Innovations Battery Manager Ultra
For well over ten years, I've been using a Battery Manager Ultra battery charger which was bought from the now defunct company "Innovations" in the UK. Many people have reported bad experiences of this charger on the web, and in fact bad reviews are almost all you find if you search for the name of the charger. However, my experience has been good. I think it's a matter of setup at manufacture time and whether customers were lucky to get a well adjusted charger.
The charger claims to be able to rejuvenate Alkaline cells as well as correctly charge both NiCd and NiMH cells. I've occasionally attempted to charge Alkalines, and this sort of works. However, the cells aren't designed to be recharged and they can leak afterwards. On the other hand, I've found this charger works well with NiCd and NiMH cells, perhaps surprising given how many negative reviews there are elsewhere on the web.
Reducing power consumption
One aspect of the workings of the charger has never impressed me. Like many consumer products, it's not particularly power efficient. The main cause of the problem is the rudimentary power supply. A plug in "wall wart" transformer produces about 12 V DC, and this runs through a 7805 regulator inside the unit to produce 5 V DC which is used to charge the ~= 1.4 V cells in the charger. i.e. out of 12 V, 10.6 V are wasted as heat. To express this in another way, the device is about 11% efficient at best. My "wall wart" actually produces about 13 V (with the 220 V electricity in the Netherlands, more with the 240 V in the UK) and so actually even more power is wasted in heating up the regulator chip.
Car style USB adaptor
I wanted to plug the charger into the 12 V solar charged power system in the garage, but I didn't want to see most of the power generated by my small solar cells going into heating. Therefore I wanted to improve this efficiency. The easiest way to do this was to replace the 7805 linear regulator with a switched mode power supply. This should more than half the power consumption of the unit as a whole because even with a low (for switched mode supply) efficiency of 80%, there would be only half the energy wasted as heat as with the 7805 which is only 40% efficiency with such a high input voltage (this is linear with the voltage drop. A low dropout regulator with a small voltage between input and output can actually outperform some switch mode supplies in some circumstances, but not these circumstances).
There's space for the adaptor inside
The cheapest way of buying a 5 V switched mode supply these days is to buy a cheap car to USB adaptor. I took one of these apart and soldered on wires to reach the pins of the board which took the 7805 inside the charger. Needing somewhere to mount the USB adaptor, I found that happily there is enough space in the front of the Battery Manager Ultra to allow it to poke through the side, so my battery charger can now also charge USB devices.
Unfortunately, after I modified the charger to use the switched mode supply, I found that it would no longer charge NiMH cells correctly. Indeed, it started to insist that all my cells were faulty, and leaving them in the charger for any period of time never resulted in a good life from them. I had transformed my previously working battery charger into the hopeless device of several other peoples' reviews !
This required some investigation. Clearly I had "broken" something. It turns out that the problem was in two parts.
This is the only adjustment possible
The first problem was easy to spot and also very simple to fix. My new power supply produced somewhat over 5 V. In fact, my Chinese SMPSU produces about 5.2 V. On the other hand, my multimeter showed that the original 7805 produced about 4.95 V in the circuit. On the small PCB which used to hold the 7805 you can also find the only adjustable part in the circuitry of the charger. This potentiometer forms part of a voltage divider across the 5 V supply and provides a reference voltage to the main PCB. Before I changed the power supply, this was producing a reference voltage of about 3.95 V, but afterwards the reference was 4.16 V.
I had noticed that the readings for cell voltage on the front panel of the charger were no longer correct when compared with the voltage which my multimeter reported. By adjusting the potentiometer I could adjust the display of the charger so that it exactly matches the display of my multimeter, which meant the reference was back around 3.95 V, and this gives the best charging behaviour. It is quite likely that other chargers which people find don't work have not been adjusted correctly.
At first, I thought this had completely fixed the charger...
After using the charger for a while with the reference voltage problem fixed, I noticed another problem: it would often reject known good batteries as faulty. In fact, it did this more often than it charged them correctly. What's more, a new odd behaviour was seen on the display: the voltage displayed was not stable, and even when there was no battery in a battery holder the charger would occasionally display a voltage for the empty space.
This problem was caused by the replacement SMPSU not producing such clean power as the simple 7805 regulator. High frequency noise was superimposed on the supply line. A simple solution to noise is a low pass filter, and this is the approach I took. From my electronics "junk" box came a bunch of resistors which paralleled to 3.2 Ohms and a 400uF capacitor. These now form a low pass filter between the output of the power supply and the charger itself, with a pole of 124 Hz. It's enough to reduce high frequency spuriae so that they don't affect the charger circuit. There is also a secondary effect of reducing the effectiveness of the regulation, and this results in the front panel green LED blinking very slightly as the charger operates.
Everything working again
The charger now charges as well as it did when new, but consumes considerably less electricity as it does so. It also has the useful extra facility of a USB socket on the side for charging modern devices.
This can be used at the same time as NiMH cells are charged.
After a few years the cheapo USB adaptor failed. I still use the charger but now the 5 V supply is from a plug-in mobile phone charger wired directly the board insisde the unit. This is a better solution than the 12 V to 5 V conversion using the USB gadget.