Recharging Ni-Cd batteries for Tektronix THS 720 STD portable oscilloscope

Merry day after Christmas everyone! I hope everyone is having a good time and doing fun projects during holidays and vacation. I’m trying to write a blog post again, so today’s blog entry will be about some of my efforts to recharge a battery! Can’t be that hard, can it?

The Tektronix THS 720 is a portable and battery-operated 2 channel 100 MHz analog bandwidth, 500 MS/s oscilloscope with isolated BNC inputs and an integrated digital multimeter (DMM). It is a very useful tool for troubleshooting electronics or servicing (switch-mode) power supplies due to its isolated input channels. It can perform “floating” measurements and measure voltage differences safely on arbitrary potentials (similar to a handheld DMM). When testing switch-mode power supplies (SMPS, having a galvanic isolation and line voltages, e. g. 230 V) with an “ordinary” oscilloscope, keep in mind that the probe ground connector (GND) provides a direct and low impedance path to the oscilloscope chassis (mains earth referenced). This is potentially very dangerous, because if you connect your GND to a “floating” Device Under Test (DUT) at a higher voltage potential, the GND connector creates an electrical short between the DUT and the oscilloscope chassis. If the DUT and chassis are somehow connected through protective earth (PE), it can short out line voltages directly via your oscilloscope and perhaps ruin your day, too (danger of being electrocuted). In order to avoid these electrical safety dangers, the serviced DUT needs to be powered through an isolation transformer as a safety measure – or alternatively –  it has to be probed by either using a high-voltage differential probe (for example Tektronix P5205A) or via a battery-powered, isolated (“floating”) oscilloscope (e. g. Tektronix THS 700 Series or Fluke ScopeMeter). There is an excellent explanation video on this matter from Dave Jones of EEVblog on YouTube.

 

While powered on, the THS 700 Series oscilloscopes have quite a high current demand, which can be only supplied reliably by a Nickel-Cadmium (Ni-Cd) type of rechargeable batteries. Unfortunately, the Ni-Cd batteries have two disadvantages: they self-discharge over longer periods of time (at a rate of approx. 1% per day, it takes about 3 months for a complete discharge, inside of the scope they may discharge even faster within a couple of weeks) and they suffer to the memory effect (capacity deterioration due to charging/discharging cycles). I bought this oscilloscope for a very fair price (~200 EUR), however, with a dead Ni-Cd battery. The EEVblog users have investigated modern-era replacement possibilities such as NiMH batteries with special adapters which will fit inside of the oscilloscope battery compartment (length and diameter-wise). The alternative to this method is to buy certain Ni-Cd battery cells and combine them together into a battery pack.

I was lucky to find an offer on Kleinanzeigen from a guy who already premade this kind of battery pack and I bought it from him for about ~50 EUR. This saved me a lot of time and tinkering because I didn’t have a spot welding machine in order to attach metal sheets to the electrode. So here is a comparison between the original THS 700 Series oscilloscope batteries and the DIY-type.

 

Four C-type cells (Panasonic Cadnica, model N-3000CR, Ni-Cd 1.2 V, 3000 mAh, Flat Top (non extending), diameter 26 mm, single cell length 50 mm, ~7 EUR/piece) were used and contacted in series and assembled with Kapton tape so they form a single battery unit. Unfortunately, the THS 720 oscilloscope demands the positive electrode (cathode) to be placed at a certain position along the battery axis – you can’t simply use the both ends of the batteries unless you want to modify your scope for this purpose. In order to meet the requirements, the positive battery electrode needs to be extended by spot-welding a thin metal sheet and connect it to a metal ring, located about 40 mm above the negative electrode (anode) of the bottom cell. The metal ring provides the positive battery supply connection to the oscilloscope. This is a very odd construction but that’s how the oscilloscope was made back in the mid 1990s. It’s important to notice that the metal sheets must be spot-welded to the battery electrodes in order to provide reliable electrical connections. Leads cannot be soldered to the battery electrodes directly for different reasons – if they come off, they can cause power loss or even shorts, the heat applied during soldering can also damage the battery cell. The metal sheet needs also to be sufficiently isolated with suitable non-conducting tape in order to prevent (potentially dangerous) battery shorts.

