Wire shielding problem? Too much wires and too much close each others?

Projects General Guidance
1

Hi guys!

Still in trouble with the second part of my first project. Before starting to unwire everything and check individually each component, I would like to try to understand if I can solve the issue in a quicker and smarter way.

I am developing a sensor box, with many sensor components, linked to an ESP32 Dev Board (AZDelivery ESP32 NodeMCU DevBoard WiFi CP2102) that sends data to an arduino mega 2560 r3.

The problem is related to the ESP32 box, and its many sensors (attached you can find a photo of the mess and I have tried to replicate the actual scheme).

In particular my problems are:

  • screen blinks slightly every 5 seconds more or less (working great before connecting al the sensors)(not a big issue but I think it is a symptom of something not working properly)
  • one sensor seems to fail to react, even if it shows apparently initial correct values (I have written a post related to the sensor itself, not yet received a feedback (SI1145 Grove Digital Light Sensor - not reactive - Sensors - Arduino Forum)
  • three sensors seem to work properly (SCD30, Dust sensor and DS18B20)
  • the other sensors don't show evident "error" (like I don't know, return simbols, return error message and so on), but readings are not correct.
    So I have tried to play with them,(calibration, checking the code and so on), but nothing seems to work

I have spoken with an arduino expert vet, and in his opinion the problem is due to so many wires too close each others (you can see the mess in the attached photo) that probably create Electromagnetic Interference that disturbs some sensors readings.

Honestly, given my background (anyway nonexistent for what concern electronics) I don't know how to fix the issue.

At first I think I should reduce wire lenghts as much as I can, but I fell it will not be sufficient.

What is your opinion? Any suggestions to solve the issue in a smart and relatively quick way?

Any idea is welcomed!

Thank in advance for your support!

Schematic_MyProject_2020-05-17_14-05-24.pdf (351 KB)

64f98dd1-1b1b-4b54-9e0c-6ff23814a0b1.jpg
64f98dd1-1b1b-4b54-9e0c-6ff23814a0b1.jpg1600×1200 245 KB

0.jpg
0.jpg2016×1512 551 KB

2

Images from Original Post so we don't have to download them. See this Simple Image Posting Guide

64f98dd1-1b1b-4b54-9e0c-6ff23814a0b1.jpg
64f98dd1-1b1b-4b54-9e0c-6ff23814a0b1.jpg1600×1200 245 KB

0.jpg
0.jpg2016×1512 551 KB

For the same reason please post your schematic as an image and make it visible in your next Reply

...R

3

Please post a link to the datasheets for each of your sensors.

Crosstalk between the wires might be the problem but, looking at your pictures, it would not be the first thing I would suspect. Poor connections, inadequate power or a program problem seem more likely.

It should be possible to confirm or dismiss the problem of crosstalk by laying the parts out as separately as possible on the workbench.

...R

4

I see several problems and I don't see any test equipment. How can you fix a problem if you don't know what or where the problem is?

What I see is you are placing everything on a static electricity generator: paper/cardboard.

What I see is the use of plastic breadboard with wires placed in adjacent pins. The brass inserts in the bread board run parallel to each other and make wonderful capacitors. Separate each wire connection with at least an unused spot between each, or better yet, a grounded set between each wire.

Paul

5

The multiple rolls of insulation tape are intriguing too!

6

Hi guys,

First of all thx everybody for your feedback and suggestions!

Sorry but it's just a couple of months that I have started with arduino and electronics, so more than a newbie.

Moreover as economist, a "non-scientist" of a no science, I have just learnt to maximize return minimizing efforts... My bad! Trying to fix it!

After succeeding in the first part of the project (a lot of problems also there but I have fixed all) I got caught up in the excitement and after wired 3 sensors and the screen on the second part of the project (and they were working as expected), I started to wire everything without being patient...

So pls forgive me and try to appreciate the good intentions and the hard work!

I have noticed that the photo I sent of the opened box was not updated (not all wired indeed), here you are more details of the situation:

![](88f804e5-bcdb-4712-b121-25f421762b64.jpg - Google Drive, faf34e10-70bc-488f-9598-65e392408d6d.jpg - Google Drive, 737f8499-fc97-4e36-afd3-fd28a90456a0.jpg - Google Drive, 6051a385-84ac-4fa0-9fca-6a07cb99095e.jpg - Google Drive)

So, I have tried to separate as much as possible jumpers on the breadboard as suggested by Paul.

