For a quick summary, scroll down to the bottom
Production of the real Go+ stopped a long time ago, and stocks have long since been depleted at all legitimate sources, but "New" devices are still widely available from Chinese merchants on online shopping sites for a fraction of the original price. Surplus factory overproduction obviously (!)
I ordered two devices from two different sellers. Unsurprisingly, both were counterfeit. Unit A came in a replica of the genuine box (with faded ink), had the Nintendo logo (in the wrong font) on both clip and wristband attachments, and the build quality was almost as good as the genuine device (not a monumental achievement) aside from the slots in the plastic part of the wristband that attaches to the device being slightly too small which made it more difficult to move the wrist strap to get at the screw. Unit B came in a white box, lacked the Nintendo logo, and the wristband was particularly horrible and weakly constructed; the clip fell off almost immediately and the plastic parts had mould flashing. The one-time safety release link on Unit B lacked the narrow sections on the sides which are supposed to break, but I'm sure that wristband wouldn't have any trouble breaking should it get caught up in something.
On to the teardown…
Both fakes have the same MAC address and the PCBs have identical layout to each other. The PCB layout is a little different to that of the genuine product and it has been depanelled in a cheaper way, using break-outs created by drilling lines of holes. The mechanical parts of each device are slightly different and appear to have been independently designed by each of the counterfeiters to closely match the original device rather than be produced from stolen designs.
The most obvious thing missing in the counterfeits is the lack of covering on the DA14580, which is in a light-sensitive WLCSP. Firing a flash-gun 15cm away from the front of the (assembled) unit will make it reset (and disconnect) instantly; the genuine device is immune to this problem even if the flash is held right next to it. There is a video in the gallery.
The second most obvious change is substitution of the boost converter. The IC in the genuine device appears to be house-marked and I couldn't find any data on it. Both counterfeit units have a boost converter whose markings suggest it's a SD6201. The SD6201 is a cheap Chinese boost converter ($0.09 in volume) that has a claimed quiescent current of 260µA, which will drain the battery in a month if it's always enabled, and it has no apparent undervoltage lockout (not that it matters when it only powers the LEDs, see below).
In reality, both units measured 22µA, which was not as good as the genuine device. This was better than the spec (above) because the boost converter is not used all the time; it's enabled when needed to power the LEDs and disabled the rest of the time. The DA14580 is powered directly by the battery.
The genuine device has a quiescent current of 2.0µA, which is in line with Dialog's IBAT(EXT_SLP)_BUCK_50kB> parameter. There's nothing close to 22µA in that datasheet so the high quiescent current draw of the counterfeits was likely due to substitution of the DA14580 buck converter components with cheaper "alternatives".
The Pokeball Plus (PBP) quiescent current was initially 32.5µA but it dropped to 13.6µA after a while. Having a Pokemon in the Pokeball didn't make any difference. PBP connected to Nintendo Switch: 10.5~11.5mA (no, not µA)
Undervoltage tests were performed on unit B. Going below 1.8V resulted in ungraceful failure; the unit disconnected and either the LED remained stuck on, or the generic error red pulses sequence was displayed, followed by the blue LED becoming stuck on. In that situation, turning the voltage back up had unpredictable results; in one case, it started advertising and connected and worked just fine, in another case it was necessary to cycle the power to restore function.
The genuine unit was not much better in terms of undervoltage lockouts. I couldn't get it to end up with LEDs stuck on, but on more than one occasion, slowly turning the power supply down below the 1.8V threshold resulted in the device drawing 5.3mA and the only way out was a complete power cycle. This wasn't too unexpected because the same power scheme (external boost converter powers only the LEDs) is used here too. It should be noted that the DA14580 isn't well suited to applications requiring an under-voltage lockout unless the designer can add an external one – quote from the datasheet: "The system should never be cold booted when the supply voltage is less than 2.5 V. A manual power up with a power supply less than 2.5 V in buck mode might create instability."
The PBP's "UVLO" was approx. 2.25V. It's rather ungraceful like the Go+, with the current measurement all over the place (maximum approx. 12mA), but it did recover when the voltage was turned back up. The device has a buck-only switching voltage regulator so the LEDs get dimmer as the voltage is adjusted below 3V.
Current draw tests
The table below shows the current draw in µA when connected at 3.0V (3.7V for PBP) with different connection priorities (priority tests done with the device idle) and test patterns (test patterns done with a high priority connection):
|Device||Default||Low-power||Balanced||High||White+Vibrate||White No Vibrate|
The low-power measurements were all over the place and need a 'scope for accurate measurement. For better responsiveness of the device in the game, you'll want to use Balanced or High anyway.
Note the increased current draw of the fakes when the connection priority is set to High. That suggests sub-par components were used with the DA14580's internal buck converter. The PBP's consumption is outrageous and suggests it was primarily intended as a Switch controller with the Go+ functionality added as a 'bonus'/afterthought. Also note the higher consumption of the PBP when using the default connection parameters – this could well be related to claims that the PGP is "faster" than the Go+.
The White+Vibrate and White No Vibrate test patterns were run by writing the following values to the LED_Vibrate_CTRL characteristic using nRF Connect:
- 00000002FFFF7F000000 (White+Vibrate)
- 00000002FFFF0F000000 (White No Vibrate)
The main difference between the fake and genuine devices is the current draw of the vibration motor, about 7mA for the real device and 21mA for the counterfeits. The counterfeits do have slightly stronger vibration, but this comes at the cost of substantially reduced battery life. The counterfeits can be expected to have less than half the battery life of the genuine device during normal use. As for the Poke Ball Plus, the less said, the better.
Although the counterfeits have higher quiescent current draw, this doesn't have a huge impact on battery life during normal usage, though they will completely drain the battery if you don't use it for a year and you don't take the battery out, unlike the genuine device which will last 10 years. If the Poke Ball Plus is to be stored for a long time, it should be recharged at least once a year, or disassembled and the battery disconnected internally.
I'll repeat the tests on a Go-tcha if I can obtain one with a faulty battery to tear down; the casing will probably get broken during the teardown so I don't want to risk a good one.
Source: Original link
© Post "Electrical analysis and teardown of counterfeit Go+ vs genuine Go+ and Poke Ball Plus" for game Pokémon Go.
Top 10 Most Anticipated Video Games of 2020
2020 will have something to satisfy classic and modern gamers alike. To be eligible for the list, the game must be confirmed for 2020, or there should be good reason to expect its release in that year. Therefore, upcoming games with a mere announcement and no discernible release date will not be included.
Top 15 NEW Games of 2020 [FIRST HALF]
2020 has a ton to look forward to...in the video gaming world. Here are fifteen games we're looking forward to in the first half of 2020.