A little while ago, I’d borrowed a Hilti TE 6-A cordless rotary hammer drill from a friend. He’d warned me that the batteries were getting very old, and barely held any charge, to the point where I might be able to drill only 2-3 holes before the battery died. So I made sure to fully charge each of his three battery packs before use, but found that one of the batteries wouldn’t accept a charge. That’s a problem. However, these being NiCd batteries, which are far simpler than lithium, it was very likely that fixing the battery would be a straightforward exercise, requiring at most a replacement of all the cells in the battery pack. So, after obtaining the owner’s consent, I started pulling things apart.
Before beginning, it was evident that the drill accepted either of two slightly different battery pack designs. The older of the two, the pack which had failed, had a square protrusion at its base, while the newer packs did not. When opened, the older pack had a circuit board and relay which was being accommodated within this protrusion in the housing. The newer packs lacked this circuit board and relay, but had a 50A fuse which the old pack did not. So a reasonable assumption is that the extra circuit board in the older pack would detect short circuits and other overcurrent faults, and used the relay to break the circuit. This may have been replaced with the fuse in order to reduce costs, or to improve reliability, since the relay solution could have been unnecessarily complex, and a failure of the relay could meant the battery pack would have stopped working if the relay failed open, or would have lost its overcurrent protection if the relay failed closed. Relays being electro-mechanical devices, they are subject to wear and do have a finite lifetime.
Now to the actual cause of the failure, which could have been either a problem with the relay or its circuitry, or with the battery cells themselves. The battery did not appear to provide any voltage even when the relay was bypassed, suggesting that the problem was with the cells themselves. It may have also been that the relay circuit somehow shorted out the battery and completely drained the cells, or perhaps they had just been left for so long without charging that they self-discharged all the way down to zero volts, or had developed an internal short due to their age or excessive discharge. (NiCd batteries can be damaged by discharging below 1V per cell, and so with this battery being made up of 30 cells in series, anything below 30V would be a problem.) However, the cause of the failure is not especially relevant. Even before failure, the battery held such little charge that a replacement of all cells is justifiable, and the simplest way to do that would be to discard the relay circuitry, and create another battery pack mimicking the new battery pack design, using a fuse instead of the relay circuitry, and re-using only the thermistor, housing, and terminals from the old pack.
[Note: Pin numbering has been chosen such that the plastic divider is located between pins 2 and 3. See image above.]
From visual inspection, pins 4 and 5 were connected to the positive terminal of the battery pack, pin 2 to negative via the fuse, pin 1 to pin 2 via the thermistor, and pin 3 was wired through the middle of the pack to what appeared to be the pack’s half voltage point. However, the voltage measured from pin 2 to pin 3 was about 0.32V lower than that measured from the negative terminal of the pack to the half voltage point, so something else was going on, and that voltage drop suggests a diode. Cutting and pulling the pin 3 wire, and removing some heat-shrink revealed a hidden 1N4148 diode and a 9.1kΩ resistor.
The thermistor had a resistance of 6.3kΩ at around 20°, increasing to 11kΩ after having been left in the fridge for a few minutes. So the thermistor can be assumed to be a 4.7kΩ NTC. (Thermistors are rated by their resistance at 25°C, and NTC thermistors increase their resistance as temperature increases.) The fuse was easily identifiable, being marked “50A”. Additionally, the wires connecting the positive and negative terminals to pins 2, 4 and 5 were 14 AWG, which is good for about 24A of current before it starts getting warm, while other connections just used light duty hook-up wire, suggesting they would not be carrying significant current.
The original pack was made up of 30 Sub-C NiCd cells connected in series. This is a common size for cordless tool battery packs, and I was able to find a pre-made 30 NiMH cell replacement designed to fit within a Hilti battery housing. Using NiMH rather than NiCd for the replacement means the battery will have a much greater capacity, and will also use far fewer toxic materials. The Hilti battery charger (model C 7/36-ACS) can handle both NiCD and NiMH batteries, so it will still be compatible.
Assembling the new pack was straightforward, with the only additional components required, aside from the batteries themselves, being the fuse, resistor and diode. (The thermistor was reused from the old pack.) An automotive blade fuse was chosen for its form factor and availability, and the resistor and diode are both extremely common and readily available components. The resistor and diode were heat-shrinked and hidden in the middle of the pack, as with the original pack. 14AWG wire was used for the current-carrying connections to battery positive and negative, and light hook-up wire (25AWG) was used for pin 3 and the thermistor.
Once the new battery pack was assembled, the drill functioned perfectly with it. We are yet to determine the lifetime of the new battery, as we are yet to drill enough holes in a day to exhaust it!