Close

An Enclosure for Closure

A project log for My Very Own Geiger Counter

Fulfilling a childhood dream

jon-kunkeeJon Kunkee 11/16/2020 at 07:410 Comments

My friend, [Ramjet], has a workshop with many tools for the shaping of matter and he knows how to use them. He is also leaving my area soon to make a COVID-safe months-long visit to family, rendering that expertise out of reach. Fortunately, he agreed to spend an entire Saturday helping me get my collection of loose circuit boards mounted in the project enclosure.

Twelve hours of masked work (less a short break or two) paid off!

(Pictures and details after the break)


Here's the front panel with it on:

Functioning front panel
Here you can see the completed front panel of the Geiger counter, complete with TFT display, OLED display, four buttons, four LEDs, buzzer, IR receiver, and potentiometer. While I'm quite happy with how it turned out, this picture lets you see some of the imperfections, like the imprecise hole placement and sizing for the lower right section, the visible Dremel learning curve in the TFT and OLED screen apertures, and the screen insets due to a lack of ideal mounting hardware.

Registration

Hole registration--getting all the holes in the case aligned with the holes in and features on the parts--was probably the single hardest part of the entire project. This is where [Ramjet]'s skills came into play; he has a simple trick where you put toothpaste on the part where it will be in contact with the surface, touch the surface with the part, and then drill holes in the surface where the toothpaste lands. He usually uses it to mark where on a wall he needs to put mollies, but it also worked well at this scale. I'm especially proud of the fit of the switches and alignment of the IR receiver.

This process worked iteratively, too. For the three buttons, there were two sets of holes: one for the buttons and one for the (M2) bolts. We toothpasted the buttons, marked the surface, drilled the holes, checked their alignment (no reaming needed on this one, as I recall!), then--because the standoffs on the button PCB could now reach the surface--we rinsed and repeated the process for them.

While it reduces catastrophic hole misplacement, it is not quite precise enough to avoid small misalignments easily solved by reaming. It also does not work on parts that don't play nice with toothpaste, like the screens, LED-button combo, and buzzer. (Well, we toothpasted the rim of the buzzer to align it.) For the LED-button combo, we used the more imprecise 'trace with pencil' method and a too-large bit to punch the hole. This is why you can see the PCB through the hole. (The left red LED hole was originally drilled for the potentiometer knob, so it needed to be reamed out pretty heavily.)

Getting the screen openings right posed a challenge, as you can tell from the larger one. There were three issues: marking the correct dimensions on the surface, kerf placement (also known as "oops, I cut too far"), and getting the Dremel tool to do its user's bidding. The smaller opening was done second with the last millimeter or so done with a file and frequent fit checks, so it came out much cleaner.

Mounting

Here's the back of the front plate:

Inside of front panel
Here you can see the various bits connected to the front panel: the button board, switch+IR board (mounted perpendicular using epoxy and wood blocks), TFT board, OLED board, and power+misc board. This is after I hooked it up and got it working again, so there are wires flying all over the place.

When [Ramjet] saw my awkward switch mounting situation (described in a previous log), he reached for some spare mix-and-set epoxy he had sitting around and devised the wooden block+epoxy situation you see here. It means I can't recover the parts on that board without considerable probably-destructive effort, but it also means they're solidly mounted.

Here are two shots of the main enclosure from different angles:

Main enclosure, viewed from left
This shot shows the BMS, I2C splitter board, ESP8266-based GK-WiFi board, and GK Plus main board. Note the wire+screw BNC connector on the right, where the wire fills in as the mounting nut. Oh, and the rat's next of wires continues.
Main enclosure, right view
More of the rat's nest from the previous picture, with more of the GK Plus board visible. Note the BNC connector visible on the right edge of the photo. Please pardon the cat toy in the background. =^.^=

Drill Template

Placing and drilling the holes in the main enclosure ran into some problems. Version 1 highlighted for me the need to more accurately place the holes, but the layout I had developed on the foam board wasn't accurate enough to catch that the I2C buss board and the GK-Plus-GK-WiFi combo would collide. (It missed the box's taper and the size of the frontplate attachment holes.) The latter is why those boards are mounted at an angle. For the former, [Ramjet] came up with the idea of first doing the toothpaste-drill-ream process on a rigid template, then using that to guide drilling in the actual case. This felt like it was going to double the work, and it kind of did--finding and fitting the template to the bottom of the enclosure was a lot more work than I expected, but for most of the holes the second round of drilling was really fast. (A few of the holes wee too close to the edge for even the Dremel tool to get to them, so we ended up hammering a finishing nail through the hole in the template to make a dimple on the outside that could be drilled into the waiting, guiding hole in the template inside.)

Drill template for bottom of enclosure
This here is the piece of acrylic that was hewn into shape to fit snugly at the bottom of the enclosure, then was drilled and reamed to eventually fit the various parts. The arrow made it so that each time we pulled it out and filed it or chiseled it, we put it back in in the same orientation. It also helped make sure all the holes ended up on the same side with the same orientation. The hole with an X drawn through it was a bad one, showcasing one of the benefits of this template approach. You can see the skiddywompus placement of the left five holes required to keep the GK WiFi and I2C interconnect boards from attempting to occupy the same space--something that classical physics frowns upon.

Mostly Done

It is really, really nice to have everything mounted. The foam board made it hard to mount and dismount the boards and didn't have anywhere to put the extra bits as I built them. Every time I unplugged a wire, I had to worry about pulling the hole board out instead.

Thanks again to [Ramjet] for all the help!

Next Steps

There are two remaining issues: the TFT (larger) screen's cabling is sensitive to noise and needs shielding, and the BNC connector needs to actually connect to the high voltage output of the GK Plus. On top of that, the ESP8266 firmware is low on RAM and lacks some features I want. Oh, and the rat's next doesn't fit in the box yet; I may yet have to get even more creative to get past that.

(Yes, a lot of this would probably have been simpler with a 3D printer or a laser cutter. Due to COVID-19 safety measures, I currently have access to neither.)

Discussions