1. Delmeister's Avatar
    Does anyone know the funtion of each of those three pins? There are no measurable voltages on them. It may be possible to make a charger/stand if one knew what they did.
    10-12-12 11:23 AM
  2. FF22's Avatar
    I think Peter of Battery Guru fame has tested the pins. I would imagine some others have, also.
    10-12-12 01:25 PM
  3. peter9477's Avatar
    There certainly are measurable voltages on them, if you're measuring on the charger. Measuring on the PlayBook would, I hope, not show you much of anything.

    There are old threads where we identified which pin was which, purely in a voltage sense. I think it was 12V, ground, 12V. I had pointed out, however, that one of those pins was connected internally to a chip in the charger, whereas the other was not, and we've also noted that reversing the connector will lead to it showing the lightning bolt symbol onscreen yet not actually charging all the time (for those with defective chargers where one of the three pins is slightly lower than the others). These things suggest that it's not a simple 12V/ground/12V situation, but rather than one of the pins has some form of data on it, superimposed on the power "signal", and that it may be more than trivial to reverse engineer any of this.

    I'm happy with the chargers as they are for now (aside from known issues) so I'm not going to get out my scope to dig deeper for now.
    FF22 likes this.
    10-12-12 02:41 PM
  4. Delmeister's Avatar
    Thanks for that info Peter. I was surprised at the 12 V since the USB charger is only 5V.

    So with your information, I got a current limited power supply, limited the current to 50 mA and applied a varying voltage to it. The first thing I noticed is that by applying the positive voltage to either end conductor drew essentially zero current. I then reversed the connection and did get a current. It maxed out at 50 mA at a voltage of about 10 volts. Also it did not matter which of the outside pins I used for the ground. They appeared shorted to each other. A measurement of the resistance between the two outside pins gave zero ohms, at the most 0.1 ohms.

    When you say that one of the pins went to a chip in the charger, would that just have been the positive output of the regulator supply that went to the middle connector?

    So in summary, this is what I found.

    - the pins are Gnd/V/Gnd where V appears to be greater than 10 volts for charging beyond 50 mA
    - the two outside pins appear to be shorted together internally
    - you see the lightning bolt at even less than 50 mA.

    All these numbers are very approximate since it is very difficult to try and simultaneously hold wires against those small pins, look at meter readings, and turn dials. I'll try to make a more rigid connection and get more accurate data. In the meantime, if anyone knows anything official, your input would be greatly appreciated.
    10-12-12 04:01 PM
  5. skyrocket9's Avatar
    Thanks for that info Peter. I was surprised at the 12 V since the USB charger is only 5V.

    So with your information, I got a current limited power supply, limited the current to 50 mA and applied a varying voltage to it. The first thing I noticed is that by applying the positive voltage to either end conductor drew essentially zero current. I then reversed the connection and did get a current. It maxed out at 50 mA at a voltage of about 10 volts. Also it did not matter which of the outside pins I used for the ground. They appeared shorted to each other. A measurement of the resistance between the two outside pins gave zero ohms, at the most 0.1 ohms.

    When you say that one of the pins went to a chip in the charger, would that just have been the positive output of the regulator supply that went to the middle connector?

    So in summary, this is what I found.

    - the pins are Gnd/V/Gnd where V appears to be greater than 10 volts for charging beyond 50 mA
    - the two outside pins appear to be shorted together internally
    - you see the lightning bolt at even less than 50 mA.

    All these numbers are very approximate since it is very difficult to try and simultaneously hold wires against those small pins, look at meter readings, and turn dials. I'll try to make a more rigid connection and get more accurate data. In the meantime, if anyone knows anything official, your input would be greatly appreciated.
    Newbs, the pins provide a magnetic connection that then activates the electrical circuit inside, i,e its a clip+switch. To test voltages you have to open it and check inside or finely splice three wires outside the pins to connect a voltmeter to once the charger is connected to the pins.