I have checked the voltage on the 3.3V (principally connected to the blinking screen) and on the 5V, and I found respectively 3,16V and 4,85. Maybe the blinking screen is due to this and maybe also the rest of the 3.3V linked sensors are experiencing a voltage drop. So I have tried to wire them on the 5V but

Where you can see the funny wiring (brown yellow red with a resistor between) is linked the temp sensor, and it is one of the only sensors working properly! :confused:

As requested here you are the links to sensors datasheets:

that's all... I know guys, just a mess... I have arrived pretty exhausted at the end of the project and once understood the different connection type (I have there SPI, I2C, Analgo and digital connections), made the scheme, wrote the code, I thought wiring should not be the problem... But I was of course wrong!

Pls be patient and tell me everything I should to to understand where could be the problem! (I don't post the code because as you can imagine it is pretty long and moreover I have just followed library example to test the sensors.

Thx so much!

7

Can you please change your pasted web addresses into clickable links? You have posted many, and most forum members use phones and tablets most of the time. Using touch screens to select those pasted addresses, copying, opening new tabs and pasting the addresses is slow and frustrating. But for the person posting the links, one extra click turns the address into a clickable link, making it easy for everyone.

8

Not able to connect to forum until now. I have just noticed that photos have not been uploaded.

Let me try again:

Btw I am proceding to disassemble everything and checking each component one by one.

I have noticed that if let the SCD30 (working in each case) and the Digital Light (not reactive when everything wired), the latter seems to work (becomes reactive) even if the minimum value (if I put him in the darkness) still remain around 260 lux.

Anyway I understand that find the problem with all this mess is impossibile, my question was more related to generaladvice, just like the fact that paper is not the best case material or jumper have to stay as far as possible on the breadboard.

I understand that the problem should be inside the code, or due to poor wiring, but I was just checking if I was making error on the basis stuff.

Thx for you patience, keeping try to fix it, anyway any advice is more than welcomed!

Cheers and have a great start of week!

9

Oh I still don't see my images, but I suppose you are able to see them.

I follow PaulRB advice, thx! I have still a lot to learn!

10

Hello Fabion,

Well done for getting this far on your project.

From the description of the problem I agree with R2 in reply #2, I think cross talk is not the most likely candidate as the cause of your problem.

Breadboard is wonderful for experimenting, useless for finished projects. Can you solder? If not you really need to learn. Until you are ready get a PCB made I suggest you build your finished project on stripboard. Soldered connections are far more reliable than wires poked into breadboard or covered in tape.

11

Hi Perry,

Thx so much for your encouragement and your advice! I am just noticing that jumpers are not the most reliable wiring system...!

Infact as I touch wires they start to give false signal and problems (also for sensors that usually work)...

I was thinking to start soldering after everything works with breadboard and jumpers, but given the relatively complexity of the project (in particular the fact that they need to be very closed inside the box) I think the time has come! (if you can help me finding a good welding torch pls let me know)

Anyway I am still trying to do some tests, good occasion to understand a bit more (just a couple of months on arduino and electronics... so the journey is just at the beginning!).

Now I am trying back with just SCD30 + Digital Light + Screen: screen sometimes start to blink when I connect the two sensors and their contacts are not good!

I didn't think screen should be impacted by others connections, but seems that poor connection of other sensors impact also screen that otherwise seems well connected.

So the game become more than tricky and I understand why it is really impossible to understand what is going wrong unless I connect and test each component one by one!

As usual more feedback on what I can improve are more than welcomed! Thx so much!

12

FABION:
(if you can help me finding a good welding torch pls let me know)

A welding torch? Errmm, possibly start with a soldering iron! There are many about, but try to go for a well known name. Antex make some good irons for not a lot of money. If you can afford a bit more, maybe a Weller soldering station, used. If you can splash out, then a nice new Weller would do you proud.

Solder is just as important. For someone new to soldering, don't even consider lead free. Go for a good brand, like Ersin Multicore of 60/40 Tin/Lead on .7mm or there abouts.

