My project started smoking.. not sure what’s wrong and if it can be fixed easily

• My project started smoking.. not sure what’s wrong and if it can be fixed easily
  Posted by paul_b_78 on 12 Jun, 2021 11:06
• Please see link to a project I have built.
  All was going so well until I forgot to plug the soldering iron into the 5 din socket.
  Arrrrrgghh
  It started to smoke a little and I’m sure it’s the irlz logic mosfet.
  Urgently could do with some assistance
  
  https://www.instructables.com/DIY-Digital-Soldering-Station/#discuss
  
  Any way I can run a test on my nano without pretty much destroying my built board to check it is giving me correct reading on all outputs.
  I am a novice but I am getting better at this.... well apart from frying something of course and choosing a board and enclose too small along with soldering which is not that great either.. so pretty much a complete fail.
  
  Also I followed most of the schematic but mainly worked off the Gerber / PDF drawing using a Perfboard and building the whole circuit from scratch. It’s not the neatest of jobs but I’m fairly confident I followed it exactly the same.
  It’s taken me quite a while to do it also as I don’t get to spend all hours each day on it due to other priorities. But I really would like to see if I can fix the issue rather than bin it
  
  Could send images of my board but I’m disappointed with my soldering since I chose a perfboard and plastic enclosure a little too small, and so things are really tight and rather messy.
  Hoping I don’t have to bin it as sent far too long on it

  ---

  
• #1 Reply
  Posted by Doctorandus_P on 12 Jun, 2021 11:13
• You could start by editing the title of your post to describe what you need help with.
• #2 Reply
  Posted by Manul on 12 Jun, 2021 11:22
• You forgot to plug iron and because of that something burned? I don't get it. Why would a soldering station self destruct if iron is not plugged? Where is schematic?
• #3 Reply
  Posted by Ian.M on 12 Jun, 2021 11:38
• If built to the schematic at your instructables link, leaving the iron disconnected could not damage the circuit.  With the temperature sensor disconnected, the OPAMP output would go as close as its able to +5V, which the code would 'see' as over-temperature and therefore shut off the MOSFET.  Also you can't break a MOSFET operating as a low side switch by not connecting a load to it.
  
  Therefore your circuit has let out the magic smoke for some other reason.  N.B. if the MOSFET has died, the odds are that its shorted internally between all three pins therefore, as its gate is wired direct to Arduino D3, its likely to have damaged the Arduino.
• #4 Reply
  Posted by paul_b_78 on 12 Jun, 2021 11:56
• The project did have an error in the begging which the owner amended but I was working off the Gerber file and recreated this to match. See image
  

  ---

  
• #5 Reply
  Posted by Zero999 on 12 Jun, 2021 17:54
• Instructables should be called destructables, because most projects on that site contain an error which can emit magic smoke. Don't get me wrong, that site has some good ideas, but one should always check first and never blindly build anything, without thoroughly understanding it first.
• #6 Reply
  Posted by paul_b_78 on 12 Jun, 2021 18:12
• @ Zero
  Thanks for the advice.
  I have stripped the whole thing down and now in the process of cleaning up the board and components, see what I can reuse and test the Nano board for all outputs. If that still works then I’m a lucky fella if not then it’s lesson learnt.
  
• #7 Reply
  Posted by paul_b_78 on 12 Jun, 2021 19:05
• Looks like I made a massive failure...
  I forgot to set up the Nano with the necessary code before I through in the 24V input..
  
  I have stripped that board down now and running some tests on components to see if I can reuse them.
  
  Thanks to all for your input though
  
  In future until I am able to understand electronics then maybe I need to either not attempt something so big and also send the created pcb file on here for someone to look over before i do any connections in future.
   
• #8 Reply
  Posted by Ian.M on 12 Jun, 2021 19:32
• Failing to program the Nano wouldn't cause MOSFET failure.  The MOSFET has a gate pulldown to keep it off if the Arduino pin isn't configured.   Worst case: if a  previous sketch made D3 an output, its possible the MOSFET would be turned on all the time, which would result in the iron overheating, but as long as you noticed it before it cooked its handle, its likely there'd be no permanent damage.   
• #9 Reply
  Posted by paul_b_78 on 12 Jun, 2021 20:57
• @ian M
  Thank you for that clarification.
  I’m starting to think I might have mis-followed the directions, although the schematic did not match the Gerber or PDF files so it all got a bit confusing
  
  I downloaded the Gerber file to Gerb Magic and my board followed the bottom copper, I can’t see what’s gone wrong.
  
  I am really baffled.
  