    Sent from my BlackBerry 9900 using Tapatalk
    10-12-12 06:06 PM
  6. Delmeister's Avatar
    I'd like to know more about this magnetic switch because I am currently charging it without the proper connector. The magnet seems to be in the playbook because the tiny alligator clips I am using are attracted to the pin area (they are made of steel I guess). Now since the magnet is in the playbook, then you need iron to complete the magnetic circuit, and the clips may be doing that (although I doubt it).

    So this is what I've found so far:
    - As the voltage is increased, nothing happens until about 8 volts. Then the lightning bolt appears but no significant current is drawn.
    - At about 9 volts the current suddenly jumps to about 25 mA.
    - The voltage goes up very slowly with increasing current. At 150 mA, the voltage is about 11 volts.
    - I then bought the Battery Guru in AppWorld. It showed a charge rate of about 0.5 W at this current.
    - After running at this charge power for about 45 min., I jacked the current up to about 300 mA. The charge rate jumped to about 1.5 W. The voltage only read 11.33V (accurate).
    - At the start, the battery was at about 20%, 3.78V. It is now at about 32% at 3.82V. Total run time is about 1.5 hours.
    - Most of this time, it was running Battery Guru in 'stay awake' mode as the only application. But I just went down to check it and it was in standby because I had been fooling with it and forgot to make Battery Guru full screen (It needs to run this way for the stay 'awake feature' to be effective.). When I took it out of standby, the instantaneous charging rate showed at just over 6 W but shortly dropped to the previous 1.5W. I assume the 6W is for the standby charging rate, but the energy going in was only 0.3*11.33=3.4W, so there are probably transient effects.
    - There is an export feature with Battery Guru in the form of a CSV file, showing voltages, power consumption, etc. I will dump it at the end of the test. Right now I'm going to jack the current up to 500 mA.
    10-12-12 06:48 PM
  7. peter9477's Avatar
    Note that if you switch your PlayBook's "Application Behavior" setting to Showcase, you can use the Keep Awake feature even when Battery Guru is not fullscreen, provided some portion of its window is still visible. It's not much help, but these restrictions are strictly the fault of the OS and there's absolutely no way around them, sad to say.

    inthemix9: "newbs"? Seriously? LOL...
    10-12-12 11:23 PM
  8. Delmeister's Avatar
    inthemix9: "newbs"? Seriously? LOL...
    Yeah, right.....geez. Anyway, too bad about the OS, but how do you get a piece of Battery Guru to show when you are focusing on another application?
    10-12-12 11:46 PM
  9. Delmeister's Avatar
    Cannot hold it at 500 mA (constant current limiting effect?). Needle jumps all over. Decided to charge at 750 mA where needle is much steadier. At this current:

    - Charge voltage is 11.9 V (input power ~ 8.9W), Charge rate is about 6.5 W, Battery voltage is 3.87 V, charge level is 35%.
    - After 30 min, battery voltage 4.04 V, Battery charge level 54%, all other conditions the same.
    - After another 30 min, battery level 71%, battery voltage 4.13 V, charge current and voltage still 750 mA, and 11.9 V respectively, but Battery Guru shows charging at 6.7 W.
    - After another 5 min, charge rate dropped to 6 W., current to 700 mA., Voltage 12.0 V (with some adjustment). Charge rate dropping.
    - After another 10 min, battery voltage 4.15 V, battery level 79%, charge rate 5 W, current 600 mA, charge voltage 12V. Playbook noticeably taking control of charging.
    - After another 15 min. Battery voltage 4.17 V, battery level 84%, charge rate 3.7 W, charge current/voltage 450 mA/12 V. Playbook definitely taking over.
    - +15 min, Battery voltage/charge level/charge rate 4.18 V/88%/3 W, charge current/voltage 375 mA/12 V.
    - +15 min, Battery voltage/charge level/charge rate 4.19 V/91%/2 W, charge current/voltage 300 mA/12 V.
    - +25 min, Battery voltage/charge level/charge rate 4.19 V/93%/1 W, charge current/voltage 250 mA/12 V.
    - sometime withing the next 5 min, the battery voltage/charge level/charge rate 4.17 V/100%/0 W, charge current/voltage 150 mA/12 V.

