ZH-YU ZB206+ Battery Capacity Tester and How It Measures Internal Resistance

Some time ago, while discussing another batch of “test subjects”—batteries—the topic of measuring internal resistance came up. Since the well-known YR1030 (and 1035) are relatively expensive, one commenter said there was a good, inexpensive little tester called the ZH-YU ZB206+.
But a few caveats also came to light, so to set the record straight, I bought one of these testers to try out.



The discussion then turned to the claim that the ZB206+ also measures at 1 kHz, as stated in the datasheets. To quote:


In my next review, I recommended this tester alongside the YR1030. But during the discussion, it turned out that unlike the 1030, it uses a more basic measurement method. So I bought the ZB206+ specifically to check this and compare its internal resistance measurement accuracy with the YR1030. Fortunately, it doesn't cost much.

This won't be a very long review. I'll try to keep things concise: description, operating modes, tests, and conclusions.

But since this is a review, I'll start as usual with the packaging.
The tester arrived in a regular plastic bag, wrapped in some kind of film.


The package includes the tester and an instruction manual.


Unfortunately, the manual is almost entirely in Chinese. You can either translate it with an app that can read text from photos, read my description below, or download the English version.

Specifications from the product page.
Supply voltage: DC12V / 5V (optional)
Operating current: <35 mA
Maximum battery input voltage: 8.5 V
Maximum discharge current error: 1% + 2 mA
Voltage measurement error: 1% -3D
Maximum error in overall measurement results: 0.1-0.2 A 2.5%, 0.3-0.5 A 1.6%, 0.6-1.0 A 1.2%, 1.1 A-2.6 A 1%
Circuit board dimensions (excluding copper standoffs): 97 x 62 x 38 mm
Weight (with copper feet): 47 g




It looks like a small, neat board. The heatsink did get a little damaged in transit, but it's nothing serious, just annoying.


The assembly looks quite neat to me.


The board has two connectors and one terminal block.
The terminal block is for connecting a battery or any other power source with a voltage of up to 8.5 V.
To the right is the connector for the measurement circuit, used with a four-wire connection. This is needed to measure internal resistance and capacity in watt-hours accurately.
Further to the right is a small board with a microUSB connector. The tester is sold in two versions, powered by 5 or 12 V. The 5 V version has the board shown in the photo; the 12 V version has a standard 5.5/2.1 mm connector instead. The price difference was negligible, so I got the version with the converter. You can always convert it to 12 V by simply replacing the board with a connector, since the board boosts the voltage from 5 to 12 V.


Controls
Three buttons: S-- (decrease), S++ (increase), SK (operating mode).

Display.
A four-digit, seven-segment display plus four additional LEDs.
The display cycles through current, capacity, and voltage readings. The corresponding LED lights up to show which value is currently on screen.
The display board can be detached from the main board, which is useful when installing it in an enclosure. The buttons aren't detachable, though, so you'll have to add some of your own. Still, it's convenient.

There's a buzzer below and to the left. It sounds unpleasant, so I disabled it almost immediately.


A power transistor is mounted on the heatsink. According to the manual, it's a P75NF75, but the marking on the transistor itself is barely visible.
The heatsink is very small. There's also a version with a fan, but I bought the cheapest one since the boards themselves are identical.


The main part of the measurement and control circuitry


The underside has only four standoffs for mounting the board in an enclosure.


1. 8S003F3P6 microcontroller with a 10-bit ADC.
2. LM358 op-amp. As far as I can tell, it amplifies the signal from the shunt and the input voltage.
3. Two 74HC595 shift registers, for controlling the display and the fan output.
4. A 50 mΩ current-sense shunt rated for up to 2 W, which gives plenty of headroom at 2.6 A.
5. Voltage converter board. The diode for reverse-polarity protection is visible to the left and below, though it's not really needed in this version.
6. TL431, used to build the power supply regulator for the display and, most likely, the op-amp.


The display has fairly good contrast, but it's easier to read with a filter, not to mention that it makes for better photos. So all the photos from here on will have one.
The board is set to a discharge current of 1 A by default.


The S-- and S++ buttons decrease or increase the load current setting. I’ll note that the buttons are arranged the way our users are used to: minus on the left and plus on the right. I’ll also point out that the store’s picture shows a current of 2.68 A, although the board can only set current from 0.1 to 2.6 A in 0.1 A increments.

