I got the fully assembled and pre-configured "kit" for a little "science project", just configured/added the Zero's WIFI setup to the boot partition and started the "Geiger" server up -- worked right away. Easy to use great looking web front end! Love it! -- Beside a minor issue / question to be figured out -- see below.
Regarding the "buzz", at startup I noticed it barely (new version better ???) -- later pretty much silence. I also 3D printed the recommended case for HV protection very advisable I'd say ;)
What I did, may be helpful for Pi Zero Newbies ;)
created this file in boot:
with the WifI configuration -- no monitor needed to be hooked up at all! ------------- country=US ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev update_config=1
Then I can ssh as pi into the tiny box via "ssh firstname.lastname@example.org" from a linux machine for example with the default raspberry password.
And then I also added a script and a cron tab entry to make the server autostart at boot time without further interactions. cat ~/start-gi.sh ----------- #!/bin/sh cd /home/pi/pi-zero-geiger-counter/pi-software ./pyGIserver.py & -----------
then as pi user call "crontab -e"
and add at then end of the file: @reboot /home/pi/start-gi.sh
To verify the pyGIserver running after a restart -- or just open a web browser and see if it's up http://raspberrypi.home:8080.
pi@raspberrypi:/boot $ps aux | grep GI pi 906 2.7 5.2 103384 23456 pts/0 Sl 02:21 1:21 python ./pyGIserver.py
PS: May be some one has an idea, I had troubles with some USB battery packs I wanted to use operating it from, while it works fine on my computer's USB ports and phone charger providing USB power. The Pi0 started always up just fine. But I eventually got I2C IO Errors when attempting to start the pyGIserver.py when on on a USB batterypack??? Any ideas, fully charged packs, even a quite large ones -- no luck :( Any clue?
Here is the scenario I found with all my battery packs:
2019-12-21 01:36:02,807 INFO pyGI.geigercounter: Starting geigercounter Traceback (most recent call last): File "./pyGIserver.py", line 25, in <module> geiger = geigercounter.Geigercounter(total=last_total, total_dtc=last_total_dtc) File "/home/pi/pi-zero-geiger-counter/pi-software/pyGI/geigercounter.py", line 58, in __init__ self.reset() File "/home/pi/pi-zero-geiger-counter/pi-software/pyGI/geigercounter.py", line 65, in reset bus.write_byte(i2c_addr, 0x00) IOError: [Errno 121] Remote I/O error
It works! There's a lot of software associated with the project, but the only file you need to test that it works is: geigertest.py. This file also has all the code you need to incorporate the board into your own projects. All the other files are for a very specific web implementation. Note the software is for python2. If you want to run it in python3, you'll need to make changes.
The board uses a standard (0.1" 2.54mm) female 2x20 GPIO connector. So naturally you may assume that you can simply plug in a Pi Zero WH and call it a day. WRONG. If you do that, it will be connected BACKWARDS, and power will be going to the wrong pins! This board is thus incompatible with the Pi Zero WH unless you want to remove the header and resolder it. The board requires that you take a headerless Pi Zero W and solder a male header BENEATH it, with the pins facing DOWN. If you want to stack another HAT on top, you'll need to find a "double-sided" or "centered" header at least 17mm long, so that you have pins going in both directions, above and below.
I can't fathom why the board was designed this way. It would have been so much easier to simply let the Pi Zero connect to a female header beneath the board, and with a male header emerging on the other (GM tube) side to allow for additional HATS. In other words: it should be designed like every other HAT on the market. Then no special soldering would be required. If you used a Pi Zero WH, then no soldering would be needed at all. At the very least, the vender needs to update the docs and make all of this manifestly clear. It's a miracle I didn't damage something with the Pi Zero WH plugged in backwards!
Moving on, the barrel connector will only take 7-9V input. That's pretty limiting. If you're going to go through the trouble of adding a buck-boost converter, why not use a TPS630701 or something and ensure a wide input range? Nominal 12V is the voltage of choice for solar applications. LiFePO4 batteries charge at around 14.6V. So a range of 3-16V would cover many possibilities—including always-on solar-powered applications without the need for an additional converter.
When the board is actively counting (as opposed to merely being plugged in), it makes a slight but noticeable high-pitched whining noise that is characteristic of tiny transformers. It's no doubt related to the electronics involved in raising the input voltage for the tube. In any case, don't expect to deploy this to any quiet room in which people reside. It will drive them nuts.
I also think the board ought to come with at least one power LED. It's disconcerting to power up a board and see... nothing. Heck, how about going cRaZy and adding TWO LEDs—one for power and one that blinks for each count! OMG! And allow the user to disable the latter. After all, there's a microcontroller and an I2C connection, yes? Shouldn't be too hard to work that out.
Overall, despite all the nits, I do like the board. Not pancake-tube sensitive, but the price reflects that. All the mounting holes are a perfect Pi-compatible M2.5. The Pi conveniently mounts right to the board (that is, once you inconveniently do a custom solder job). I just wish more care had gone into the docs and following Raspberry Pi standards and conventions.