Help with converting breadboard to PCB

I'm currently in the process of designing a PCB to transfer my Arduino Nano breadboard project onto. I come from a programming background and so have little experience with electronics beyond the research I have done in my free time. As such, I have some concerns with my design that I would appreciate if someone more knowledgeable could shed some light on. The purpose of the board is basically to break out the Nano's pins so that they can be used for various data reading purposes, including measuring battery voltage, measuring current through a shunt, reading values off temperature and other sensors, etc. Some of the various components I am using to assist in this include a LoRa radio (RFM95W) for transmitting the data, a GPIO extender (SX1509) to allow for more connections, and a discrete ADC (ADS1115) for better resolution in analog signal measurement. The power to these components will be provided by a buck converter set to 5V.

(The dimensions of the board are 100x100mm. The width of each trace is .152mm, which was the default setting in the software I'm using. This is also about the minimum clearance I have tried to allow between each trace.)

Above is what I've done so far. Please note that this design is not in any way final; not all of the traces are drawn yet and many of the silkscreen labels for the pads are inaccurate. My main concern stems from grounding. I will have several analog signals coming in that will share the ground plane (not shown in picture, will consist of the entire second layer) and return to their source (an external battery) through a connection (also not pictured) which ties the battery negative terminal and the power source's ground together. I'm afraid that the presence of digital components connected to the same ground will interfere with the measurement of these signals. I've read about ground loop, ground bounce, etc. but I don't understand any of it very well. My question is, are there any special design considerations I need to follow when creating this board to prevent these problems? I can tolerate the analog readings being off by a bit, but they still need to be somewhat accurate to be useful. Thank you.

Where is this being used? Indoors, outdoors, length of cables, any other metal work around, etc. ? But assuming it is a sort of normal bench top situation, then ground everything to a solid single point (electrically) and do not use screened leads with screening terminated at both ends. Draw a precise diagram of all the cables, screening, and devices connected, ensure the ground, shieldingd and -ve terminals are solidly connected at a single star point. wrt the board layout, maybe best to separate the analogue tracks from the digital, but maybe easiest to try it and see.

It will be used in an outdoor setting, with wires running from various different components to the board. The lengths of these wires will be in the range of a few feet, depending on what they are being run to. The only source of potential EMI present that I can think is a brushless motor. I should also note there will be two separate battery systems, one to power the low-voltage components such as the electronics here, as well as a high-voltage (relatively anyway) system that is being measured. As such, I will need to tie together the grounds of these two systems to get an accurate reading, if that is relevant.

As for the shielded wires, you are saying to only connect the shield/metal part of the cable at the ground end, correct? Or did I misunderstand?

Also, should I go ahead and try a simple design with the bottom layer being the ground plane and no other tricks? If I do that, I may have to create some islands in the ground plane in the form of vias/crossings for traces due to limited space (like the blue wire in the diagram); that won't cause any extra problems will it?

Flood both sides with ground plane. If there's islands, drop a via there to connect to the other side, which probably isn't islanded at that point.

It used to be a design goal to reduce the number of vias. They were "expensive" with older PCB construction methods. Nowadays the price for a square inch of PCB is the same for 1 via or 1000. So don't be afraid to use lots of them.

I see only 3 mounting holes and 2 of them aren't aligned with the grid. Does this fit some specific plastic box you have in mind? (Some of them have the internal mounting posts at strange coordinates.) If you're likely to use this box again for multiple projects, create that box as a library part and then you can drop the fully-tested hole locations onto any design.

Also look for through-hole parts that may fall on top of mounting posts in the box. Even if you're not planning to screw into that particular post, it makes the board sit crooked if there's a component lead poking through there. A "keepout" can be drawn in the library part to remind you not to put components in those places.

Your module connectors are nicely labelled but the outline of the module isn't. Once again, it's pretty quick to create a library part for a module and then it can appear on the schematic as a named block and on the PCB with an outline and name drawn on the silkscreen.

If those connectors at the bottom-right and bottom are off-board connectors (to cables) then this is going to be a nightmare to put into a box. Try to keep all connectors along one edge of the board. Two adjacent edges is acceptable. Three edges is too many for 85% of all designs. The reason is the physical layout in the box. An outdoor box should have no cable-entry points facing upwards where rain can get in. Cable entries on the sides are less weatherproof than putting the cable entry on the bottom. For indoor type boxes, you often mount them on a wall or they sit on the back edge of your workbench, so it's always a fight with the cables coming out on all sides of the box.

I'm guessing that the big connectors across the left and top are screw connector strips? It will be neater to place them as two strips in parallel (facing out from each other) than to have them at these odd angles. Try to put those strips at a strategic point between mounting holes because most of the mechanical stress is from your screwdriver and you don't want the bottom edge of the board flapping around with no mounting.

The module at the top appears to have no power?

Thanks for advice on the layout, it's very helpful. I'll probably reroute a lot of these traces to shorten them with the via suggestion in mind. Things like the mounting holes, lack of power to certain components, etc. are because this is a temporary design that I wanted to clarify some things on before continuing. I plan on fixing them shortly. I should also mention there is no particular container I have in mind for placing this in, likely what I will do is put it in a junction box along with some other components that don't fit as naturally in the circuit board (like the buck converter) as well as the terminal strips for outgoing connections where necessary.

Aside from these things, do any problems stand out to you electrical-wise that I may run into? In particular, will splitting the ground plane cause any unwanted side effects?

Make your pads and traces larger.

Looks like trace widths can be 30-50 mil.

Looks like pads can be 80 mil.

As for the shielded wires, you are saying to only connect the shield/metal part of the cable at the ground end, correct? Or did I misunderstand?

You did not misunderstand. If connected at both ends, the shield will be carrying a portion of the load current, hence 'ground loop'. If you are measuring voltages/whatever on conductive pipes/cables coming from different locations, then you will have to consider their ground connections and your equipment's star point ground, too.