13

My soldering iron is a Weller W61. Mid price, reliable and bought after my previous 30 year old Weller iron died.

I didn't think screen should be impacted by others connections, but seems that poor connection of other sensors impact also screen that otherwise seems well connected.

If your system is manifesting some problem or problems, and at the same time there are things obviously wrong with the system, so in this case the things you list in your first post and the obvious loose wiring, don't dismiss the things that are obviously wrong as not being the cause of the problems you see. Fix the obvious stuff first and then see what problems, if any remain. You will probably find that when you fix the obvious stuff then either the faults will go away, even if you never conclusively establish the link between the cause and the fault, or you will at least reduce the number of problems you have to deal with, making whatever is left easier to fix.

14

My 40-year old Weller station is still running fine. And I have another one on the shelf for when it quits working.

OP- You found the problem- you just don't realize it. "Infact as I touch wires they start to give false signal ".

The breadboards are notorious for intermittent connections.

Also, I would like to see a schematic diagram. Pictures of components with wires drawn between them is not a schematic. Yours communicates practically nothing. Especially when the parts have no labels or descriptions.

15

Hi guys!

Sorry for the late reply, but I was trying to fix the mess while remote working (fortunately I am trader so the story of "not good opportunities arisen" ever works with my boss...).

First of all, please tell me what a good schematic should be for an engineer: I just try to show how I wired them taking the picture of each component like a child doing bricolage!

Secondly, let me try to sum up my discoveries! Please don't shoot on me, but I think they will help other puppets like me in the future!

One big problem (I didn't think it was but now I realize it) it was that I have tested my analog sensors on MEGA, and pretended they worked on ESP32 without adeguate the analog output bits... I was still using .../1023 instead of /4095...

So a big point in favour of most of you (and also to my quote "REMEBER: THE ANSWER IS INSIDE YOU(R CODE)!!") that suggested me to check everything else a part EMI lol...

After fixed this, I got another "surprising" effect! The PH sensor was not working when I run the complete code, but it was working great if I run its code alone...

So I started the debug (yday for the precision lol) of the main code (more than 500 lines and plenty of library :S), line by line, section by section, and I figure out that this sensor was not working once WiFi was on! (I was using pin G25)(I had a lot of issue last week to link the SPI screen and I2C sensors at the same time, due to the fact that some pins were used by flash memory and other stuff (but my question from a super-newbie point of view is: why having pins if I cannot use them? but leave this question to another post in the future lol))

I changed the pin to G35 and now everything seems to be working fine, despite the fact that in order to check everything I "destroyed" the first wiring scheme that most of you blamed as the first cause, and now it is working in an even worst situation lol

My question is, how I should have checked the fact that using G25 creates conflict with WiFi (a part getting mad and debug it line by line)?!

Here you are the diagram of the ESP32 I am using:

3393×2476 826 KB

Anyway now it is time to SOLDER and make this prototype reliable!

Thanks so much for your patience, trust and time!

F

16

First of all, please tell me what a good schematic should be for an engineer:

Read other posts on these fora, there are lots of schematics. Here's a simple one of mine common ground tutorial. Components drawn as simple outlines with pins labelled according to function. Lines to represent wires joining them up. Highest voltage at the top (usually Vcc), lowest voltage at the bottom (usually ground or 0V). Signals moving from left to right. Pencil and paper then photographed is fine. The two usual pieces of software used are Eagle and Kicad, both are free. No, I don't know which is better, I use Kicad, never tried Eagle so I can't compare.

Why having pins if I cannot use them?

A typical micro-controller has lots of built in hardware functionality, such as serial ports, SPI and I2C ports, input and output pins, analogue pins and lots more. If every function had its own dedicated pin there would be a ridiculous number of pins, so instead 1 pin will have several possible functions. However, that means if you use it for one function it's not available for another function. The manufacturer of the chip doesn't know what you are going to use the chip for, so they make it as flexible as they can so you can decide what you need to use. Typically there will be several hardware function on a chip that you don't use. Have a look at Microchip's web site and the range of different chips they have. They are aiming to have one that does what you need whatever the application, and they have a huge range. That's chips, when you buy a ready made board like a Uno or NodeMCU then someone else has made the decision for you about the choice of chip, and some of the compromises of which pins are available for what have already been made by the board designer.