HTC Desire battery shutdown analysis

I have had the same problem as some other HTC Desire users complaining on different forum: my smartphone shut down when the battery remaining capacity percentage  get down to around 15%.

I tried to understand which was the problem, so I first record a full discharge at constant current level of around 250mA thanks to  »Battery Monitor Widget » program.

The following chart confirms a sudden stop of the smartphone around 15%. I also checked a Nexus one original battery  on my Desire and observed the same behavior

HTC Desire battery shutdown analysis batterieshtcvergb

The part number of the checked batteries are 35H00132-01M for the Nexus one and 35H00132-05M for the Desire.

At that time I was wondering if the problem was coming from the batteries or from my smartphone.

I decided to buy two new batteries, one genuine HTC and an other one unbranded, then  I proceed the same way as previously: full discharge at a constant current of around 250mA. The results are given on the next chart.

batterieneuvesdechvergb

The unbranded battery shows an identical decrease linear ramp as the old HTC ones except it stops at 0% level as normally expect for all batteries!!!!

The voltage curve during discharge is very common for both batteries. The only difference is on the time of working.

tensionbatterydechvergb

As the discharge current is constant and almost identical in all tests, it is clear that the old batteries ( Part number xx-01M and xx-05M) have both lost about 15% of their initial capacity as they stop about 15% earlier.

Why the HTC Desire ( and may be some other smartphone ?) is not able to take into account such a capacity loss?

Previous analysis I made on HTC Windows phone showed that the % remaining battery capacity was calculated from the measured battery voltage and current during working. This evaluation process was independent of the effective battery capacity even if the precision of the % figures was not so precise.

On this new (Android related?) HTC smartphones the figures looks very precise while the battery holds its original capacity. As soon as the battery capacity begins to decrease, the shutdown problem occurs. This is worse than having a poor percentage precision………IThe percent remaining charge is calculated from the original battery capacity and the total current consumption since the last full charge.

As a consequence the smartphone battery managing program needs a calibration process to update the current battery capacity. Then the discharge battery curve should be changed from the red line to  the green dash line as shown  in the following graph

Expected-battery

Unfrotunately the  callibration process will not be   be able  to improve the battery total capacity. It has no influence on the chemistry of the battery……

One calibration process proposed (http://forum.xda-developers.com/showthread.php?t=755903) deals with a precise setting of only the 100% charge of the battery. It looks to me inaccurate to solve the battery shutdown problem.

It is obvious that a full charge and discharge cycle is necessary to evaluate the actual battery capacity and to introduce this value in the battery managing program. It seems that the Battery Calibrator  ( https://market.android.com/details?id=net.jonrichards.batterycalibrator.ui&feature=also_installed  )  is able to solve this problem, unfortunately it works only on  »rooted » smartphone.

Battery Monitor Widget designer explain that the calibration could be obtained on unrooted smartphone : http://www.3c71.com/android/?q=comment/234#comment-234  …. may be on their  »Pro » version. I tested it and does not work on  my unrooted HTC desire.
The genuine HTC battery part number 35H00132-06M exhibits a very strange discharge curve.  The % remaining charge data are  meaningless. The  »Battery Monitor Widget » was unable to screen any significant values of battery voltage,  current and temperature.

pn06m

tempdech

 I think that neither  »Battery Monitor Widget » nor any internal program of the HTC Desire are failing in this case.

I checked the electrodes of the batteries with a multimeter:

connexionsbattdesire.jpg

The resistance between  contacts  2 and – is very similar for all models: 4.7kOhm. This is the one wire data bus introduced in new battery system

On batteries xx-05M and unbranded model the resistance between contacts 1 and – is 3.3 MOhm with the + measuring polarity on contact 1. This value is very sensitive to temperature. On the reverse polarity the reading is about 180 KOhm

On battery xx-06M the resistance on the same contacts is 53 KOhm and non sensitive to temperature.

Li-ion batteries are designed with an internal safety circuit. Most of them delivers temperature information to external charger. The contact N°1 of  »normal » batteries gives this information. The internal safety circuit of this battery 35H00132-06M ( S/N : 2TGA1095012360)  does not give any temperature information as confirm by  the previous graph. This is very dangerous for customers who uses this model.

 


 

 



4 commentaires

  1. Marcio dit :

    Hi, could you tell, what was the brand of the unbrabded battery and where did you buy it? Thanks.

  2. remote control helicopter reviews dit :

    Great information, thanks for the share!

  3. Fran dit :

    Hi Michel

    Good piece of research. I’ve been looking at possible differences between an HTC 35H00132-06M and an unbranded battery labelled for Nexus One/Desire. I have an HTC Desire which seems to have occasional problems with the unbranded battery.
    I have checked the parameters between the battery terminals between the two batteries and found the following (using the same terminal designation as you):

    (a) between and – terminals the voltage is roughly the same, just over 4v off-load.

    (b) between – and (2) the resistance is about 4.7k ohms on both batteries, either way. Terminal (2) is marked T on the unbranded, possibly the temperature sense.

    (c) between – and (3) the resistance on the unbranded is 49.5k and 47k respectively switching the meter probes. On the HTC it’s about 190k one way, and over 2M ohms the other, but it’s less clear as I think there are non-resistive impedances in the circuit which affect readings on this battery. On the unbranded battery, this terminal is marked DQ, and I’d like to know what this means.

    So there appears to be a difference between the HTC and unbranded for terminal (3). The question is: what affect does it have on the phone?

  4. 78michel dit :

    The terminal with 4.7kOhm is the data line communication between the smartphone and the internal safety circuit in the battery.
    What you call the (3) terminal is a separate connection to an internal battery temperature sensor .If you heat up your battery with an hair dryer while measuring the 2Mohm impedance you should notice a quick value decrease. This is quite dangerous do not overheat the battery
    Then if you put your battery immediately back in the smartphone you will see on any applications that gives you the battery tempreature that the battery temperature has change.
    My problem with the HTC 35H00132-06M model I got is that the temperature detected by the smartphone never change: this means that the samrtphone will not switch off if the battery temperature get too high. I wonder if this is really safe???? So I decided to do not use this battery
    Very recently , it has been reproted that 2 Iphone have exploded due to battery overheating. This confirm that battery temperatue control is critical so I am not ready to take any risks on that point

Le Lion du guidon |
LGN tunisie |
Bonsai amateur |
Unblog.fr | Créer un blog | Annuaire | Signaler un abus | Veolia Corse Tour 2011
| olivaurelie
| Bienvenue chez Patty