Modifying a knock-off battery charger to be safer

i-tv5pN2R-X3

Sometimes buying a low-cost clone off of eBay is a great option, but [Martin] wisely decided to test his counterfeit IMAX B6AC, and found it grossly lacking. His detailed breakdown shows an alarming array of problems, including poor design and construction, and a lack of warning if the balance circuit fails. In addition, the charger wasn’t properly calibrated. By using a precision multimeter, Martin found that the charger actually brought cells above critical voltage. So really, using a charger like this out of the box can both destroy your battery pack and/or start a fire. One other interesting detail – this model can only be calibrated once. Sweet features.

[Martin] detailed his fixes in a well-illustrated blog post. He first had to re-enable the calibration menu using this method which requires bricking the device first! Once un-bricked, however, he could do the recalibration using a voltage divider and a reliable power source.

This project really underscores the need for a precisely calibrated multimeter. Not only would [Martin] not have been able to test his charger properly, but the re-calibration wouldn’t have been as accurate as needed. As hobbyists, this is a reminder that we can only trust our tools if they are accurate.

Comments

  1. captnmike says:

    What about the part about buying good stuff in the first place and not counterfeit junk?

    • Rob says:

      Supporting the few remaining quality manufacturers we have left does seem like a far better option to which to resort. Nonetheless, if you’ve already got some craptastic hardware sitting around, going through it and overhauling it (if it even can be overhauled) is an excellent way to learn operational theory and troubleshooting skills. Great write-up!

    • ka1axy says:

      The problem is being able to tell the good stuff from the counterfeit junk.

      We have a couple of chargers that look just like that here at work. Different brand name, but probably from the same Chinese factory.

      Missing from the article was the name of the vendor whose charger was counterfeited by IMAX

      • gabriel says:

        Exactly. there’s no way to know what is junk and what isn’t. Specially that if you open the junk to see that it’s junk you can’t even return it!

        I bought a 9v charger after extensively reading review… the unit that arrived had the extra solder on one of the rails lose in the case! a 2″ slob of solder, just jiggling around.

  2. Gigawipf says:

    I have the Voltcraft b6ac from Conrad. It was way more expensive (about 99€). I thought it is identical to the imax devices for <35€. So these cheap chargers look identical but are poorly fabricated? Maybe i will have a closer look at it but it seems to be the most expensive one of these models.

    • If it look identical i’d check if it actually is :)
      I’ve seen that same board other times on internet. It seems that the carger is being branded with lots of different names.

      I agree with Sven that the voltages are acceptable. But the guys says in another post that the charger did not detect a failure on the per-cell connector and the lipo was catching fire for being overcharged. Now, this is the worrying part imho.
      For the supply, yes it would be better a ground connection.

      On a side node, there are many crap power supplies that provide poorly regulated voltage. Since everything is made in china anyway, it would be great to know some good brand.

      • The power supply does not need a ground connection, it is in an insulated case. Chances are those heat sinks are electrically “hot”.

        • I thought too that the psu case is insulated after all. But considering poor pcb insulation clearance in some very cheap chinese power supplies, there would be still the risk of the mains going to the 2 wires that supply voltage to the main board. It’s OK for the plastic, i just think that the outer metal case should be grounded, not the heatsinks.

      • Gigawipf says:

        i opened the Voltcraft charger and it has almost the same components but it seems to be better. At least it is soldered properly and has golden connectors. But the psu does not look better but i hope it is. I payed over 50€ more -.-

    • gabriel says:

      First rule of buying stuff from china: never buy the ones that looks like the non-china (or expensive, established brand made in china)

      If the china factory is doing a product they believe, they will have their own brand/case etc.

      If it’s just stealing the plans and removing componenets to churn out cheaper ones, you will probably see the same case as the leading brand with with another color, or sometimes not even this. Avoid like the plague.

  3. Sven says:

    Yes, cheap chargers can be bad, but this writeup makes some huge assumptions, such as that lipo cells cannot handle anything over 4.2V.

    the “critical” voltage is actually 4.4V, the limit of 4.2V is set so there is a margin for error, and the “overvoltage” is only 27mV, or 0.6%, even at 1% accuracy this is fine. Those voltages are not off in any way.

  4. Chris C. says:

    I think [Martin] might be a bit alarmist here.

    First, I wouldn’t think that going a mere 27mV above 4.2V, for a brief period as intended (not continuous trickle charging), would cause any unusual battery failures. The manufacturer datasheets I’ve seen that specify 4.2V say it exactly like that, and do not have additional significant digits like 4.20V; so there should be a little wiggle room as long as the voltage, rounded to the nearest tenth of a volt, is still 4.2V. But if anyone knows for certain otherwise, please share. Still, if it’s up to me, I stop a little short of 4.2V just to be on the safe side. There’s rarely any need to try to push batteries to the maximum possible charge anyway, and if do you have such a need, you really should get bigger batteries.

    And although [Martin] used a multimeter which appears to cost $350 (!), that really isn’t necessary; nor is accuracy to three decimal places. But checking your multimeter against a reasonably well calibrated voltage source is a good idea, or at least comparing against one or more other cheap meters, to be sure there isn’t a gross miscalibration. Do it at least once.

    Finally, [Martin] says “The ground pin is not even connected to the charger casing!” Why would that be necessary? The switching power supply is not open-frame, it’s already totally enclosed and isolated in its own plastic casing – or at least it was, until [Martin] removed it for some of the photos. It’s common not to ground the outer casing of devices powered by a plastic-enclosed SMPS. Doing so may interconnect the AC line’s ground and the DC ground, and I’ve seen a reasonable argument that this may cause problems, detailed beginning in the section “SMPS Kill Equipment” here:

    http://sound.westhost.com/articles/external-psu.htm

    I do at least agree that is a cheap SMPS, and that the undetected failure of the balancing circuit and inability to calibrate more than once, are unacceptable.

