Unusual effect on nanoKenbak-1

[fcpr...@gmail.com]

I just received my two nanoKenbak-1s today and noticed they both do this sort-of strange thing.:

When running from the battery, everything works as expected.

But when running off the USB power (i.e.: while charging the battery), touching the outer metal case of the crystal with my finger causes the processor to either reset or go into a locked-up state. When I touch it, I am touching nothing else - the board is laying on the bench - and I'm wearing rubber-soled hiking shoes! I doubt it's static electricity because the effect is repeatable without me moving even a tiny bit. I would not expect the outer case of a crystal to be electrically connected to anything. But it's like some sort of charge builds up there and my finger is discharging it.  But only with the USB power. No big deal, just very odd.

[famousd...@gmail.com]

I'll give that a try, but I swear I've touched the crystal before...

[fcpr...@gmail.com]

I just tried it again - it's more consistent if I dampen my fingertip a bit.

[fcpr...@gmail.com]

So, guess what... as of this morning, they are both fully charged and with the charger plugged in, neither is showing this effect!!!

I can only suppose that it was a combination of running the ATMega328 at 3.6 volts and the fact that the battery being less than fully charged dragging down the charging controller output below a point where the processor could run reliably. Since they are now both immune to this effect, I'll just keep an eye on them after I've discharged the batteries a bit to see if it returns.

I took out my old Weller WM-120 12-watt iron with the tiniest tip I have and reflowed the solder on the dropping resistor for the STOP LED, and now the LED is stable (had been flickering with any flex of the board). Another problem solved.

My next project will be hacking in a connection to the serial port on one of them. The challenges there are threefold:

1. The TX/RX pins are right between INT0 and RESET so I have to be super careful soldering tiny wires to them.

2. The ATMega328 spec says the max voltage on the signal pins is Vcc+0.5 volts, so I'll have to work out something to guard against 5v in.

3. This looks to NOT be a multi-layer board, so I'm hoping it will be safe to drill holes for a 4-pin header clear of the traces I can see.

[fcpr...@gmail.com]

Serial port done. Works fine. Yes, my 74 year old hands did manage to bridge RX to RST even with a tiny tip - dug out the bridge with the tip of a steel straight pin and blew it clean - no way I was going to get it out with heat. Had a USB-FTDI adapter with a jumper for 3.3v, so no worries about the voltage on RX. Drilled no holes - epoxied the header horizontally instead so everything would be on the top of the board. A drop of epoxy on top of the chip and wires for strain relief. I'll call this the NanoKENBAK-1 Plus.

[fcpr...@gmail.com]

If I tried to power an ESP-01 module (rated for 3.3 V) from the NanoKENBAK-1, do you think the 3.7 V - 4.0 V would kill it?

[fcpr...@gmail.com]

To answer my own question (ESP-01s are cheap enough to be sacrificial), NO the slight over-voltage doesn't kill the ESP-01 and YES it does work on the nanoKenbak-1 but NO it won't work for very long because it draws 60-75ma. but the LIR2032 can only be charged at 35ma. max, so the battery will discharge completely in about an hour even when connected to the charger (and even faster when not). Without an external power source for the ESP-01 (which sort of defeats the purpose), it's a fail, so back to my FTDI umbilical cord. For a quick upload of a program that doesn't itself require the serial port, I can plug it in, do the upload, and pull it out.

My mostly failed hookup for WiFi telnet to nanoKenbak-1: