Battery Testers

© Brooke Clarke 2007 - 2022

There are a number of reasons why you might want to test the capacity of a battery.  For household type batteries it's good to know if it's the battery that has died or the device holding the battery.  When working with rechargeable batteries it's good to know the state of charge, i.e. what's the capacity.

I've found a number of cases lately where the Radio Shack battery tester shows in the middle of the green, yet the battery is not powering the device.

These testers estimate the remaining life of a battery.  For Carbon Zinc and to a lesser extent Alkaline this can be done by measuring the battery voltage while a light load is applied.  If the proper load is chosen for a given battery capacity the loaded battery voltage works well as an indicator of remaining life.  The idea is to determine the state of charge using as little of the batteries capacity as possible.

HP 66311A Mobil Communications DC Source

This test instrument was specifically designed to measure the power consumed by battery operated devices, like cell phones, and to measure the associated battery and battery charger.

Aug 2007 - Added test jack table to TS-183 web page so that the Radio Shack load resistors could be compared to the TS-183 load resistors.  The Radio Shack tester showed a 9 volt battery as good yet the battery would not work in the Stamps.com postal scale.  The TS-183B showed the resistor as bad after abut 5 to 10 seconds.  There's a 100% difference in the load resistors for this voltage.  Need to check the others.

In addition to the test leads this battery tester also has metal contacts for a 9 volt battery on the front panel, the negative contact and be used on the bottom of AA, C, D battery and the red lead to make testing easier.  On the right side is a slot for button cells and part of the front panel can be depressed to make contact to them.  For most common batteries I have used this tester, but now am no longer as confident in it's abilities.
 

Since the above Radio Shack tester does not work very well I've been on the lookout for an improved tester.  ZTS makes a number of handheld battery testers and the Mini-MBT (Multiple Battery Tester) is what got.  They have a utility and a design patent.

6823274 Apparatus and method for testing remaining capacity of a battery, Nov 23, 2004, ZTS, 702/63 ; 320/125; 320/136; 702/64; 702/65; 702/79

uses voltage at end of resistor load pulse.  D515444 Battery Tester - packaging for the 6823274.

When a battery is connected it's immediately under load and after a few seconds the loaded voltage is read.  The capacity is based on the loaded voltage at the end of the test time.   The patent has data on the load resistance for 8 different battery types as well as capacity information in 20% steps for 3 & 6 Volt versions of the CR123 photo battery and for the 1.5 volt alkaline terminal.

The four battery types that the MBT will test are:

Cell Type	Details
Ni-MH	1.2 volt
Alkaline	AA, AAA, C, D, N
Lithium	CR123, CR2, CRV3
Alkaline	9V

I checked the same battery that the Radio Shack tester showed in the green and this tester either shows it as bad or on the second try it failed to start the test, meaning it's dead.
Also checked some CR123 cells that have been here for years.  They test at different levels.  I haven't made any controlled discharge tests to see how accurate this tester is, but at a first look I like it.
The Mini tester uses a socketed 18 pin micro controller without a seperate MUX IC and without an external crystal.  The circuit board is soldered to the positive and negative battery terminals and so can not easily be removed.  There's a 4 post 0.1" header for testing the unit, but it's between the PCB and back cover so is not accessible after the front cover is removed. 

Instead of using a switch like on the Radio Shack tester, this one uses 4 different positive battery terminals and a common negative test lead.  There's a groove around the tester to hold the lead and a pocket for the short probe body on the left side not shown in the photo.  The On button is on the right side.

There are two lower priced ($30) models this one (Mini-MBT) and the MINI-MBT9R.  The difference is that on this one the lower left terminal is for the CR123 Li-MnO2 primary battery and on the other one it's for a Rechargeable 9 volt battery.  I specifically wanted to try the CR123 battery test since I use these in some of my products (5BA, 90BA Battery Adapters).