   
  
• #10 Reply
  Posted by Labrat101 on 12 Jun, 2021 21:27
• There is the possibility you picked up a bad MosFet to start with if you got it from China & co.
  you may have been unlucky to get a fake . MosFets are pretty hardy . Not to say they can't be destroyed.  If  it did go complete short . you may be lucky it may only have damaged one port .
  the nano does have internal protection . Just try another port and change the program to show the
  new port number .  And if you bought more Mosfets from the same supplier check the marking to
  see if its a genuine one . from the manufactures data sheet . the print should be lazer printed not
  able to be washed off with Acetone .
   Good Luck 
• #11 Reply
  Posted by WattsThat on 12 Jun, 2021 22:03
• When building one-off projects on perf board, strip board, whatever your method, do not build the entire unit, apply power and hope for the best. That’s a recipe for failure because when you’re starting out, you will make mistakes. Mis-wire a power supply and you will absolutely let the smoke out of something.
  
  Build your circuit in blocks. Wire up the 24 vdc buck converter. Solder a resistor on the output. Power it up. Set the correct output voltage. Check the ripple voltage. Turn it off. Wire the Nano and the LED. Change the Blink demo to use the external led. Power it up, download the program and test. Get the Nano running. Install and connect the lcd. Write or find a hello world sketch and test the lcd. Build the opamp. Connect a pot in place of the thermistor. Download the final program. Test the opamp and the analog input. Only when the circuit works, add the mosfet. Then you can connect your soldering iron.
  
  Se how that works? One step at a time. Get each functional block working, one at a time. Individual components you can test before installation. Resistors, caps, transistors and similar parts - test them before installing. Only when you’ve got some experience can you just throw the whole thing together and have a reasonable expectation it will work on the first power up.
• #12 Reply
  Posted by paul_b_78 on 12 Jun, 2021 23:03
• @watts that
  I completely see what you mean with your theory.
  
  This has been noted and will be my process in future.
  
  Just one last question.. when you say add a resistor to the output..what type / size resistor and on what side of the buck converter.
  I have an oscilloscope but never used it so guess now is the right time.
• #13 Reply
  Posted by paul_b_78 on 12 Jun, 2021 23:07
• @ labrat101
  
  My Nano only came with 1 port
  I have even tried a second Nano which isn’t connected to anything and that also shows the port greyed out on Arduino IDE program on my laptop
  What am I doing wrong... can’t even program the correct port for Laptop to read Nano
  
• #14 Reply
  Posted by WattsThat on 12 Jun, 2021 23:53
• You would want to load the power supply to at least what your project will consume, if possible and practical. Sometimes it is, other times not practical and finally, some times where it’s not necessary to test. One of those cases would be when it’s a fixed output and it’s a complete, pre-manufactured & tested module.
  
  Many linear regulator ic’s require a minimum load, most times 10-20ma is more than enough. If in doubt, check the datasheet for the part.
  
  A typical 328 based Arduino will be around 50-100ma max plus the i/o loading so 250 milliamps is more than adequate for a test. R=E/I so 5/.25 = 20 ohms. That will dissipate E*I watts so you’d want a 2 watt resistor so it doesn’t overheat. Mind you, it will get toasty so don’t go squeezing it. If all you have are 1/4 w devices, go with two in parallel and keep the power below 50% of the rating so you can reuse the resistors. Push them past their power rating for any length of time and they will probably drift in value.
  
  A different Nano may show up on a different com port. Go to the tools menu and then port to see if the new board pops up. It should be there.
  
• #15 Reply
  Posted by Labrat101 on 13 Jun, 2021 08:15
• I think you have made a mistake drawing that much power from a nano I/O you will sure fry the outputs . Max out is 1 led .
  5 or 10ma . If it's a clone.  For sure not .. .  330 ohm I believe is the safe load on an i/o .
  A led is better for testing.  As you are a novice.  .
  
• #16 Reply
  Posted by WattsThat on 13 Jun, 2021 18:13
• Quote from: Labrat101 on 13 Jun, 2021 08:15

  I think you have made a mistake drawing that much power from a nano I/O you will sure fry the outputs . Max out is 1 led .
  5 or 10ma . If it's a clone.  For sure not .. .  330 ohm I believe is the safe load on an i/o .
  A led is better for testing.  As you are a novice.  .
  

  
  
  We’re talking about testing a power supply here, not a Nano.
  
  Regardless, your comments are quite misguided and flat out wrong. Being a clone board has no impact on its current specifications. It’s the design and the parts that determine the specs, not who puts it together.
  
  The atmega328 is limited to a maximum current of 200ma, all pins combined. The practical limit per pin is 20ma, with the absolute max at 40ma.
• #17 Reply
  Posted by paul_b_78 on 14 Jun, 2021 09:22
• @wattsthat
  I managed to get the IDE to recognise Port Com4.. I’m now going to try out your instructions, by starting over and adding the circuit in blocks.
  I have a DC power supply so I had already done the necessary fine adjustments that way.. that’s how I knew to adjust it to the 6.5v output.
  What I realised though is that the Instructable is miss-leading...
  I had never wired up an LCD before and with the Instructable it shows 2 methods but as one.....so I had wired the LCD connections along with the I2C connections together. That of course wasn’t causing the smoking, but now wondering if there are some other issues with the Instructable which caused my project to start smoking.
  Note to self: maybe before I follow an Instructable in future, to send it over here for an experienced member to over look and see if it has flaws.