 

Anyways, the Panasonic Cadnica replacement batteries turned out to work exceptionally well and provided enough power for a certain amount of time (depends on oscilloscope usage, perhaps 1 day with intermittent usage). I didn’t use the oscilloscope very much in the past and the battery was discharged as expected. In order to recharge the battery, I originally used the oscilloscope’s internal charging circuit, which “kinda” did the job. The charging at supplied voltage of 12 V and 1 A took around 24 hours. I noticed a significant heat-up of the battery packs up to ~45 °C, maybe 50 °C so I wasn’t quite sure about the unsupervised safety of this charging procedure. The max. charging temperature according to the datasheet is specified at +45 °C so I really hit the temperature limit. To avoid this in the future, I was looking for a fast Ni-Cd charger where I could quickly recharge the batteries without a significant heat build-up. There is a Tektronix THS7CHG Battery Charger on the 2nd hand market (e. g. eBay), however, it’s too rare and too expensive and not a viable option. An EEVblog user has constructed a similar charger with off-the-shelf components and 3D-printed housing. For a single Ni-Cd cell, the rapid charging time should be in the order of 1 … 2 hours at a nominal Voltage of 1.2 V and maximum currents of up to 4500 mA – compared to the 16-24 hours at standard charging currents (300 mA).

For this purpose, I bought a 2nd hand discontinued universal charger, which are very common in the fields of radio-controlled toys and RC models (drones, cars, boats etc.). It was manufactured by a German company called Robbe (type: Power Peak Infinity 2), which is capable of charging and discharging different types of batteries (NiMH, Lead Acid, Ni-Cd). In order to operate it, one needs a (switch-mode) power supply which provides the necessary voltages and currents for operation of the charger (e. g. 13.8 V and 0.1 … 5 A). The battery charger’s microcontroller monitors and regulates the output voltages and currents, which suits the charging profile of the batteries being recharged. In case of 4 Ni-Cd batteries, we need at least 4x 1.2 V = 4.8 V with a current limitation of 4.5 A (according to the Cadnica datasheet). The charging speed is limited by the battery temperature, the build-up of internal pressure and the so-called charge rate C. Typical charge rates are around 0.1 C while “fast” charge rates are around 0.5 C up to 1 C. Due to limitations of Ni-Cd batteries (memory effect, ~500 recharging cycles), they should be discharged first, then recharged at low rates (e. g. 0.1 C) to increase their longevity. For long-term storage of Ni-Cd batteries, Robbe user manual recommends to discharge them first and store them in a cold and dry place (e. g. in a fridge at 4 °C) in order to mitigate the performance deterioration due to the memory effect.

My charging/discharging setup looks as follows: the switch-mode power supply (JAMARA Germany DC Regulated Power Supply, Output: 13.8 V & 0-20 A) is connected to the charger. The charger output is then connected to the battery pack fixture. For my battery fixture, I used some spare optics parts I had at hand. The spring-loaded electrode helps to keep the battery in place and also helps to provide a reliable electrical contact. The recharger was powered on and set up to the “DISCHARGE -> CHARGE” mode, which is recommended by the Robbe manual. The battery under test had a remaining voltage of about 2.4 V and was discharged with a current of about 0.1 to 0.3 A. As soon as the battery was discharged to a certain threshold voltage (e. g. 1 V open circuit voltage), the charging cycle began by a slow ramp-up of the voltages and currents. After few minutes, the rapid charging is reached at approx. 6 V and 4.5 A (about 30 W), which corresponds to a charge rate of 1.5 C. I could observe some kind of a charge duty cycle, where the current flow stopped for a certain amount of time before continuing again – perhaps for thermal management purposes. The charging process took about 60 minutes, the estimated restored battery capacity (= transferred charge) was in the order of 3000 mAh. The temperature of the setup was monitored during the charging process with a thermal camera, however the battery pack didn’t heat up significantly in comparison to the mentioned THS 720 internal battery charger. The voltages were monitored by a digital multimeter HP 34401A and an oscilloscope Tektronix TDS 3064B, the current was monitored by a Tektronix TCP 202 current probe. This allowed me to estimate the power demand during the charging process.

I was surprised how well and how fast this recharging process went. The DIY battery pack proved to be suitable alternative to the original battery pack Tektronix THS7BAT. Unfortunately, the Ni-Cd batteries age and need to be replaced over time. However, an inexpensive alternative with modern-day parts for a reasonable amount of money (~30 … 40 EUR) can be used to replace the original battery pack and extend the lifetime of the oscilloscope usage, at least until the internal electronic parts start to fail (e. g. the opto-couplers of this scope).