The TS-183 works the same way.  You pick the test jack to match the battery instead of using switches.  The load is permanently associated with the jacks.  For example on the TS-183 jacks 1, 2, 3, & 4 are for single cell batteries and 1.40 V is the 80% indicator.   Jack 3 is for a "D" Zinc Carbon battery and uses a 6 Ohm load.  The ZTS tester is using a 9.1 Ohm for alkaline cells with a 80% voltage of 1.41 V.  ZTS could eliminate the whole relay circuit and thus lower the cost of the tester since one of the connections to the cell under test is by the probe.  That's to say whoever is testing the battery has to hold the probe on the cell   If there was a socket holding the cell then the relay has some merit, but in this case it's not likely someone will "forget" and keep holding the probe on the cell for an extra hour or more.

There are some  aspects that could be improved. 

• Using a post for the battery positive terminal seems to be upside down.  Either use a socket for the positive terminal or make the post for a negative terminal.  It's not as bad as balancing a pin on the point of another pin, but not that far different.
  
• The display is at the top so when you are holding a battery upside-down on one of the posts with one hand and holding the probe with your other hand the display is somewhere behind your hands.  Better to place the display at the bottom.  If you turn the unit upside down then the display reads backwards and the labels are all upside down, and you need to be left handed, but it still a better way to use the stock tester.
• The On button seems to get activated whenever the unit is handled.  Maybe put it on the top narrow face where you don't normally hold the tester.
• The word Pulse" does not seem to have anything to do with how it works.  The load resistor is always connected to the test terminal so the battery sees the load as soon as the circuit is compleated.  When powered up the tester scans all the terminals and when it finds on with a voltage starts a timer.  At the end of the time (always 3 seconds?) it reads the loaded battery voltage.  So there's no pulse involved.  Then, after making the measurement the relay opens the load resistor to keep from draining the battery.  Maybe this is what they mean by pulse? 
  

These are nit picks.  This is be far the best battery tester I have.

They make more expensive testers that can test a longer list of batteries.

ZTS MBT-1

U.S. Patent 6823274 shows on the front panel.

Army dry battery tester.  Works by measuring loaded voltage.

The set can be carried on a belt and tests radio batteries, Vietnam era.  Uses high current load to activate Magnesium batteries.

LS 91 & LS 94 State of Charge Tester for LiSO4 primary batteries BA-5590/U, BA-5598/U and BA-5588. 

4725784  Method and apparatus for determining the state-of-charge of batteries particularly Lithium batteries, Feb 16, 1988, 324/427 ; 320/149; 320/150; 320/162; 340/636.15

Uses pulse loading and voltages measured at various times in relation to the load pulse as well as temperature correction.  Capacity depends on manufacturer as well as model.

The idea here is to drain the battery using a known load until it's voltage is down to the end voltage and record the delivered Amp hours, or better delivered Watt hours.  Note that when the battery starts the test it's voltage is highere than at the end where the voltage is the pre assigned ending voltage.  So the early amp hours are at a higher voltage than the ending amp hours.  The early Watt hours are higher than the ending Watt hours.  So reporting the run time using a load resistor or constant current source is not as accurate as reporting the delivered Watt hours.

The results for a given battery will be different as a function of the load. 

Coulomb Counting

By measuring the current and the elapsed time the product of them is how many coulombs (Wiki) have flowed.  Most modern rechargeable batteries use coulomb counting to determine State of Charge and that can be translated into remaining run time.

Baiway BW-TF01N Ver 6

Fig 1

Fig 2

This is a stand alone coulomb counting circuit available on eBay lu_cytime208 with a title of "Capacity Tester Indicator for 0V-30V Lithium LiPo / LiIon Battery LCD Display" for about $30. 
It can be connected into either a discharge circuit to measure battery output or into a charging circuit to see how much charge in put into the battery.  Note that you always put more charge into a battery than you can extract.  That's to say the efficiency is always less than 100%.
This unit has no provision to turn the load on or off.  It is not a load tester nor is it a charger.  The Set Voff voltage is used to control when the TF01N goes into sleep mode.  The idea is to build this coulomb counter into a battery pack so it's state of charge can be monitored and to minimize the power used by the TF10N.  There is an internal current shunt that can be used for currents up to 3 Amps and after that an external shunt can be used up to 50 Amps.
It has no provision for an external power source, but rather runs on the battery pack it's connected to and that pack needs to have a voltage of more than 8 volts (up to 50 Volts).

