I've had problems with the N3 review sample, and I've just received the full investigation report.
The report is copy-pasted below, with the written permission of the test engineer who conducted the investigation.
Personal note: This is the epitome of responsible behavior from a manufacturer, words below restored my faith in Thumbnet, despite the fact that I have to unplug the keyboard when I want to use a dongle.
Full unedited report below:
The N3 was connected to a 4 AA battery pack supplying 6.2 V, then connected to a Lenovo Z570 Model 1024 laptop via a ferrited data cable.
When the USB connection was made, the laptop screen and two connected screens went black, accompanied by sound from the speakers.
The user examined the set-up. When the power leads from the user’s own external battery pack were removed from the Phoenix screw terminals on the N3, they discovered a lot of small wires remained in the empty socket.
The user noted that their view of the 5V and GND symbols on the N3 power connector is obscured when the connectors are plugged in.
The user discovered that 2 of the 4 USB ports on the laptop no longer functioned. Remedial action attempted so far has consisted of removing all external peripherals and battery, and pressing the laptop power button for at least 30 seconds.
After the incident, the battery pack was found to have a voltage of 6.0V.
N3 philosophy and circuitry
The N3 is designed specifically for use in ThumbNet stations, where they will be carefully handled and will not require multiple reconnections, so there is minimal risk of over-voltage, short-circuits and reverse polarity from the external power supply. Additionally, such mis-use is tolerated, because there is additional protection on circuits that the N3 connects to.
The N3 contains many protective features, mainly “self-centred” –
Safety features in the N3 include:
· USB power line series protection diode (to prevent spurious voltages being pushed back up the USB cable power feed)
· Robust MCP1826 LDO. Survives reverse polarity and over-current faults (contains short circuit current limiting and overtemperature protection).
· RF limiter diode on the input. Protect the RF chip against over-large RF signals.
· 1K “bleed” resistor. To reduce ESD susceptibility and prevent static build up in aerial systems.
· USB common mode choke. To eliminate conducted interference on the USB cable.
Nonetheless, there are some mechanisms that can potentially lead to failures in the N3, external power supplies, or USB ports. These are discussed in the next section.
Procedurally, protection has been in the form of stated limitations:
· Normal external power supply limits 4.5-6.0V (but 5V +/-10% is recommended)
· Do not apply a reversed polarity supply, otherwise damage may result to the N3 or power supply circuits through overcurrent.
The N3 is also supplied with the external power connectors in pairs, so the user only needs to mate bare wires to the power supply once. After that, the Phoenix plug should normally be used, ensuring correct polarity and a clean mate each time.
In reality, the N3 is capable of use with power supply voltages as low as 4.0V and as high as 7V or even 10V for a short time, depending on heat dissipation.
Possible effects of mis-use of N3
· Short circuit on +5v in. May cause an over-current of the USB supply
· Reverse polarity +5v in. May cause 3V3 LDO overheat and/or USB supply over-current
· Over-voltage on +5v in. In extreme case, may cause 3V3 LDO to overheat and capacitors and ICs to fail.
· Ground loop through N3. Extreme case may cause power supply over-current, PCB damage.
These failure modes and effects are not readily apparent and straightforward. For example, in principle, the LDO acts as a diode to a reversed power supply. The switching FET in reverse looks like another (substrate reverse) diode, so to a negative supply, the N3 power socket appears to be two silicon diodes in series.
However, there is a more complex mechanism that occurs: The reversed power supply only drops -0.7v across the inverse substrate diode of the regulator, but as the USB power rail connects to this point via the protection diode D2, it tries to raise this voltage point to just below +5v – i.e. it tries to raise the voltage directly against the mis-connected external supply. The reversed supply can’t put negative volts up the USB cable (D2 again) but it can make the USB supply think it is driving a direct short circuit, and a battery pack can shunt a lot of current.
If the N3’s regulator has already burnt out then the USB power supply directly “sees” the batteries, and tries to completely counteract their applied voltage.
Similarly thorough investigations have also been carried out for all potential failure modes and effects.
Nature of the particular incident being investigated
In this incident being investigated, the external battery pack was connected before the N3 was plugged into the USB port of the laptop. If there was a short-circuit in the connector, it would have started conducting the moment that the battery power was applied, and at least for a few seconds before the N3 was plugged into the USB, so the user should have noticed a significant increase in battery temperature and the battery voltage should have dropped by more than 0.2V (assuming the usual battery chemistries). Neither of these things was noted.
The above would also be true even if the battery was inadvertently connected with reverse polarity.
Without any other evidence coming to light, the most likely reasons for the failure was:
· The external power supply was inadvertently connected with reverse polarity.
· The laptop being used did not have a resettable fuse on the USB port.
Thorough analysis and testing has shown that the N3 is resilient and performs exceptionally well when operated according to the prescribed limits.
Most laptops and PCs have resettable fuses, so the risk of this type of event happening again are very small.
While the N3 is specifically made for ThumbNet – i.e. more controlled use, very few connector mating operations, and more tolerant external circuitry, it is recognized that the N3 has proved to be popular among general experimenters, and therefore should be made more tolerant of misuse.
1. Re-advise users about voltage limits and the requirement for correct polarity.
2. Advise users to mate power supply wires to the Phoenix plugs first (after checking polarity and that there are no stray wires) and then routinely use the plug and socket to attach power, rather than repeatedly removing wires from the terminal blocks. Additionally, this avoids the issue of connectors obscuring the +5V and GND symbols on the N3 case.
3. Advise users that, as with many electronics that are used for experimentation, it is advisable to use power supplies (including USB ports) that contain their own protection (thermal resettable fuse or over-current trip).
Figure 2 Additional protection diode
The recommendations in 1 to 3 above will work perfectly if followed correctly, but it is recognized that users are not infallible. A Schottky diode can be fitted to protect against short circuits and reverse polarity on the external supply. This will protect the USB supply in the event of a battery short, and will stop the N3 from trying to “charge” the battery if it goes under-voltage.
4. When using a battery supply, users may wish to fit a 1A (minimum) diode (Schottky for 6v or “conventional” for 7v2, depending on battery voltage) between the battery positive and the N3 input.
5. Existing N3 owners who want additional protection can relatively easily retro-fit a diode by cutting the track that is accessible on the rear side of the N3 board between the +5V connector and Q1, and solder in a >0.5 Amp, low-forward-voltage-drop diode. A SOD523 package will fit well. This is a simple procedure, but if this enhancement proves to be popular and users need further advice, we will provide it.
Long term enhancements
6. The N3 design is already being updated for better performance. The protective Schottky diode will be fitted as standard inside all future N3s.
Addendum 15th December 2016
The N3 that was involved in this incident was passed directly to the test engineer without any intermediate handling, and was found to be functioning correctly, with no damaged components. Current draw is 405mA.