    So the charge level suddenly shot up to 100% where just before that it was 93%. The battery health according to Battery Guru had been showing at 93% throughout the whole test, so I guess PB took this to be the max. possible charge level, and when it reached it, it called it 100% of actual life. The charge rate dropped to zero, and the input current/voltage was just that required to keep the PB operating.

    Dumped the CSV file.
    peter9477 likes this.
    10-13-12 12:16 PM
  10. peter9477's Avatar
    Delmeister, you can't, as you've more or less figured out, keep a piece of BG's window showing when any other app is fullscreen. Only option there is to minimize both, with them next to each other, and squint...

    Great information there on the charging process, and congratulations on getting farther than anyone else has managed or bothered so far with this (at least, anyone who reported it). Some "newb" you are. ;-)

    I wouldn't pay much attention to the 93% thing. It's a coincidence that this matches the Health reading, which is itself merely an estimate that the OS calculates and adjusts over time based on how much charge it sees you putting into the device and taking out of it, and extrapolating from the measurements to estimate the batteries' (there are two) current combined max capacity.

    The fact it jumped from 93% to 100% is likely either because you've got a relatively new unit and it hasn't fully calibrated the level indicator yet, or perhaps because you've started charging it differently than it is used to, and various numbers "don't add up" for it. It may well adjust things to cover the full 0-100% range on its own after another few charge/discharge cycles... or not. How many recharges does Battery Guru report that you've done on this device so far?

    By the way, you may want to do some "full" discharges and recharges to nudge the recalibration. If you search the forum for "drain test" and my username or whatever, you can find my recommended approach to doing that, which doubles as a way of independently measuring the "true" health of the battery. Might be useful to you in tracking what side effects, if any, your own charging methods have over time.
    10-13-12 02:10 PM
  11. jpash549's Avatar
    Perhaps someone will explain how the PB reduces the 12 volts down to the voltage required at the battery terminals to charge. Roughly the figures above show to me about a 70% energy conversion efficiency.
    10-13-12 02:59 PM
  12. Delmeister's Avatar
    Here is the graph I put together from the BG CSV file. The file includes other data such as battery voltage and current, but the two traces of battery level and charge rate tell most of the story.
    Attached Thumbnails Fast charger pin assignments??-pb-bgchargetest.jpg  
    peter9477 likes this.
    10-13-12 07:34 PM
  13. Delmeister's Avatar
    @ Peter:

    There were 19 charge cycles at the time. On the 20th, it also showed the jump when charging from 50% with a low power USB. On the current 21st cycle, it has charged smoothly all the way to 100% from 4%, using the standard USB charger. However in both these cycles, the health has only shown 85%. From searches, I see this health issue is pretty crazy. Edit...it's now showing 93% health on discharging.

    I really would like to have a look into one of these rapid chargers but can't turn up anything from searching.

    Peter, why do you not display the charge power (negative) in BG?
    Last edited by Delmeister; 10-14-12 at 12:12 PM.
    10-14-12 10:41 AM
  14. peter9477's Avatar
    Peter, why do you not display the charge power (negative) in BG?
    I do display the power, numerically, but for the first release I chose not to have the graph expand to show it for several reasons. For one thing, I didn't think it would be that interesting to many people, as the primary purpose is to monitor battery drain, not charging (at least, graphically). Another reason is that I didn't want to add graph scaling/panning at first, to keep things simple.

    You'll note that the line actually does go negative, below the zero axis... it's drawing it, but you can't see it because you can't zoom the graph or pan down there. Since it goes as high as 10W with the rapid charger, the positive portion of the graph would be squished to only 3/8 of its current vertical size, were I to try fitting it all in by default.