Resistor

A fixed resistor is the load.  The load current decreases as the battery voltage goes down. 

QS-906 eBay title: .5v~12v Battery Capacity Meter discharge Tester 3.7v 18650 lithium lead-acid

14 Sep 2016 This unit came from the same eBay seller as the first non functional one, but this one has a resistor marked R020  instead of a resistor marked R200 like the first (improperly assembled) unit. See below for battery fixture.	Discharging an eneloop. More later .. . . 2.65 AH vs. label capacity of 2.45 AH. roughly works without calibrating.
Fig 5	Testing 18650 battery: 15 Ohm 5W resistor (4.1v / 15 Ohm = 273 mA) close to the 300 mA spec for this Li-Ion battery rated at 1500 mAh, which tests between 1450 & 1470 mAh (0.777 mAh shown) K&J Magnetics parts to make connection to battery use 6-32 flat head screws. 0.625" dia (RA22CS-P) and 0.25" dia (R422CS-P)

This is also known as a ZB2L3 ZHIYU Battery Capacity Tester External Load Discharge Capacity Test (I have one of these on order 29 AUG 2016)
The current sense resistor can clearly be seen with matching markings on the PCB and resistor of R020.

Function

This is a battery capacity tester that uses an external resistor as the load to drain the battery.  The load resistor has nothing to do with the device calibration.  Once properly calibrated the load resistor can be changed and the readings will still be correct.

Load Resistors

The load should never exceed 20 Volts or 6 Amps because of the MOSFET switch. The MOSFETs are in parallel so the current can be higher, but I doubt as high as 12 Amps.
There may be a firmware limit of 15 Volts or 3.1 Amps.
The current through the resistor will vary from a maximum value set by the fully charged battery voltage and resistance.
For example an eneloop AA battery may start out at 1.4 Volts with a 7.5 Ohm resistor for a current of 0.187 Amps.
The power dissipated in the resistor is V * I or 1.4 V * 0.187 A = 0.261 Watts.
Since the supplied resistor is rated for 5 Watts this is no problem, but for more powerful batteries a resistor with a higher power rating may be required.

Circuit Description

The tester uses a micro controller mounted under the 4-digit 7-Segment LED display.
An 8205s (pdf) Dual MOSFET each FET  speced: 28 mOhms RDSon with a max rating of 20V or 6 Amps are connected in parallel and switch the load to battery positive (see Fig 2). 
The battery current is sensed by R2 (0.2 Ohms) between the negative battery terminal and ground. 
The battery voltage is sensed by (a guess - hidden circuit under display) a voltage divider R3 & R4.
Just above the Ah LED is a 0.1" pitch 4-hole pattern that's probably for in circuit programming of the uC. 
The LM1117 (TI) Low Dropout voltage regulator is set for a nominal 3.3 Volt output (3.29 actual).
The uC is hidden so don't know make or model.

Fig 2 Input Schematic

Internal connection between R (J1-4)and +In (J1-3). Guessing at R3, R4 Voltage divider. R2 is the battery current sense resistor. Q1 & Q2 in parallel to switch load to ground. For all the parts except R2 the marking on the part matches the nomenclature.  But for R2 the PCB is marked: R020              and the part is marked: R200 Photo here shows a resistor marked R020.

Calibration

Battery and Resistor disconnected.

Step 1.

Display shows: 0A0u (zero Amps and zero Volts), Q1 & Q2 are open so there is no load on the battery.  The battery voltage drives the microcontroller ADC.
Battery removed and either left open or shorted.  Which works best is TBD
When OK is pressed the zero calibration for both battery voltage and current is stored (but not saved)

Step 2.

Display shows: 100u (10.0 Volts). Q1 & Q2 are open so no load on battery or heat in resistor.
Remove battery.
External 10.0 Volt power supply to battery +In and -In terminals.
When OK is pressed the voltage at ADC V-In is measured and a ratio stored to get to 10 Volts.