If you start a discharge without a battery connected, an error message appears. There are 6 possible messages in total:
Err1 - Measured voltage is above 8.5 V
Err2 - Measured voltage is too low or absent.
Err3 - The battery has too much internal resistance or a poor connection, causing the voltage to drop too much under the set load current.
Err4 - Power transistor fault
Err5 - Power overload. By default, the board has a maximum power limit of 12 W, but this limit can be disabled
Err6 - Supply voltage is outside the 11–14 V range.



Settings.
To open this menu, power up the board while holding down the SK button. The -/+ buttons change the value, and SK moves to the next item. After going through all the items, the board switches to normal operation and saves the settings.
1, 2. - Two-wire or four-wire measurement mode.
3. Automatic battery type detection and corresponding automatic setting of the discharge cutoff voltage. I recommend turning this off, since the type is determined based on the battery voltage: for a charged Li-Ion battery, that’s 3 V; LiFe, 2.5 V; nickel batteries, 1 V. So the automatic detection can get it wrong, and sometimes you need to discharge to a different voltage anyway.
4. Disable the buzzer. I turned it off almost immediately because it was annoying.
5. Disable the maximum power dissipation limit.
6, 7. - The power in watts at which the fan turns on. At 00, it runs continuously during discharge.
8. The maximum setting is 10 W, but this mode is selected in a rather unusual way: not by stepping through 7-8-9, but in reverse, 3-2-1-10.


The internal resistance measurement mode is activated by powering up the board while holding down the S-- button. The display can show either the internal resistance in mΩ or the battery voltage; press SK to switch between them. The only way to exit this mode is to restart the board.


About the fan.
As you can imagine, even dissipating 5 W with such a small heatsink is difficult, let alone more power, so the manufacturer provided for a fan to be connected.
The fan’s negative wire connects to the F- point (on the left), and the positive wire to the input voltage. Since my fan was rated for 12 V, I connected it after the converter and, moreover, after the protection diode, although you can connect it before the diode as well (the diode is at the very top of the photo).
I connected it at the closest point, just to keep things a little neater. While I was at it, I replaced the measurement connector with a terminal block.


The board’s current consumption barely changes with the operating mode and is 65 mA at a 5 V supply (with the converter).
The fan, of course, increases the current draw. In my case, I used a 12 V, 0.5 W fan, and with it the board draws 170 mA.
This fan was more than adequate for testing a 4.2 V battery at maximum current.


Checking voltage measurement accuracy.
At low voltages, the measurement accuracy is simply excellent—practically spot on. Above 5 V, it reads 0.02 V low, but I’d still call that excellent.
The description specifies a maximum voltage of 8.5 V, but if you raise the voltage while the load is already on, it measures up to almost 10 V. It displays 9.95, and the reading doesn’t change even if you raise the voltage further.
Trying to start a test at a voltage like that produces an error.


Before checking current-setting accuracy, I measured the battery current draw in standby mode. No issues here either: at 4.2 V, the current is 0.1 mA; at 8.5 V, it’s 0.2 mA.


Current-setting accuracy was also good. In my opinion, it’s excellent for such a cheap device. What’s more, at the end I checked how much the current drifted after warm-up, and the difference was within the measurement error.


But friends, I bought this board to find out how accurately it can measure the internal resistance of batteries.
To test it, I gathered a number of different batteries, varying in manufacturer, capacity, maximum current, voltage, and chemistry. I could have used even more test subjects, but decided that 12 would be plenty. I also used a four-wire connection and a good-quality holder—at least, I don’t have a better one.


And here are the results. In almost every case, the ZB206+ showed some “jitter” of 3–5 counts. For example, instead of displaying 17 mΩ, it would show 16–18, briefly jumping to 15–19. More on that later. Some batteries showed more variation (VTC6), others less (NCR18650B).


I also checked the measurement accuracy with discharged batteries, which I had on hand from other tests.
The last pair of screenshots is in the right order: VTC5A on the left, VTC6 on the right.


All well and good, but you probably noticed something odd in the results of the previous test. I mentioned above that the readings are often very unstable, changing about 2–3 times per second, which makes it hard to tell what the resistance actually is.
But when I tested discharged batteries, the effect became even more pronounced with some of them. The photos below show the same battery, taken about 30 seconds apart. In other words, I inserted the battery, checked the resistance, and left it in the holder. The readings started “drifting” like this: for about half a minute, 10–14 mΩ, then a gradual rise to 16–19 mΩ, then after another half minute, back to 10–14 mΩ.
It may not have been exactly half a minute, but the intervals looked roughly the same.