  5. pcf11 says:

    A craftsman never blames his tools.

  6. All this to save a couple dollars? A brand new real IMAX B6 is a whopping $40.

  7. tekkieneet says:

    You would stress the battery more by charging at a higher voltage and
    get less lifetime/cycles out of the battery.

    There might also be additional voltage drops across the wiring +
    connectors resistance during charging. Actualy voltage at the battery
    itself would actually be less by the I*R drops.

    Unless they are using oversampling, they would only be getting
    25.5V/1024 = 24.9mV resolution for the Atmega 10-bit ADC. 27mV is not a
    big issue. If you can calbriate it, great. If you use a charger that
    exceed that voltage by more than 50mV on a protected battery, you might
    risk tripping the overvoltage circuit inside the battery pack.

    For my own charger design, I calibrated my by collecting voltage points
    vs ADC values over the entire output range in Excel and use the slope &
    intercept point in regression analysis. Looking at my results, I
    would not use only 2 data points ofr calibration if I can help it.

    I am hoping that the charger doesn’t go above the preset voltage due to
    software issues (such as bad PID coefficients) or output ripples in the
    switcher. They use Atmega for PWM if I remember the schematic. There is
    a lot of tradeoff in PWM frequency vs resolution. So either way, you
    would have a bit of output ripples.

    Only way to make sure is to chart the battery voltage over time with an
    external data logger (that you can trust) and look at the output ripples
    with a scope.

  8. FZ says:

    It sounds a lot like an advertising video on youtube: http://www.youtube.com/watch?v=kziSFSQvxr4
    The “reverse polarity circuit” is missing. What is missing?
    The precision for his voltage measurements ist just silly. And what exactly is the problem with using a laptop charger? A little bit more info please! Otherwise it just seems like the usual “buy the original, support the developers, everything else is dangerous” without good reasoning.

    • Sven says:

      Most of his complaints are crap for something this cheap. The PSU however is a legitimate complaint, it is missing some filter components (The main common mode coil position on the board is bridged) and *looks* like the frankly lethal supplies i have seen before.

  9. tekkieneet says:

    Google for “BC6 charger schematic”. First result points to a forum with
    Atmega I/O and a link to another page with partial schematic.

    You can’t put in a diode at the charger output and expect it work.
    Current would still flow out if battery is hooked up backwards.

    The only way to make sure that the battery is plugged in the right
    directions is to look at the voltage at the battery terminal (via ADC0).
    If it is positive, then the battery is hooked up the right way. I have
    done that for my design. Looks like they can something similar for the
    BC6 clones. There is NO hardware missing.

    Input power polarity protection is done via DA2:2 comparator by looking
    at the + and – terminals in the BC6. I used a diode for mine.

  10. BiOzZ says:

    if your even thinking about using a cheap ripoff charger for somthing that its not uncommon to EXPLODE if charged right even after modifications your not right in the head

  11. Martin says:

    Thanks for your remarks, I will update my posts soon to make them more accurate, taking them into account:

    From my readings, it seems the RC community usually agrees that 4.20V is a max for unprotected Lipo, and that they otherwise tend to puff or vent. My approach was more intended to slightly decrease the max voltage of each battery cell rather than improve the accuracy (it did anyhow).

    My opinion about the earth pin not connected to the casing of the charger comes from the cheap power supply inside: cheap Chineese power supplies have electrocuted before (google: iphone electrocute womman) and I would not mind that extra protection. I’ll post Chris C. arguments since it perfectly makes sense, especially with a decent DC adapter (I have yet to find a good replacement for that one).

    I borrowed the multimeter I’m using from work this is what I usually do, considering mine is off. It is obvious that so many decimals is overkill, no argument with that.

    Finally, I recommend anyone not to buy knock-off chargers they can be dangerous and I nearly got a lipo fire home. I somehow trust my charger now but would never leave it unattended when charging lipos/li-ions.

    That thing tekkieneet says about no hardware missing is interesting. It indeed makes perfect sense and I should check my sources first. I’m however intrigued by that ‘missing’ component. Could that be a capacitor then?

    • Chris C. says:

      Thanks for the clarifications. I did Google that report about the iPhone charger. Disturbing. It’s the first time I’ve read about one possibly causing a death, but not the first I’ve read about those chargers – the counterfeit ones are *notoriously* bad. I doubt yours is as poor as that extreme example, but can appreciate why you’re wary of it.

  12. carlitos chatran says:

    This is such an alarmist post, some of it is reasonable, but a lot of it is utterly nonsense. Most of it addressed by posts above (what is the problem with using a laptop psu, the 20 mV negligible difference, etc).

  13. rewolff says:

    FYI, I haven’t read the article to verify the claim of “non-authenticity” of the charger. But my verified genuine Imax B8 has very similar problems.

    We charged a battery and decided to verify the voltage at regular intervals. Of course we ended up forgetting for a few moments and when we checked again, the voltage on one cell was up to 4.26 according to the balance display. We told it to discharge until we thought the voltage had dropped enough. We then verified with a multimeter: The calibration was off: we were now left with significantly below 4.2V.

    In the past I’ve seen it charge a battery well beyond 4.2V per cell. Verified with a multimeter. So.. that’s a genuine B8 from Imax.

  14. Pedro says:

    original imax B6 cost about 35€.

    this counterfeit one + parts needed to make it better cost more than 35€. yes I’m not counting the high end multimeter.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 96,353 other followers