Step 3.
Display shows: 200A (2.00 Amps). Q1 and Q2 are turned on. So if resistor installed it will heat because of the internal connection between R (J1-4) and+In  (J1-3).  But the resistor can be removed during calibration to avoid overheating it.
Remove battery.
Connect positive source of 2.00 Amps to -In (J1-2 in Fig 2 above) and the ground to J1-1 (bottom R).
When OK is pressed the voltage drop across R2 is measured and stored.
Note the circuit impedance is a few tenths of an Ohm so the current source does not need to have a high voltage compliance, like was needed if you mistakenly try to calibrate using the battery terminals for this step.

Step 4.
When OK is pressed the above values are checked for reasonableness (how is a mystery) and if they seem OK they are saved in EEPROM in the uC.

Note this cal only needs to be done one time.

Operation

With the load resistor and battery connected power up the QS-906.
Press (+) or (-) to see the termination voltage and then press (+) and/or (-) to set the termination voltage.
Then press OK to start the test.
The first Ah display should be zero.
The first A display should be the battery starting voltage divided by the load resistor value.
The first V display should be the battery fully charged voltage minus a little for the effect of the load because of internal resistance.

Error Codes

Err1: the battery voltage higher than 15V
Err2: the battery voltage is lower than the termination voltage setting.
Err3: the battery is unable to support the load current or of the test lead resistance is too large.
Err4: the current is too large (current is more than 3.1A.

Example (29 Aug 2016)

eneloop AA cell after calibration using short in step 1 as above.  Results before test finished:
Ah:  3.076 [ended at 7.761 a little more than 3X the actual value]
A:  0.665 (current still reads 4.01 x too high)
V:  1.24 (DMM: 1.277 V, implies a current of 1.277 / (7.5 + 0.2) = 0.1658 A
Still not reading correctly.  Maybe because of the R2 value problem (PCB: R020, resistor R200).

Hacking

A way to fix the current calibration is to change the current used for the cal and tweaking the value to get the current to read correctly.

I'd like to use this to test BB-2590 military batteries that are 4 cell Li-Ion, i.e. fully charged at 16.8 Volts, so a bit above the 15 V firmware limit.
So . . . . by cheating on the voltage calibration by say 17 V / 15 V and after finding the correct current for a normal cal change that by 15V / 17V.
This would require modifying the cutoff voltage by 17/15, but then the Ah would read correct.

Fig 3 Checking Current Power Supply on + & - battery terminals Note PS= 4.50 V, measure 3.94 V because of resistance of test leads.	Fig 4 Checking Voltage Load resistor disconnected. PS on + and - battery terminals. PS= 13.76 Volts (just under err1 voltage), measure= 13.98 V. Note:  err1 only happens when tester in run mode. In startup mode I increased input to 18 Volts and the tester displayed it without any error message.