And what about the 1 kHz frequency? Nothing doing: the measurement is performed using short pulses, effectively with DC.
1, 3. - Voltage waveform at the transistor gate
2, 4. - Voltage waveform across the shunt.

The pulse repetition rate is about 4 Hz, with a pulse width of around 3 ms. The circuit operates as a current regulator, set to about 1–1.2 A. For DC resistance measurements, a current of 0.5–1C is usually recommended, so 1.2 A is a good fit for batteries with capacities in the 1.2–2.4 Ah range.

In any case, this is not an AC measurement, let alone one at 1 kHz. If we roughly extrapolate from the 3 ms pulse width, assuming the pulse and pause were each 50% of the cycle, the frequency would be 166 Hz.


While I was at it, I checked how the device behaves when measuring in two-wire mode.
Resistance measurement doesn't work unless the measurement leads are connected.
1. If you connect just one lead, the resistance reading is 1.5 times too low, and the voltage reading is exactly half.
2. If you connect the second lead, both the resistance and voltage readings return to normal.
Switching to four-wire mode afterward makes no difference, so I still don't understand the point of the 2/4-wire setting.


And of course, I also tested battery capacity.
I picked a few batteries for the tests and decided to check their capacity.

Murata VTC5A.
Discharge current: 2.5 A, which is 1C.


When you start a test by pressing the SK button with automatic battery detection disabled, you'll be prompted to select the cutoff voltage. That's why I recommended disabling automatic detection: setting the cutoff this way is convenient, and you can also choose it more accurately, since the automatic setting uses a cutoff of 3 V.
The adjustment range is 1–6 V. The upper end is fine for 2S packs, but 1 V isn't quite right, since even Ni-Cd batteries are tested down to 0.9 V.
I set the voltage to 2.5 V.


During the test, the display cycles through capacity in Ah, capacity in Wh, battery voltage, and the set discharge current. Capacity is shown for 2/5 of the time, and each of the other readings for 1/5.
You can't stop the test partway through and view the result. Well, you can stop it, but the device switches to current-setting mode, so the result is unavailable. Once the test is complete, the display flashes, and you can view all the results in this mode.


At the end, I got 2.409 Ah or 8.67 Wh.


My previous test showed 2.421 Ah or 8.693 Wh, practically identical. And considering that I ran more tests after my first one and the battery had lost a little capacity, the result is excellent.


Next was the Murata VTC6. But I got a little confused and accidentally set its cutoff to 2.5 V instead of the 2.0 V used in their test.


Because of the mistake, the test took two stages: discharge to 2.5 V, then down to 2.0 V. The total came to 2.975 Ah or 10.585 Wh. In the original test, this battery delivered 3.015 Ah or 10.67 Wh at a 1C discharge rate.

An interesting detail: after the test is complete, you can press the SK button to display the battery voltage. It then settles at a certain value (the two photos on the right); apparently, this is the average voltage over the course of the test.


I also tested two more batteries, the Samsung 30Q and 35E. Discharge rate: 0.5C.


Results: the 30Q is shown at the top.


The 30Q ended up showing a slightly higher capacity than in the original test, but in this case it was a little overcharged (you can check the voltage in the internal resistance test).


The 35E was the opposite: it was slightly undercharged, so the result was lower than in the original test. You could say everything checks out.



Conclusion.
No issues with the current, voltage, or capacity measurements. I’d also like to point out a few positives:
Convenient controls for selecting the load current and discharge cutoff voltage.
The fans are controlled according to the load power.
Overall, the build quality is pretty good.

One thing deserves special mention—I’d even put it in bold: if you disconnect the power during a test and then reconnect it, the load resumes the test without resetting the readings! In my view, this alone outweighs all the drawbacks, because every other device I know of starts the test from scratch.

And now the drawbacks:
The internal resistance measurement mode is awkward to activate; it really needs a button with normally closed contacts in the power circuit.
The minimum voltage is 1 V, not 0.9 V.
Internal resistance measurements have a wide spread, and the readings drift even with a quality holder and a four-wire connection.
And of course, there’s no AC measurement or 1 kHz frequency measurement.

It’s a shame I mistakenly concluded from the initial information that this tester measured internal resistance accurately. That’s why I ordered it in the first place—to check that claim. Now I’m wondering what to stick it on use it for.
Still, I recommend it for its decent specs, ease of use, ability to resume a test, and price. It’s hard to find anything more interesting for 7 dollars.


That’s all from me. I hope you found this useful.