers from a crippling 500mA output current rating – not enough for modern devices.
Unfortunately for me, I had connected it to a device that expected 2-3A of input, and was happy to try drawing it regardless of what the voltage was. Without so much as any sign of stress, the PowerMonkey ceased to function and became entirely dead. Oh well.
The unit itself is built inside a metal enclosure with a metal trim plate on the front which is self-adhered. Peeling the trim off reveals a clear plastic window – I would have much preferred to see the power bank like this, if I had known. Four screws are removed to extract the unit from its shell.
Quite a neat construction, utilizing two PCBs joined by a pair of wires.
As promised, the cell is a YOKU branded 3500mAh Li-Poly pack. Both PCBs are dated 18th June 2012 with a project name of 1028 Ver 1.2. The output side PCB contains the switching converter components including a chunky diode and ceramic SMD multilayer caps on the reverse. Quick probing seemed to reveal the diode was still intact, so probably the internal MOSFET within U3 had failed dead short, which itself may have tripped the Li-Ion protection circuitry. I noted that even applying 5V to charge the unit did not “wake it” and neither did disconnecting/reconnecting the cell.
The LEDs and charging seemed to be controlled by a Holtek MCU, as is common in so many cheap powerbanks from China. There is a DW01 lithium-ion protection IC, and a few transistors scattered about, as well as a zener diode (for protection?). There also seems to be a charge controller (U1).
Given the limited reward of repairing this unit – namely the limited capacity, 500mA output, and the barrel-jack charging input that is NOT a USB connector, I felt that it would just be best to scavenge the battery and junk the rest.
For every success story, there are always many untold failures and maybe even defeats where I have just simply given in and given up. Nothing wrong with that though – at least I tried, and tried to make the most of the experience.
When it’s a piece of equipment you really like, a failure sometimes feels a bit hollow and disappointing – must be the sort of feeling surgeons get when they fail to save a patient. The regret of not having done better lingers over you for a while … but eventually, you just chalk it up to experience and actually do better in the future. Way better than not trying.
In that regard, at least I’m not blindly throwing things into the bin once they stop working … some products become good “
organ parts donors”, which means free repairs in the future or parts for the junk box.