29 Aug 2016 resolved Prior Problems - only read for background Under $5 including shipping from Hong Kong.  It does NOT work as it should, needs updated firmware. Fig 1  After OAOu:Short, 100u:10V, 200A:2A Cal actual current is 1.37V/7.5Ohm = 0.182A, Display current is 0.659Amp, Why? Operation An external 5W 7.5 Ohm resistor is switched to the battery when the OK button is pressed and switched off when the battery voltage drops below the limit voltage indicated by Pnnnn. But After the battery is connected press + or - to change the value of P and when correct press OK to start the test. YouTube video of QS-906 circuit running. Maybe calibrated for Voltage (10V ref) and Current (2A ref)  So the tester can calculate the current given the loaded voltage.  If the test leads have too much resistance: Err3: the battery is unable to withstand the load discharge current resistance is too large or line. I have not yet figured out why some of the error codes appear or how to do the Calibration.  If you know more contact me. The tester has NOT been calibrated and the default cal is way off. When testing a Panasonic eneloop pro AA battery with the 7.5 Ohm load resistor the tester shows: 1.69 A at 1.28 V equivalent to a resistance of 0.757 Ohm load.  So way out of cal. But when the same battery (fully charged) reads 1.409 V open circuit and 1.403 V with the 7.5 Ohm load applied.  Actual current is V / R = 1.403 / 7.5 = 0.187 A = 187 mA The battery internal resistance is (1.409 - 1.403) / 187 mA = 0.032 Ohms. Short current current might be = 1.403 / 0.031 = 43 Amps (but not likely). Tried cal (holding down all 3 buttons then connecting power.  After start up displays: ---- (indicator for Calibration mode), 0.004 (maybe offset voltage) , OAOu (maybe zero Amps & zero Volts) when OK is pressed the display changes to: 10.0u when OK pressed again display is 200A, and when OK pressed again reverts back to Ah -> A -> V -> Ah cycle.  Doing the cal requires a power supply that can deliver both 10.00 Volts and 2.00 Amps. But after trying to do the cal first with 10V and 1A and twice trying 10V and 2A the cal fails, i.e. when testing the eneloop shows 1.8 Amps, way too high. Also tried with current in negative direction, but still get 1.8 Amps on the eneloop, not the 0.187A is should be.  Maybe some factor of 10 error? Note: with the supplied 5W 7.5Ohm resistor when the input is 10 Volts the power in the resistor is 13.3 Watts, and when 2.0 Amps are flowing the voltage drop across the resistor is 15.0 (the max rated input voltage) and the power in the resistor is 30 Watts. This means the calibration must be done very quickly. Maybe the Ah display is showing 10X the actual capacity.  Will test an eneloop to completion to see if it comes our near 2.4 Ah X10 = 24 Ah. The QS-906 shows 1.36 Volts and the DMM reads 1.348 Volts on the screw terminals.  Shows 1.741 Amps Voltage drop across the 7.45 Ohm resistor is 1.308 for a current of 1.308 / 7.5 = 0.174 Amps Note current is reading 10.0057 X higher than actual, so it looks like a 10x error. Next to see how the battery capacity compares with the MH-C9000 (tested at 1.0 Amp load, but can be rerun at  something closer to 0.174A.) If 2.4 Ah capacity and 0.174A load then  13.8 Hours to discharge.  About 12:30 now. About 1:00, was not watching, the mode indicator LED is flashing for Ah, A and V instead on being on steady.  Displays: 1.003 Ah, 1.692A (actually 0.1692A) and 1.31 V (cutoff set to 1.20 Volts).  Do not understand whey flashing.  Will let run longer to see if V display (and DMM) go down to 1.2 V.  Load resistor has 1.264V voltage drop ( 0.168A) so still connected. At about 11:00 pm the voltage was 1.24 V so not done.  Next morning very fast flashing and the capacity shows at 19.71 Ah.  Pressing OK twice shows the voltage as 1.26, i.e. the cell has recovered a little.  Pressing OK some more times to restart testing shows: 0.000 Ah, 1.583 A & 1.23 V. So, it looks like the reported capacity might be 10x the actual capacity.  Now testing an eneloop at 200mA discharge to see what capacity it has.eneloop has 2480 mAh capacity when discharged at 200 mA, so the19.71 Ah (divided by 10 = 1.97 Ah is still way off. Tried calibration: 1. Short, 0A0u -> OK 2. 10 V, 100u -> OK 3. 2 Amps, 200A -> OK but when fresh eneloop battery installed the displayed current is 0.669 Amps, much higher than the actual .174 Amps (384 % off).  The next day the display shows: Ah: 7.125 (clearly much higher than 2.4 the actual capacity) V: 1.26

Filament Flashlight Bulb

The resistance of the bulb changes in a non linear way with the battery voltage.  The bulb gets dimmer as the battery voltage gets lower.  How dim does the bulb need to get to say the battery is dead?  Kind of a fuzzy feely thing maybe not the best capacity test method.

Constant Current

When a constant current load is used the delivered Watt hours are directly proportional to run time.  This is the most constant load type. 

Constant Power

Switching Mode Power Supplies, like used in most modern battery powered devices consume what's pretty close to a constant power in Watts.  This means at higher battery voltages they draw smaller currents than when fed lower battery voltages.  This is the oppsite of the resistor load.

This is a bench top electronic load test instrument that can be set to look like a constant resistance, current, power or voltage load.  By using a computer to control the load and read back the battery voltage and current almost any test sequence can be accomplished.  There is a low voltage limitation of  2.5 Volts.

