A short while back, I posted about this generic battery tester, which was a cheap and nasty clone of a model which was quite popular, which I used frequently as a child. In my quest to buy it, I came across another model which seems to be widely available and inexpensive – the “generic” BT-168 that costs about AU$2.46. Unfortunately, while I ordered both units simultaneously, this one got lost in the post (or so it seems) and I ended up receiving a refund for it. Because I wanted to see what was inside, I decided to order it again from another seller, and this time it managed to arrive.
The unit looked like the image in the listing, save for some slightly “bleeded” printing and a scratch in the window of the meter itself. The meter window protrudes out from the surface of the meter and has a fairly “sharp” edge, making this unit look and feel a bit odd.
Initially, judging from the way the meter looked, I assumed that the red sliding positive terminal would be spring loaded and automatically “clamp” the battery in a way similar to some battery chargers. Unfortunately, this is not the case, and the slide is a manual adjustment requiring some finger pressure during testing to maintain a good contact. Test contacts for 9V cells are located on the side, and all contacts are made from very thin sheet metal which is not too sturdy or well formed resulting in some residual bow. This is similar to the other generic tester’s terminals.
The unit is assembled with two screws, and has a moulded table in the rear showing the range of voltages that are considered good, low and replace, not that this is highly useful information on its own as you will see later.
Two screws is all it takes to get into the case, however, it should be noted that both screws come pre-rusted, so I suppose maybe they’re not very well made.
Rather annoyingly, the unit is mostly empty space, and features four resistors as in the other generic unit. There are some differences though – namely that the resistors are all 1% metal film 5-band code resistors, so they should be able to handle more power dissipation, and that these resistors are “air-wired” whereas the other SMD resistors necessitated using a PCB.
The meter sensitivity/resistance isn’t known directly as it is unmarked. As I didn’t want to desolder anything, I got a little lazy and instead worked it out as follows:
- With no battery connected, the meter has two parallel sets of equivalent resistances across it, namely 3.9 || 1.5k and 220 || 9.1k. Resolving these into one resistance over the meter, this should represent an equivalent resistance of 1503.9 || 9320 = 1294.944336 ohms.
- Measuring across the meter terminals, a resistance of 488.8 ohms was measured.
- Solving for x in 1294.944336 || x gives me:
1294.944336 * x / 1294.944336 + x = 488.8
1294.944336 * x = 488.8 * 1294.944336 + 488.8 * x
1294.944336 * x – 488.8 * x = 488.8 * 1294.944336
806.144336 * x = 488.8 * 1294.944336
x = 488.8 * 1294.944336 / 806.14336
= 785.181424 … ohms.
The design is virtually the same as the previous generic tester I looked at, but the resistances are quire different. The meter has a higher resistance, and the load resistances are lower. In the previous computations, I only assumed that the load resistor is the only load across the battery. In this teardown, I decided to factor in the influences of the meter and other resistances. As it turns out, this just creates a mess and results in an equivalent resistance that’s not far from the load resistance itself (3.89 instead of 3.9 ohms, and 215 instead of 220 ohms).
The load that this tester puts on the batteries is much higher than the other tester, drawing about 385mA from a 1.5v battery and 41.8mA from a 9v battery. The other only draws 56mA and 13mA respectively. This higher load will bleed off surface charges more quickly, and give a more indicative reading sooner than the other meter. However, because a moderately high load is being put onto the cells, it is likely to report premature low-indications for batteries which may still be serviceable under lower loads (e.g. batteries with high internal resistance). This is why the cell voltage is not a really useful piece of information on its own – it may be surface charge that is drawn down quickly, and it will vary significantly as a function of load due to the internal resistance of the cell. If the load isn’t specified, a voltage alone doesn’t precisely indicate state of charge.
There really isn’t anything to using one of these – it’s “intuitive”.
However, that being said … the results you get from these units can vary significantly.
The same battery is being tested in both pictures – the battery is flipped over for the other unit due to the order of the terminals. One claims that it’s very much good (at the top of the scale) whereas the other is saying a borderline good-low. Ironically, it’s the one that places more load on the battery that’s saying it’s very good, so what’s happening?
Well, as it turns out, the load is only one part of the test equation. Of course, a higher load will draw down the battery voltage more and a good cell would be expected to have a lower voltage under a higher load as compared to a lower load. But the part that we’re missing is the meter sensitivity. The d’Arsonval movement is essentially acting as an not-very-calibrated voltmeter. Because this particular battery tester has a more resistive coil (higher sensitivity), assuming the same meter spring return force, I would presume that this higher resistance coil would move further for the same voltage applied, as the voltage divider it forms would have more voltage across the coil than the other meter. This is helped by the fact that the voltage divider resistors with the meter in this meter have a lower value, meaning that the meter receives more voltage as well.
The measurement of batteries is a rather complicated science, and such cheap battery testers are widely available and used to give a crude indication of remaining battery charge. This meter shares its configuration with the other examined generic meter, although the values are significantly different, thus resulting in this meter putting a larger load on the batteries, but also being more sensitive to voltage. As a result, the indications you get from the two testers can, at times, differ significantly.