Hello, this is my first post in the forum. I'm looking for some insight from experienced users/ designers.
I've been working on repairing an automated bag filling machine that a former coworker built. It is pretty simple in terms of coding and electronics (just Arduino UNO, Hx711 module, a load cell, a relay module and a pair of buttons). Although this machine works, it is prone to malfunction, and I struggle to understand the nature of its problems. Like I calibrate it, try it out and tune it until it works fine, just for it to break down again a few days later, and opening the enclosure to upload sketches usually messes up some of the connections of other components. I figure this is related to the makeshift nature of the circuitry. So my question is how do you go about stepping up your proven, working prototypes to circuits that are reliable and not-so-makeshift (first thing I thought of was making a pcb but I'm not entirely sure if it can be done with an Arduino UNO, or if I would need to go with just AT Mega). Also, what are your prefered methods for feeding arduino boards in actual working projects, my former coworker was using small phone chargers to get 5V from 220V (pic related)? What type of casing of enclosure do you use for keeping all your circuitry together? I have this electrical box where all components are screwed or hot glued, and I feel like there should be a finer way (pic related). Finally as general advice, how to ensure reliability from the circuits/projects that one may develop, as in having your circuit function as it is intended every time (or as close to) someone uses it.
I understand and appreciate all your questions, but this quoted statement makes me wonder.
What kind of break downs or malfunctions do you get after a few days?
Why do these breakdowns require changes to the firmware?
a7
It looks like there are two Uno's there. Why?
Does that make it a Due?
No. A DuemilanUno perhaps though 
HX711 is prone to drift. This may be the source of the problem.
As for the wiring, LarryD has a good start: attach everything to an internal panel, not the actual enclosure. It makes it much easier to make changes.
- Arrange wiring neatly.
- Solder or crimp connections. Barrier strips as shown are also good, but they should be secured, not hanging in midair. If you're using any kind of screw terminal, either use solid core wire, or ferrules on stranded wire.
- Do not just stick "du Pont" wires into connectors. You can use them, but zip tie them securely so they can't come loose.
- To expand on that nothing should be floating in midair. Assume it's going to be tossed in the trunk of a car or fall on the floor: parts should not be dislodged easily.
Here's a good reddit thread on panel building: Reddit - Dive into anything
Also, for really well done wiring (as a goal for perfection) visit Reddit - Dive into anything (yes, it's SFW).
You could make the usb terminals available from outside the box (without opening it). That would save you a lot of hassle inside this narrow box.
Personally I would scrap the whole thing and start again with an ESP32,on a PCB, use screw terminals rather than Dupont connectors, connect to it wirelessly, power it with a 'phone charger and use OTA to update it if required
However, I do understand that this is probably not going to happen !
OP says it is a prototype... so there is hope...
Option 1. Gradually improve prototype.
Option 2. Start over again and take learning points from prototype into account.
Both options are well tested in engineering. Sometimes option 2 can bring you a jump forward, sometimes it can introduce new unforeseen problems...
(Boeing 737 max vs JSF)
A dab of hot-melt glue can help hold the connections in place.
Start again ONE Nano or Nano Every. Use breadboard for testing then, when you are happy it works as required, make a PCB.
What is your hardware + coding proficiency; I am asking if you completely understand this contraption? Is the 2nd UNO a "hot spare"?
If you completely understand the circuitry and software, would you be comfortable rebuilding the prototype if employer paid for Arduinos and misc such as decent power supply, proper enclosure, etc.
Or, did prior employee take the 'need' upon him/her-self to build and integrate this unit? Are you in an environment where OSHA/public safety/large equipment maintenance agreements apply? (Often, it is a violation to modify industrial machinery.) Does any liability exist for failure: downtime, slower production, etc.
These kinds of projects are often improvised explosive devices in terms of OSHA, liability insurance, municipal codes, just to name a few.
If your employer has assumed all liability for this device being "in-use" and you are paid/expected to maintain it, then you have a choice to make: 1) continue as-is, 2) ask your employer to provide in writing the nature of this unit (critical, nice-to-have, non-critical, remove asap...), or 3) take on a rebuild and request employer to formalize that in writing stating they will pay for parts and authorize your new build.
I guess #4 would be to find a new job.
Whatever you decide, bear in mind that pin-jumper wire is completely inappropriate... there are 'shields' available to provide solder/screw-down connections. The current AC wall-wart is in violation of several best practices, but primarily it is illegal in many jurisdictions to use an electrical device for any-other-use than the originally certified/approved use.
Good luck.
Ray
There are actually 2 filling machines which use load cells, each UNO operates a machine independently tho they share some of the 220 V cables
This can come in handy. Thanks!
Thanks a lot! This is useful info. Do you have any recommendations for an Hx711 replacement?
Why ESP32? Enlighten me please.
And yes, probably not gonna take it that far but still appreciate your suggestion!
My hardware experience has mostly been limited to lab-level stuff, thats why I'm struggling with coming up with something that works in an actual work environment. As of coding: usually dont write code so I just take bits I find online/have written myself before, but I am able to understand and modify code to suit my needs. In regards to this contraption I have managed to have a good grasp of how it works. I am indeed trying to rebuild this prototype and employer does provide funding (yet I'm not even sure what type of hardware to look for in terms of power supply, enclosure, etc).
OSHA/public safety/other agreements don't apply, only liability is having to fix the machine every time it breaks down.
The machine works with a load cell and I usually have to tune its scaling and weight limit ( when this weight is reached the loading valve will close). Even if I dont upload firmware, it is useful to connect the UNO to a computer in order to use the serial monitor and get an idea of what may be causing trouble (other times when the only reading I get from the load cell is -42, I have a hint that some cable may have come loose, etc.)
That is why LCD displays are used!