ZPB30A1 "Constant Current Electronic Load 9.99A 60W 30V Battery Discharge Capacity Tester"

The quoted title of this paragraph is the same as the eBay title for this made in China tester.  eBay seller ayanhu81 (about $20).  The same seller also has a "Four Wire Battery Holder Battery Resistance Test for 18650 26650" that works with the capacity tester.  See a nicer BF-1L 4-wire battery holder below.
Note this unit has no charging capability, it's only a constant current load or capacity tester.  Can also be used to test sources like wall warts, power banks, &Etc.
Technical Review -
Youku video (Mandarin): Use as Constant Current Load -
Youku video (Mandarin): Use as Battery Capacity Tester -

Fig 1	Input power: 12 VDC <= 0.5 (0.1 typ) DC power connector: 5.5 x 2.1 mm. Power up display: 03.0u Volts & 0.21 Amps
Fig 2 as received was in Constant Current Load mode (Fun 1)	The top display stays fixed on Volts Before inserting the power plug, hold down the red button, then plug in power. Turn knob so that Fun 2 appears, then press knob. unplug power. Now unit will come up in Battery Test (Fun 2) mode. The other parameter that shows up after the Function is the beep mode on or off.
Fig 3 after changing into Fun 2: Battery Capacity mode	1.00 Amp constant current load, 14.4V cutoff The top display rotates between: Voltage, Ah & Wh . . . . This battery is a 4S Li-Ion BB-2590 (BT-70791BK1). Probably 225 Wh, but we'll see. Note the voltage displayed (16.07) matches what I read with a FLuke 87 DMM (16.07), so for a 1 Amp load and these short wires the IR loss is about zero.
Fig 4 At end of capacity test	Note the State Of Charge bar graph on battery shows 1 bar here and in fig 3 above shows all bars. Tester reads 6.747 Ah, 15.18V & 102.4 Wh Test was run with 1.00 Amp load & cutoff @ 14.4V.

Watt's Up

Maha 777Plus

Maha MH-C9000

Triton2

iMAX B6 LCD Screen Digital Lipo NiMh Battery Balance Charger

1.5v~12v Battery Capacity Meter discharge Tester 3.7v 18650 lithium lead-acid

BF-1L Battery Fixture (Single Cell)

This is fixture was laser cut from clear acrylic and the saw tooth bed has markings for 34, 44, 49 & 65mm long cells.  It is a Kelvin fixture with separate wires for forcing current and sensing voltage.  The negative terminals are spring loaded. Fastech BF-1L

BF-1L battery Fixture with standard AA cell on 44mm setting

BF-1L battery Fixture w/o a battery

To make a measurement first do the calibration by:

Then to make a measurement:

The meter I'm using from EVB in Portugal, is the same circuit designed by Bob Parker and sold by Dick Smith Electronics  in Australia as a kit.  I saw mention of battery testing at Hints for techs using Bob Parker's ESR meter(kit)...  and an email from Bob suggested the switch and resistor to reduce the likelihood of damage to the meter.

I'll be using it to test the battery packs for my 5590BA Battery Adapter.

HP 4328A Milliohm Meter

ESR-Cap meter

New for me Oct. 2007 are the "Ready to Use" rechargeable batteries.  Sanyo eneloop brand in AA Ni-MH format.  Capacity of 1.9 to 2.0 AH.  Testing capacity can be typically done in less than 24 hours, but shelf life may take months or years.  See the eneloop Ready To Use battery web page for the test plan.

self-discharge (Wiki) causes most rechargable batteries to loose charge just sitting.

Chemistry	% / month
Ni-MH	30
Ni-Cad	15 to 20
Li-xx	2 to 3
Ready to Use Ni-MH	1.5

Also see the paragraph on Li-Ion battery maintenance and the associated video.

This is the current a battery can source when a load of 10 milli Ohms is applied for 200 milli seconds.  So far the only mechanical relay I've found that can handle 40 Amps (needed for "D", "F" and No. 6 dry cells) with a contact resistance will below 10 milli Ohms  is the one used in car starter motor circuits typically rated for 400 Amps.  Solid state devices have way too much resistance.  A method of getting the required load is to connect just under two feet of 14 AWG copper wire across the terminals.  The problem is only connecting it for 200 milliseconds and then disconnecting.  If left connected longer serious problems with the battery or load may occur.