    Future feature improvements...
    10-14-12 11:24 PM
  15. Delmeister's Avatar
    Perhaps someone will explain how the PB reduces the 12 volts down to the voltage required at the battery terminals to charge. Roughly the figures above show to me about a 70% energy conversion efficiency.
    The playbook almost certainly drops the voltage like almost every other voltage converter where efficiency in space and power is important. It's by high speed switching. The efficiency is usually quite a bit higher than the 70% you found but don't forget that the power output I was quoting is what Battery Guru was giving me, which was the charging power. To this you have to add the power that as being supplied to keep the Playbook running which is another 1-2 Watts. Also, currents are switched at very high speeds and are not uniform, so measurements with standard instruments have to be taken with a grain of salt.

    To understand switching power supplies, you have to understand the behavior of one component really well. It is the solenoid, or inductor. You will see lots of these in switching power supplies in the form of donuts wrapped with wire. The donut core (or toroid) is the best form for concentrating the magnetic field in the vicinity of the winding itself. The underlying expression is:

    V = L (di/dt) Where V is the voltage that is applied across the winding, L is the inductance of the coil (a constant depending on the number of turns, the coil shape, the core material) and has the units of henries, i is the current, and t is the time. The ratio di/dt is a differential and represents the rate of current change with time.

    For example, in the playbook, one end of the coil would always be connected to the +4 volt battery. When the rapid charger is connected, an electronic switch in the playbook connects the other end to the to +12 volts. So you have 8 volts across the coil. If the inductance is 1 microhenry, the current rises through the coil at a rate of 8 million amps per second. After 1 microsecond, the switch disconnects from the charger and connects the coil to ground (the latter is actually done instantaneously with a diode).

    The current at this point is 8 amps through the coil (and into the battery + playbook if it is turned on), and the battery voltage is applied to the coil to try and make the current go in the other direction. It's effect is to decrease the current at the rate of 4 million amps per second, so that after two microseconds, the current through the coil drops to zero. As the current starts to reverse, the coil is disconnected from ground (actually done automatically with the diode since it will allow current to only go in one direction). At some point, depending on power demands, the coil is again connected to the +12 volts and the cycle repeats.

    In the above example (and it is only an example), a complete cycle is 3 microseconds, and if carried out continuously would represent a frequency of 333,333 Hz. The current is sawtooth and always in the same direction. Its average value over a period is 4 amps. The battery, with the help of capacitors, absorbs this easily and its voltage over the period is fairly steady. If you have followed this so far, you might want to go further and prove to yourself that the efficiency of power transfer is 100%.

    Hope that helps. I would draw some pictures but Im too useless at that.
    10-15-12 09:44 AM
  16. jpash549's Avatar
    Thanks for the reply to my question. I googled the topic and found a lot of information most of which would have required too much time for me to begin to comprehend. Your explanation is to the point and direct enough to allow those of us without much circuit knowledge to at least get the general idea. I also liked your graph of your experimental results.
    10-17-12 03:05 PM
  17. Delmeister's Avatar
    You're welcome. These switchers can look pretty complicated but they essentially are all based on a coil, a diode, and an electronic switch. I actually ended up doing a plot of the waveforms along with a picture so I might as well post it.
    Attached Thumbnails Fast charger pin assignments??-pbswitcher.jpg  
    10-17-12 08:43 PM

Similar Threads

  1. Are the Voltages on the pins of the Rapid Charger known?
    By captainbdsc in forum BlackBerry PlayBook
    Replies: 3
    Last Post: 05-08-12, 09:58 AM
  2. Replies: 9
    Last Post: 04-20-11, 09:30 PM
  3. BB Alerts - Assigning color to PIN Messages?
    By CipherDias in forum BlackBerry OS Apps
    Replies: 10
    Last Post: 12-15-08, 06:01 PM
  4. Assign PIN to Group?
    By arell12 in forum BlackBerry Administrators
    Replies: 4
    Last Post: 07-18-07, 06:27 PM
LINK TO POST COPIED TO CLIPBOARD