I am trying to prototype with an Adafruit board that I thought would be a good match for a Boron, but it turns out it’s a bit problematic. I see the same behavior with 2 different Adafruit boards.
If I plug a boron with connected antenna into this thing, then plug in the battery, I get about 2 seconds of a fast green flash then about 2 seconds of breathing white, repeating. It won’t connect.
If I unplug from the board, the Boron boots up and connects normally. There is only one switch that pulls the enable pin low, and the rest of the board pretty much connects a terminal block to the Boron. Any ideas? What am I missing?
At first blush, I’d say that the Adafruit boards have some kind of issue. Obviously they’re about as simple as they can be, so I can’t imagine what.
You don’t happen to have one of their Featherwing microcontroller boards, do you? I guess it’s possible that you’ve uncovered some kind of “difference” between their stuff and ours, but AFAIK we follow the Featherwing spec to the letter.
BTW while we’re talking about such things, NCD makes some cool accessories for Gen2 and Gen3.
Hi All,
A short update on this, that may be of help to others.
I think I understand what happens with this board and my difficulty connecting.
When a Boron is connected to this board, and using the antenna that comes with the boron, the Boron has a hard time connecting. If I disconnect from the terminal board, or use an NCD board in it’s place, it doesn’t have any trouble connecting. I believe it’s an RF problem, that the Adafruit board is affected by the transmit from the antenna enough that it won’t connect. If I move the antenna as far as possible from the adafruit board, it will connect.
Unfortunately I’m trying to make a compact assembly with board, sensors, battery and enclosure, so even though this is a very compact and useful to me form factor, it just doesn’t work when you put the antenna in the equation.
@rickkas7 did some testing on the Adafruit Terminal Board. These are his findings:
I believe the problem is RF interference. The switch on the terminal block board is an inexpensive switch, and it's installed practically under the u.FL antenna connector. I was able to reproduce the problem with the Boron LTE, but not with the Boron 2G/3G, however my Boron 2G/3G also has a SMA connected distant antenna so it's not as good of a test.
If I run cloud debug, the Boron LTE runs normally until the a modem connection is attempted, then the device reboots. I guessed that this was because the enable line was deactivated.
The enable pin shuts off the voltage regulator, which would cause the device to reset. It's active high, with a weak 100K pull-up. The terminal block power switch turns off power by grounding EN, otherwise it leaves it floating.
By connecting a jumper between EN and VUSB, thus forcing EN high, the Boron LTE boots normally. Since that runs the risk of a direct short if you flip the power switch, I then tried a 10K resistor. That works as well.
Conclusion:
The design of the Adafruit terminal block FeatherWing makes it susceptible to RF interference on the EN (power enable) line because the switch is located right under the antenna connector. This seems to be worse on the Boron LTE, but could happen with any device.
If powering by USB, connect a 10K resistor between USB and EN.
If powering by battery, connect a 10K resistor between BAT and EN.
With the resistor added is seems to be sufficient to offset the interference that causes the device to power off when using the cellular modem.
@rickkas7, great work! I was just going to use some other board, but this form factor suits my project much better.
I’ll try the resistor, but have a question, if I use the resistor between Bat and EN, do you think that will still be adequate even when I’m charging via USB? Should I have the resistor to both Bat and USB?
As long as you have a battery attached you should be fine with just the BAT to EN resistor.
If you switch between sometimes having a battery or not having one at all then things get more complicated because you’ll need some diodes because you can’t have pull-up resistors to both because current will end up flowing in bad ways between the two inputs.
The resistor will only affect power consumption if you are powering down the Boron by connecting EN to ground. As long as you leave EN unconnected there is no difference.